Consensus recommendations for a standardized brain tumor imaging protocol for clinical trials in brain metastases

Semi-Supervised Learning Allows for Improved Segmentation With Reduced Annotations of Brain Metastases Using Multicenter MRI Data

BACKGROUND

Deep learning-based segmentation of brain metastases relies on large amounts of fully annotated data by domain experts. Semi-supervised learning offers potential efficient methods to improve model performance without excessive annotation burden.

PURPOSE

This work tests the viability of semi-supervision for brain metastases segmentation.

STUDY TYPE

Retrospective.

SUBJECTS

There were 156, 65, 324, and 200 labeled scans from four institutions and 519 unlabeled scans from a single institution. All subjects included in the study had diagnosed with brain metastases.

FIELD STRENGTH/SEQUENCES

1.5 T and 3 T, 2D and 3D T1-weighted pre- and post-contrast, and fluid-attenuated inversion recovery (FLAIR).

ASSESSMENT

Three semi-supervision methods (mean teacher, cross-pseudo supervision, and interpolation consistency training) were adapted with the U-Net architecture. The three semi-supervised methods were compared to their respective supervised baseline on the full and half-sized training.

STATISTICAL TESTS

Evaluation was performed on a multinational test set from four different institutions using 5-fold cross-validation. Method performance was evaluated by the following: the number of false-positive predictions, the number of true positive predictions, the 95th Hausdorff distance, and the Dice similarity coefficient (DSC). Significance was tested using a paired samples t test for a single fold, and across all folds within a given cohort.

RESULTS

Semi-supervision outperformed the supervised baseline for all sites with the best-performing semi-supervised method achieved an on average DSC improvement of 6.3% ± 1.6%, 8.2% ± 3.8%, 8.6% ± 2.6%, and 15.4% ± 1.4%, when trained on half the dataset and 3.6% ± 0.7%, 2.0% ± 1.5%, 1.8% ± 5.7%, and 4.7% ± 1.7%, compared to the supervised baseline on four test cohorts. In addition, in three of four datasets, the semi-supervised training produced equal or better results than the supervised models trained on twice the labeled data.

DATA CONCLUSION

Semi-supervised learning allows for improved segmentation performance over the supervised baseline, and the improvement was particularly notable for independent external test sets when trained on small amounts of labeled data.

PLAIN LANGUAGE SUMMARY

Artificial intelligence requires extensive datasets with large amounts of annotated data from medical experts which can be difficult to acquire due to the large workload. To compensate for this, it is possible to utilize large amounts of un-annotated clinical data in addition to annotated data. However, this method has not been widely tested for the most common intracranial brain tumor, brain metastases. This study shows that this approach allows for data efficient deep learning models across multiple institutions with different clinical protocols and scanners.

LEVEL OF EVIDENCE

3 TECHNICAL EFFICACY: Stage 2.

Diagnostic Confidence of Contrast-Enhanced T1-Weighted MRI for the Detection of Brain Metastases: 3D FSE versus 3D GRE-Based Sequences

BACKGROUND AND PURPOSE

This retrospective study evaluated the utility of contrast-enhanced T1-weighted 3D fast spin-echo-based sampling perfection with application-optimized contrasts by using different flip angle evolutions (SPACE) sequences for brain metastasis detection on 3T MRI compared with a gradient-recalled echo-based 3D FLASH sequence.

MATERIALS AND METHODS

We identified all patients at a single institution who underwent SPACE and 3D FLASH sequences as part of a practice quality-improvement project. Their medical records were retrospectively reviewed. Five certified neuroradiologists reviewed the images, with at least 2 weeks' separation between scoring sequences for the same patient. We evaluated the following parameters: number of metastatic lesions, number of indeterminate lesions, lesion margin, contrast-to-noise ratio (CNR), extent of image artifacts, and overall image quality. The CNR was also quantified for solidly enhancing lesions of >1 cm.

RESULTS

We identified 220 patients who underwent SPACE and 3D FLASH sequences (the order of the sequences was equally distributed). Of these, 79 had brain metastases on imaging, and 7 were excluded; thus, 72 patients were included in the study. Twenty patients were scored by 2 radiologists. Of the 92 evaluations, SPACE detected more lesions than 3D FLASH in 35, while 3D FLASH detected more lesions in 10. More indeterminate lesions were seen on 3D FLASH (n = 27) than on SPACE (n = 9). For the lesion margin, CNR, and overall image quality on a Likert scale, SPACE performed significantly better than 3D FLASH, with fewer image artifacts (P < .00001). Higher quantitative CNRs were found on SPACE than on 3D FLASH images, though this result was not statistically significant (median = 22.9 versus 15.5, respectively, P = .134). There was a high interreader lesion detection concordance with the Krippendorf α ordinals at 0.962 for SPACE, 0.870 for 3D FLASH, and 0.918 for the 2 sequences combined.

CONCLUSIONS

Compared with 3D FLASH, the SPACE sequence detected more metastatic lesions and was rated higher for image quality, lesion margin, and CNR, with fewer artifacts. Importantly, the SPACE sequence resulted in increased reader confidence, with fewer indeterminate lesions detected.

Contrast Agent-Specific Parameter Optimization for T1-Weighted Fast Spoiled Gradient Echo Imaging: Use Cases for Gadoterate Meglumine and Gadobutrol at 1.5T and 3.0T

OBJECTIVES

The objective of this study is to derive a contrast agent-specific theoretical framework to optimize acquisition parameters for contrast-enhanced magnetic resonance imaging (MRI) based on the physiochemical properties of gadolinium-based contrast agents, focusing on fast spoiled gradient recalled echo sequences. The goal is to enhance the lesion-to-background contrast for improved diagnostic sensitivity in clinical applications.

MATERIALS AND METHODS

Signal equations for fast spoiled gradient recalled echo sequences were derived for nonenhancing and enhancing tissues using gadoterate meglumine and gadobutrol, characterized by distinct longitudinal (r1) and transverse (r2/r*2) relaxivities. Simulations were conducted at 2 field strengths, 1.5T and 3.0T, and various scenarios were considered, including hypothetical lesions with T1 ratios ranging from 1.1 to 1.8. The signal behavior was analyzed across a range of initial conditions, including different spin densities and field-strength dependent variations in tissue relaxation times. The optimal flip angle and repetition time combinations were determined to maximize contrast. In vivo validation was performed on 2 patients undergoing contrast-enhanced MRI of the brain, using the proposed acquisition parameters.

RESULTS

The modeling and simulations revealed that the flip angle that maximizes signal intensity for a contrast-enhancing lesion (Ernst angle) differs from the flip angle that maximizes T1-dependent contrast between lesion and healthy tissue in unenhanced MRI (Pelc angle) and also differs from the flip angle that maximizes the same in contrast-enhanced MRI. The theoretical simulations indicated possible contrast gains of 24%-28% using optimized parameters. The in vivo acquisitions demonstrated contrast gains of 19%-44% for a diffuse enhancing lesion and 91% for a weakly enhancing focal lesion, when comparing the optimized acquisition parameters to manufacturer's default settings.

CONCLUSIONS

Adjusted repetition time and flip angle values, derived using the proposed framework, improved the image contrast between healthy and diseased tissues, enhancing the visualization of abnormalities. This approach can be used to optimize routine clinical MRI protocols and balance scan time with contrast enhancement. This may translate to more precise lesion detection, potentially leading to earlier and more accurate diagnosis or treatment monitoring in clinical practice.

Metastasis Detection Using True and Artificial T1-Weighted Postcontrast Images in Brain MRI

OBJECTIVES

Small lesions are the limiting factor for reducing gadolinium-based contrast agents in brain magnetic resonance imaging (MRI). The purpose of this study was to compare the sensitivity and precision in metastasis detection on true contrast-enhanced T1-weighted (T1w) images and artificial images synthesized by a deep learning method using low-dose images.

MATERIALS AND METHODS

In this prospective, multicenter study (5 centers, 12 scanners), 917 participants underwent brain MRI between October 2021 and March 2023 including T1w low-dose (0.033 mmol/kg) and full-dose (0.1 mmol/kg) images. Forty participants with metastases or unremarkable brain findings were evaluated in a reading (mean age ± SD, 54.3 ± 15.1 years; 24 men). True and artificial T1w images were assessed for metastases in random order with 4 weeks between readings by 2 neuroradiologists. A reference reader reviewed all data to confirm metastases. Performances were compared using mid- P McNemar tests for sensitivity and Wilcoxon signed rank tests for false-positive findings.

RESULTS

The reference reader identified 97 metastases. The sensitivity of reader 1 did not differ significantly between sequences (sensitivity [precision]: true, 66.0% [98.5%]; artificial, 61.9% [98.4%]; P = 0.38). With a lower precision than reader 1, reader 2 found significantly more metastases using true images (sensitivity [precision]: true, 78.4% [87.4%]; artificial, 60.8% [80.8%]; P < 0.001). There was no significant difference in sensitivity for metastases ≥5 mm. The number of false-positive findings did not differ significantly between sequences.

CONCLUSIONS

One reader showed a significantly higher overall sensitivity using true images. The similar detection performance for metastases ≥5 mm is promising for applying low-dose imaging in less challenging diagnostic tasks than metastasis detection.

Expert Panel on Neurological Imaging, Ivanidze J, Shih RY, Utukuri PS, Ajam AA, Auron M, Chang SM, Jordan JT, Kalnins A, Kuo PH, Ledbetter LN, Pannell JS, Pollock JM, Sheehan J, Soares BP, Soderlund KA, Wang LL, Burns J. ACR Appropriateness Criteria® Brain Tumors

Brain tumors represent a complex and clinically diverse disease group, whose management is particularly dependent on neuroimaging given the wide range of differential diagnostic considerations and clinical scenarios. The introduction of advanced brain imaging tools into clinical practice makes it paramount for all treating physicians to recognize the range and understand the appropriate application of various conventional and advanced imaging modalities. The imaging recommendations for neuro-oncologic clinical scenarios involving screening in patients with increased genetic risk, screening in patients with systemic malignancy, pretreatment evaluation in patients with intra- and extraaxial brain tumors, posttreatment-surveillance in patients with known brain tumors after completion of therapy, and subsequent workup in the context of suspected radiographic progression are encompassed by this document. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision process support the systematic analysis of the medical literature from peer reviewed journals. Established methodology principles such as Grading of Recommendations Assessment, Development, and Evaluation or GRADE are adapted to evaluate the evidence. The RAND/UCLA Appropriateness Method User Manual provides the methodology to determine the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where peer reviewed literature is lacking or equivocal, experts may be the primary evidentiary source available to formulate a recommendation.

Interleaved flow-sensitive dephasing (iFSD): Toward enhanced blood flow suppression and preserved T1 weighting and overall signals in 3D TSE-based neuroimaging

PURPOSE

To develop and validate a 3D turbo spin-echo (TSE)-compatible approach to enhancing black-blood (BB) effects while preserving T1 weighting and overall SNR.

METHODS

Following the excitation RF pulse, a 180° RF pulse sandwiched by a pair of flow-sensitive dephasing (FSD) gradient pulses in the phase- (y) and partition-encoding (z) directions, respectively, is added. The polarity of FSD gradients in z direction is toggled every TR, achieving an interleaved FSD (iFSD) configuration in y-z plane. The technique was optimized and evaluated in 18 healthy volunteers and 32 patients with neurovascular disease or brain metastases. Comparisons were made among TSE with and without one of BB preparations: iFSD, delay alternating with nutation for tailored excitation, and motion-sensitized driven equilibrium.

RESULTS

iFSD-TSE achieved the best blood flow suppression indicated by venous sinus SNR and parenchyma-to-sinus contrast-to-noise ratio (CNR). iFSD-TSE yielded slightly lower white matter SNR (106.6 ± 32.9) and white-to-gray matter CNR (27.3 ± 8.1) compared to TSE (111.4 ± 31.5 and 28.6 ± 8.8), which were significantly higher than those of delay alternating with nutation for tailored excitation-prepared TSE (84.3 ± 25.0 and 16.8 ± 4.8) and motion-sensitized driven equilibrium-prepared TSE (77.3 ± 26.6 and 15.9 ± 5.3). At the neurovascular wall lesions, iFSD-TSE yielded the highest wall-to-lumen CNR among the three sequences with a BB preparation, all of which significantly outperformed TSE. iFSD-TSE effectively suppressed slow-flow artifacts that otherwise mimicked an atherosclerotic lesion or strongly contrast-enhancing vessel wall. In diagnosing brain metastases, iFSD allowed for highest inter-reader agreement (κ 0.75) and shortest reading time.

CONCLUSION

iFSD is a promising approach compatible with 3D TSE for robust blood flow suppression and preserved T1 weighting and overall SNR.

Pearls and Pitfalls of T1-Weighted Neuroimaging: A Primer for the Clinical Radiologist

All T1-weighted images are built upon one of two fundamental pulse sequences, spin-echo and gradient echo, each of which has distinct signal characteristics and clinical applications. Moreover, within each broadly defined category of T1-weighting, acquisition parameters can be modified to affect image quality, contrast, and scan duration; each tailored sequence has unique advantages, drawbacks, clinical indications, and potential artifacts. In this review, we describe key features that distinguish different types of T1-weighted sequences and discuss the utility of each sequence for specific clinical settings, including neuro-oncology, vasculopathy, and pediatric neuroradiology. In addition, we provide case examples from our institution that illustrate common artifacts and pitfalls associated with image interpretation. The findings described herein provide a framework to individualize the imaging protocol based on patient presentation and clinical indication.

Volumetric and diffusion MRI longitudinal patterns in brain metastases after laser interstitial thermal therapy

OBJECTIVE

To characterize MRI changes of brain metastases (BM) following laser interstitial thermal therapy (LITT), particularly in lesions exhibiting durable response or early progression.

MATERIALS AND METHODS

Longitudinal scans from patients with LITT-treated BM were retrospectively analyzed. Treatment response was categorized as durable response, long-term disease control (i.e., stable at 1 year), stable disease < 1 year, or progression < 1 year. Volumetric and diffusion MRI changes after LITT were analyzed for each subregion (contrast-enhancing, central non-enhancing, whole lesion). Volumetric changes were modeled with bi-exponential fits in responding lesions and progressors.

RESULTS

295 MRI scans from 47 lesions across 42 patients (57.8 ± 14.3 years old, males:females 21:21) were analyzed. Overall, the post-LITT scan showed a lesion enlargement (p < 0.0001 for all subregions), more pronounced in the contrast-enhancing (CE) component (median = +77%, p < 0.0001), and a reduction in the apparent diffusion coefficient (ADC) (p < 0.001), especially in the central non-CE component (median = -224 × 10-6 mm2/s, p < 0.0001), with no significant differences between responders and progressors. Based on mathematical modeling, the responding lesions shrank to half of the post-LITT size after 79.83 days (median "pseudo-half-life"), and the progressing lesions shrank for a median of 27 days (median time-to-growth) before regrowing. The estimated optimal timepoints for follow-up scans were 23 days and 125 days, yielding accuracy/specificity/sensitivity 0.82/1.0/0.55 in identifying progressing lesions (p < 0.01).

CONCLUSION

BM typically exhibit an early volume increase with diffusion restriction after LITT. Responders then show bi-exponential shrinkage with gradual diffusion increase. Progression can usually be detected only after 3-4 months, because earlier radiographic patterns may overlap with responding lesions.

KEY POINTS

Question Laser interstitial thermal therapy (LITT) is an emerging local treatment for brain metastases, but the radiographic patterns following this treatment have not been thoroughly described. Findings Responding lesions showed a typical radiographic pattern with early volumetric enlargement and diffusion restriction (not exclusive of responders), followed by a bi-exponential shrinkage and diffusion elevation. Clinical relevance Being aware of the typical radiographic changes in brain metastases responding to LITT is informative for the interpretation of follow-up images. Early volumetric and diffusion changes (< 3-4 months) do not appear to be reliable markers to predict treatment success.

Efficacy of reduced-dose gadobutrol versus standard-dose gadoterate in contrast-enhanced MRI for the evaluation of diabetic foot osteomyelitis

BackgroundGadobutrol is a macrocyclic gadolinium-based contrast agent (GBCA) with higher relaxivity than other GBCAs, suggesting the potential for dose reduction without compromising image quality.PurposeTo evaluate whether a 30% reduced dose of gadobutrol is as effective as the standard dose of gadoterate for lesion visualization and enhancement in diabetic foot osteomyelitis.MethodsThis study included 132 patients with preoperative contrast-enhanced foot MRIs prior to amputation surgery for diabetic foot osteomyelitis from November 2020 to January 2023. Sixty-six enhanced foot MRIs with reduced dose gadobutrol (0.07 mmol/kg) and 66 MRIs with standard dose gadoterate (0.1 mmol/kg) were reviewed by two radiologists. For the primary study objective, two parameters (lesion border visualization and subjective lesion enhancement) for qualitative lesion visualization were assessed between the two agents using a noninferiority analysis. In the quantitative assessment of lesion enhancement, lesion-to-background ratio and enhancement percentage were compared between the two agents.ResultsThe mean scores for lesion border delineation and the visual degree of contrast enhancement were nearly identical between the two groups. For both readers, the lower limit of the 95% confidence interval (CI) for the difference did not drop below -0.35, which is above the noninferiority margin. Regarding quantitative analysis, no significant differences were observed in the enhancement percentage and lesion-to-background ratio between the two agents (p > 0.5).ConclusionA 30% reduced dose of gadobutrol (0.07 mmol/kg) is as effective as the standard gadoterate dose (0.1 mmol/kg) for lesion visualization in contrast-enhanced MRI of diabetic foot osteomyelitis, with similar enhancement efficacy.

Efficacy of 3D-TSE sequence-based radiosurgery in prolonging time to distant intracranial failure: A session-wise analysis in a histology-diverse patient cohort

BACKGROUND

Stereotactic radiosurgery (SRS) for patients with brain metastases (BM) is associated with a risk of distant intracranial failure (DIF). This study evaluates the impact of integrating dedicated 3D fast/turbo spin echo (3D-TSE) sequences to MPRAGE in BM detection and DIF prolongation in a histology-agnostic patient cohort.

METHODS

The study population included adults treated with SRS from February 2019 to January 2024 who underwent MPRAGE alone or dual sequence with the addition of 3D-TSE starting from February 2020. Median times to DIF were estimated using the Kaplan-Meier method.

RESULTS

The 216 study patients who underwent 332 SRS courses for 1456 BM imaged with MPRAGE and 3D-TSE (primary cohort) were compared to a control cohort (92 patients, 135 SRS courses, 462 BM). In the session-wise analysis, the median time to DIF between the cohorts was significantly prolonged in the primary vs. control cohorts (11.4 vs. 6.8 months, P = .029), more pronounced in the subgroups with 1-4 metastases (14.7 vs. 8.1 months, P = .008) and with solitary BM (36.4 vs. 10.9 months, P = .001). While patients relapsing on immunotherapy or targeted therapy did not significantly benefit from 3D-TSE (7.2 vs. 5.7 months, P = .280), those who relapsed on chemotherapy or who were off systemic therapy (including synchronous metastases) exhibited a trend toward longer time to DIF with 3D-TSE integration (14.7 vs. 7.9 months, P = .057).

CONCLUSIONS

Implementing 3D-TSE sequences into SRS practice increases BM detection across all patients and translates into clinical relevance by prolonging time to DIF, particularly in those with limited intracranial disease and those not receiving central nervous system-active agents.

Multisite Benchmark Study for Standardized Relative CBV in Untreated Brain Metastases Using the DSC-MRI Consensus Acquisition Protocol

BACKGROUND AND PURPOSE

A national consensus recommendation for the collection of DSC-MRI perfusion data, used to create maps of relative CBV (rCBV), has been recently established for primary and metastatic brain tumors. The goal was to reduce intersite variability and improve ease of comparison across time and sites, fostering widespread use of this informative measure. To translate this goal into practice, the prospective collection of consensus DSC-MRI data and characterization of derived rCBV maps in brain metastases is needed. The purpose of this multisite study was to determine rCBV in untreated brain metastases in comparison to glioblastoma (GBM) and normal-appearing brain by using the national consensus protocol.

MATERIALS AND METHODS

Subjects from 3 sites with untreated enhancing brain metastases underwent DSC-MRI according to a recommended option that uses a midrange flip angle, GRE-EPI acquisition, and the administration of both a preload and second DSC-MRI dose of 0.1 mmol/kg gadolinium-based contrast agent. Quantitative maps of standardized relative CBV (srCBV) were generated and enhancing lesion ROIs determined from postcontrast T1-weighted images alone or calibrated difference maps, termed Δ T1 (dT1) maps. Mean srCBV for metastases were compared with normal-appearing white matter (NAWM) and GBM from a previous study. Comparisons were performed by using either the Wilcoxon signed-rank test for paired comparisons or the Mann-Whitney U nonparametric test for unpaired comparisons.

RESULTS

Forty-nine patients with a primary histology of lung (n = 25), breast (n = 6), squamous cell carcinoma (n = 1), melanoma (n = 5), gastrointestinal (GI) (n = 3), and genitourinary (GU) (n = 9) were included in comparison to GBM (n = 31). The mean srCBV of all metastases (1.83±1.05) were significantly lower (P = .0009) than mean srCBV for GBM (2.67 ± 1.34) with both statistically greater (P < .0001) than NAWM (0.68 ± 0.18). Histologically distinct metastases are each statistically greater than NAWM (P < .0001) with lung (P = .0002) and GU (P = .02) srCBV being significantly different from GBM srCBV.

CONCLUSIONS

Using the consensus DSC-MRI protocol, mean srCBV values were determined for treatment-naïve brain metastases in comparison to normal-appearing white matter and GBM thus setting the benchmark for all subsequent studies adherent to the national consensus recommendation.

Deep Learning-Based Signal Amplification of T1-Weighted Single-Dose Images Improves Metastasis Detection in Brain MRI

OBJECTIVES

Double-dose contrast-enhanced brain imaging improves tumor delineation and detection of occult metastases but is limited by concerns about gadolinium-based contrast agents' effects on patients and the environment. The purpose of this study was to test the benefit of a deep learning-based contrast signal amplification in true single-dose T1-weighted (T-SD) images creating artificial double-dose (A-DD) images for metastasis detection in brain magnetic resonance imaging.

MATERIALS AND METHODS

In this prospective, multicenter study, a deep learning-based method originally trained on noncontrast, low-dose, and T-SD brain images was applied to T-SD images of 30 participants (mean age ± SD, 58.5 ± 11.8 years; 23 women) acquired externally between November 2022 and June 2023. Four readers with different levels of experience independently reviewed T-SD and A-DD images for metastases with 4 weeks between readings. A reference reader reviewed additionally acquired true double-dose images to determine any metastases present. Performances were compared using Mid-p McNemar tests for sensitivity and Wilcoxon signed rank tests for false-positive findings.

RESULTS

All readers found more metastases using A-DD images. The 2 experienced neuroradiologists achieved the same level of sensitivity using T-SD images (62 of 91 metastases, 68.1%). While the increase in sensitivity using A-DD images was only descriptive for 1 of them (A-DD: 65 of 91 metastases, +3.3%, P = 0.424), the second neuroradiologist benefited significantly with a sensitivity increase of 12.1% (73 of 91 metastases, P = 0.008). The 2 less experienced readers (1 resident and 1 fellow) both found significantly more metastases on A-DD images (resident, T-SD: 61.5%, A-DD: 68.1%, P = 0.039; fellow, T-SD: 58.2%, A-DD: 70.3%, P = 0.008). They were therefore able to use A-DD images to increase their sensitivity to the neuroradiologists' initial level on regular T-SD images. False-positive findings did not differ significantly between sequences. However, readers showed descriptively more false-positive findings on A-DD images. The benefit in sensitivity particularly applied to metastases ≤5 mm (5.7%-17.3% increase in sensitivity).

CONCLUSIONS

A-DD images can improve the detectability of brain metastases without a significant loss of precision and could therefore represent a potentially valuable addition to regular single-dose brain imaging.

Leptomeningeal metastatic disease: new frontiers and future directions

Leptomeningeal metastatic disease (LMD), encompassing entities of 'meningeal carcinomatosis', neoplastic meningitis' and 'leukaemic/lymphomatous meningitis', arises secondary to the metastatic dissemination of cancer cells from extracranial and certain intracranial malignancies into the leptomeninges and cerebrospinal fluid. The clinical burden of LMD has been increasing secondary to more sensitive diagnostics, aggressive local therapies for discrete brain metastases, and improved management of extracranial disease with targeted and immunotherapeutic agents, resulting in improved survival. However, owing to drug delivery challenges and the unique microenvironment of LMD, novel therapies against systemic disease have not yet translated into improved outcomes for these patients. Underdiagnosis and misdiagnosis are common, response assessment remains challenging, and the prognosis associated with this disease of whole neuroaxis remains extremely poor. The dearth of effective therapies is further challenged by the difficulties in studying this dynamic disease state. In this Review, a multidisciplinary group of experts describe the emerging evidence and areas of active investigation in LMD and provide directed recommendations for future research. Drawing upon paradigm-changing advances in mechanistic science, computational approaches, and trial design, the authors discuss domain-specific and cross-disciplinary strategies for optimizing the clinical and translational research landscape for LMD. Advances in diagnostics, multi-agent intrathecal therapies, cell-based therapies, immunotherapies, proton craniospinal irradiation and ongoing clinical trials offer hope for improving outcomes for patients with LMD.

Improved detection of brain metastases using contrast-enhanced 3D black-blood TSE sequences compared to post-contrast 3D T1 GRE: a comparative study on 1.5-T MRI

OBJECTIVES

Brain metastases are the most common intracranial malignancy in adults, and their detection is crucial for treatment planning. Post-contrast 3D T1 gradient-recalled echo (GRE) sequences are commonly used for this purpose, but contrast-enhanced 3D T1 turbo spin-echo (TSE) sequences with motion-sensitized driven-equilibrium (MSDE) technique ("black blood") may offer improved detection. This study aimed to compare the effectiveness of contrast-enhanced 3D black blood sequences to standard 3D T1 GRE sequences in detecting brain metastases on a 1.5-T MRI.

MATERIALS AND METHODS

A retrospective analysis of 183 patients with suspected or follow-up brain metastases between May 2022 and September 2023 was conducted. Among these patients, 107 were included in the final analysis. Both post-contrast 3D T1 GRE and 3D black blood sequences were acquired on the same scanner with similar acquisition times. Two neuroradiologists independently evaluated the images for the number, size, and location of metastases. Interobserver variability and statistical analysis were performed.

RESULTS

Among the 107 patients (mean age 60.8 years ± 13.2 years; 55 males, 52 females), 3D black blood sequences detected a significantly higher number of brain metastases, particularly small lesions (< 5 mm), compared to 3D T1 GRE sequences (p < 0.05). There was no significant difference in detecting large metastases (≥ 5 mm) between the sequences. In addition, the black blood sequences provided better conspicuity of metastases in the majority of patients (85%).

CONCLUSION

Contrast-enhanced 3D T1 TSE with MSDE ("black blood") sequences offer improved detection of brain metastases, especially small lesions, on 1.5-T MRI compared to standard 3D T1 GRE sequences.

KEY POINTS

Question Accurate identification of the number and location of brain metastases using MRI is essential for planning and managing effective treatment. Findings Contrast-enhanced 3D T1 TSE black blood sequences detected significantly more small brain metastases than standard 3D T1 GRE sequences on 1.5-T MRI. Clinical relevance The use of 3D black blood sequences on 1.5-T MRI may have the potential to improve the accuracy of detection of brain metastases, leading to better treatment planning and potentially improved patient outcomes.

Modernizing Neuro-Oncology: The Impact of Imaging, Liquid Biopsies, and AI on Diagnosis and Treatment

Advances in neuro-oncology have transformed the diagnosis and management of brain tumors, which are among the most challenging malignancies due to their high mortality rates and complex neurological effects. Despite advancements in surgery and chemoradiotherapy, the prognosis for glioblastoma multiforme (GBM) and brain metastases remains poor, underscoring the need for innovative diagnostic strategies. This review highlights recent advancements in imaging techniques, liquid biopsies, and artificial intelligence (AI) applications addressing current diagnostic challenges. Advanced imaging techniques, including diffusion tensor imaging (DTI) and magnetic resonance spectroscopy (MRS), improve the differentiation of tumor progression from treatment-related changes. Additionally, novel positron emission tomography (PET) radiotracers, such as 18F-fluoropivalate, 18F-fluoroethyltyrosine, and 18F-fluluciclovine, facilitate metabolic profiling of high-grade gliomas. Liquid biopsy, a minimally invasive technique, enables real-time monitoring of biomarkers such as circulating tumor DNA (ctDNA), extracellular vesicles (EVs), circulating tumor cells (CTCs), and tumor-educated platelets (TEPs), enhancing diagnostic precision. AI-driven algorithms, such as convolutional neural networks, integrate diagnostic tools to improve accuracy, reduce interobserver variability, and accelerate clinical decision-making. These innovations advance personalized neuro-oncological care, offering new opportunities to improve outcomes for patients with central nervous system tumors. We advocate for future research integrating these tools into clinical workflows, addressing accessibility challenges, and standardizing methodologies to ensure broad applicability in neuro-oncology.

Gadolinium-Enhanced T2 FLAIR Is an Imaging Biomarker of Radiation Necrosis and Tumor Progression in Patients with Brain Metastases

BACKGROUND AND PURPOSE

Differentiating radiation necrosis (RN) from tumor progression (TP) after radiation therapy for brain metastases is an important clinical problem requiring advanced imaging techniques that may not be widely available and are challenging to perform at multiple time points. The ability to leverage conventional MRI for this problem could have a meaningful clinical impact. The purpose of this study was to explore contrast-enhanced T2 FLAIR (T2FLAIRc) as a new imaging biomarker of RN and TP.

MATERIALS AND METHODS

This single-institution retrospective study included patients with treated brain metastases undergoing DSC-MRI between January 2021 and June 2023. Reference standard assessment was based on histopathology or serial follow-up, including the results of DSC-MRI for a minimum of 6 months from the first DSC-MRI. The index test was implemented as part of the institutional brain tumor MRI protocol and preceded the first DSC-MRI. T2FLAIRc and gadolinium-enhanced T1 (T1c) MPRAGE signal were normalized against normal brain parenchyma and expressed as a z score. The mean signal intensity of enhancing disease for the RN and TP groups was compared using an unpaired t test. Receiver operating characteristic curves and area under the receiver operating characteristic curve (AUC) were derived by bootstrapping. The DeLong test was used to compare AUCs.

RESULTS

Fifty-six participants (mean age, 62 [SD, 12.7] years; 39 women; 28 with RN, 28 with TP) were evaluated. The index MRI was performed, on average, 73 [SD, 34] days before the first DSC-MRI. Significantly higher z scores were found for RN using T2FLAIRc (8.3 versus 5.8, P < .001) and T1c (4.1 versus 3.5, P = .02). The AUC for T2FLAIRc (0.83; 95% CI, 0.72-0.92) was greater than that for T1c (0.70; 95% CI, 0.56-0.83) (P = .04). The AUC of DSC-derived relative CBV (0.82; 95% CI, 0.70-0.93) was not significantly different from that of T2FLAIRc (P = .9).

CONCLUSIONS

A higher normalized T1c and T2FLAIRc signal intensity was found for RN. In a univariable test, the mean T2FLAIRc signal intensity of enhancing voxels showed good discrimination performance for distinguishing RN from TP. The results of this work demonstrate the potential of T2FLAIRc as an imaging biomarker in the work-up of RN in patients with brain metastases.

Size matters: Early progression of melanoma brain metastases after treatment with immune checkpoint inhibitors

Background

Immune checkpoint inhibitors (ICIs) are effective treatments for patients with metastatic melanoma, including patients with brain metastasis (BM). However, half of patients with melanoma BM have intracranial progression within 6 months after the start of ICIs. We investigated whether size affects response to ICIs in patients with melanoma BM.

Methods

In this single-center cohort study, patients with melanoma BM who were treated with ICIs between 2012 and 2021 were included. Clinical and radiologic features were collected at baseline. Longest axial diameter of all BMs was measured on baseline and follow-up MRI, and segmentation was performed for all BMs on baseline MRI. Lesion-level logistic regression analysis and patient-level survival analysis were performed for early BM progression (ie, within 6 months after start of ICIs) and intracranial progression-free survival (PFS), respectively.

Results

A total of 82 patients were included with a total of 464 BMs. At baseline, 37.8% of patients had ≥ 4 BMs and 53.7% of patients had at least one BM with a diameter ≥ 10 mm. In multivariable analysis on the lesion level, baseline BM diameter was associated with early BM progression (odds ratio 1.10, 95%CI 1.05-1.15, P < .001). On the patient level, having at least one BM ≥ 10mm was associated with shorter intracranial PFS (hazard ratio 2.08, 95%CI 1.64-5.56, P < .001).

Conclusions

Large BM diameter was associated with a higher risk of early progression after the start of ICIs. Therefore, local therapy should be considered for patients who are treated with ICIs and who have melanoma BMs ≥ 10 mm.

Consortium for Intracranial Metastasis Academic Research (CIMARa): Global interdisciplinary collaborations to improve outcomes of patient with brain metastases

Brain metastases (BrM) arising from solid tumors is an ever-increasing and often devastating clinical challenge impacting hundreds of thousands of patients annually worldwide. As systemic anticancer therapies, and thus survival, improve, the risk for central nervous system (CNS) recurrence has increased. Historically, patients with BrM were excluded from clinical trials; however, there has been a shift toward increasing inclusion over the past decade. To most effectively design the next generation of clinical trials for patients with BrM, a multidisciplinary team spanning local and systemic therapies is imperative. CIMARa (Consortium for Intracranial Metastasis Academic Research), formalized in June 2021, is an inclusive group of multidisciplinary clinical investigators, research scientists, and advocates who share the collective goal of improving outcomes for patients with BrM. CIMARa aims to improve outcomes through the development, coordination, and awareness of multi-institutional clinical trials testing novel therapeutic agents for this unique patient population alongside the translation of preclinical research to the clinical setting.

The Brain Tumor Segmentation - Metastases (BraTS-METS) Challenge 2023: Brain Metastasis Segmentation on Pre-treatment MRI

The translation of AI-generated brain metastases (BM) segmentation into clinical practice relies heavily on diverse, high-quality annotated medical imaging datasets. The BraTS-METS 2023 challenge has gained momentum for testing and benchmarking algorithms using rigorously annotated internationally compiled real-world datasets. This study presents the results of the segmentation challenge and characterizes the challenging cases that impacted the performance of the winning algorithms. Untreated brain metastases on standard anatomic MRI sequences (T1, T2, FLAIR, T1PG) from eight contributed international datasets were annotated in stepwise method: published UNET algorithms, student, neuroradiologist, final approver neuroradiologist. Segmentations were ranked based on lesion-wise Dice and Hausdorff distance (HD95) scores. False positives (FP) and false negatives (FN) were rigorously penalized, receiving a score of 0 for Dice and a fixed penalty of 374 for HD95. The mean scores for the teams were calculated. Eight datasets comprising 1303 studies were annotated, with 402 studies (3076 lesions) released on Synapse as publicly available datasets to challenge competitors. Additionally, 31 studies (139 lesions) were held out for validation, and 59 studies (218 lesions) were used for testing. Segmentation accuracy was measured as rank across subjects, with the winning team achieving a LesionWise mean score of 7.9. The Dice score for the winning team was 0.65 ± 0.25. Common errors among the leading teams included false negatives for small lesions and misregistration of masks in space. The Dice scores and lesion detection rates of all algorithms diminished with decreasing tumor size, particularly for tumors smaller than 100 mm3. In conclusion, algorithms for BM segmentation require further refinement to balance high sensitivity in lesion detection with the minimization of false positives and negatives. The BraTS-METS 2023 challenge successfully curated well-annotated, diverse datasets and identified common errors, facilitating the translation of BM segmentation across varied clinical environments and providing personalized volumetric reports to patients undergoing BM treatment.

Enhancing prognostication and treatment response evaluation in primary CNS lymphoma with 18F-FDG-PET/CT

BACKGROUND

The role of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG-PET/CT) in the prognostication and response evaluation of primary central nervous system lymphoma (PCNSL) remains inadequately defined.

METHODS

We conducted a retrospective analysis of 268 consecutive newly diagnosed patients with PCNSL between 2006 and 2020. Of these patients, 105 and 110 patients were included to evaluate the prognostic value of baseline and post-treatment 18F-FDG-PET/CT scans, respectively. Tumor uptake was considered positive when it exceeded that of the contralateral brain upon visual assessment. Quantitative analysis of baseline 18F-FDG-PET/CT included measurement of the maximal standardized uptake value (SUVmax), total metabolic tumor volume (TMTV), and total lesion glycolysis (TLG).

RESULTS

The median age of the 268 patients was 62 years (range: 17-85), with 55% being male. The median progression-free survival (PFS) was 24.5 months (95% CI: 19.9-29.1), and the median overall survival (OS) was 34.5 months (95% CI: 22.9-46.1). The average SUVmax was 15.3 ± 5.7 and the mean TMTV and TLG were 12.6 ± 13.9 cm3 and 135.0 ± 152.7 g, respectively. Patients with a baseline TMTV ≥ 17.0 cm3 had significantly shorter OS (12.5 vs 74.0 months, P = .011). Post-treatment metabolic response by 18F-FDG-PET/CT significantly predicted PFS (median: 10.5 vs 46.0 months, P = .001) and OS (median: 21.0 vs 62.0 months, P = .002), whereas anatomic response by contrast-enhanced MRI showed no statistically significant differences in PFS (P = .130) or OS (P = .540).

CONCLUSION

Baseline TMTV and post-treatment metabolic response, as assessed by 18F-FDG-PET/CT, are significant prognostic factors in patients with PCNSL.

The sensitivity of MIPs of 3D contrast-enhanced VIBE T1-weighted imaging for the detection of small brain metastases (≤ 5 mm) on 1.5 tesla MRI

OBJECTIVES

To evaluate whether the use of Maximum Intensity Projection (MIP) images derived from contrast-enhanced 3D-T1-weighted volumetric interpolated breath-hold examination (VIBE) would allow more sensitive detection of small (≤5 mm) brain metastases (BM) compared with source as well as 2D-T1-weighted spin-echo (SE) images.

METHODS

We performed a single center retrospective study on subjects with BM who underwent 1.5 tesla brain magnetic resonance imaging. Two readers counted the number of small BM for each of the seven sets of contrast-enhanced images created: axial 2D-T1-weighted SE, 3D-T1-weighted VIBE, 2.5 mm-thick-MIP T1-weighted VIBE, and 5 mm-thick-MIP T1-weighted VIBE; sagittal 3D-T1-weighted VIBE, 2.5 mm-thick-MIP T1-weighted VIBE, and 5 mm-thick-MIP T1-weighted VIBE. Total number of lesions detected on each image type was compared. Sensitivity, the average rates of false negatives and false positives, and the mean discrepancy were evaluated.

RESULTS

A total of 403 small BM were identified in 49 patients. Significant differences were found: in the number of true positives and false negatives between the axial 2D-T1-weighted SE sequence and all other imaging techniques; in the number of false positives between the axial 2D-T1-weighted SE and the axial 3D-T1-weighted VIBE sequences. The two image types that combined offered the highest sensitivity were 2D-T1-weighted SE and axial 2.5 mm-thick-MIP T1-weighted VIBE. The axial 2D-T1-weighted SE sequence differed significantly in sensitivity from all other sequences.

CONCLUSION

MIP images did not show a significant difference in sensitivity for the detection of small BM compared with native images.

The Role of Delayed Imaging at MRI in Rare Non-enhancing Prostate Cancer Brain Metastases: A Case Report

Brain metastases in prostate cancer are rare (<2% of cases). In magnetic resonance imaging, nearly all brain metastases exhibit contrast-enhancement, which may be affected by the time elapsed since the administration of the contrast agent. We discuss a case where the brain metastases in a patient with prostate cancer do not show a clear contrast-enhancement on magnetic resonance imaging using a standard brain metastases protocol. It also emphasizes the usefulness of delayed imaging in identifying blood-brain barrier damage. We present the case of a 69-year-old man diagnosed with prostate adenocarcinoma, currently in castration-resistant phase (last value of serum prostate-specific antigen: 45.1 ng/mL) with bone, mediastinal and inguinal lymph nodes, pulmonary, and hepatic metastases. In a contrast-enhanced whole-body computed tomography examination, the appearance of intra-axial brain lesions suspicious for metastases was documented. The subsequent contrast-enhanced brain magnetic resonance imaging showed the presence of 5 intra-axial lesions consistent with brain metastases. These lesions exhibited hyperintense signals in T2-fluid-attenuated inversion recovery images; after contrast agent administration, a ring-like contrast-enhancement was more clearly visible in T1-weighted images acquired later (about 15 minutes after contrast agent administration) than in those acquired earlier (about 5-7 minutes after contrast agent administration). In conclusion, for oncological subjects with multiple brain lesions lacking obvious contrast-enhancement using a standard magnetic resonance imaging protocol, we suggest acquiring late images. These might allow for the detection of even minimal post-contrast impregnation, improving confidence in the diagnosis of brain metastases.

Magnetic Resonance Imaging Assessment of Morphological Changes and Molecular Behavior to Evaluate Treatment Response of Brain Metastatic Lesions After Stereotactic Radiosurgery

BACKGROUND

Brain metastases (BMs) are the most common type of intracranial tumors, frequently arising from primary cancers such as lung, breast, melanoma, and renal cell carcinoma. Magnetic resonance imaging (MRI) plays a crucial role in assessing both the morphological and molecular characteristics of BMs, particularly in evaluating treatment response following radiosurgery. However, the interpretation of these imaging changes remains complex, often influencing clinical decision-making.

OBJECTIVE

This study aims to evaluate the morphological changes and molecular behavior of BMs postradiosurgery using MRI to assess treatment response.

MATERIALS AND METHODS

A retrospective review was conducted at a high specialty medical center, including 41 patients with BMs treated with stereotactic radiosurgery (SRS) from 2018 to 2022. Patients had a baseline MRI (pre-SRS) prior to treatment and follow-ups at 2-3 months (MRI-2) and 5-6 months (MRI-3). The response assessment in neuro-oncology brain metastases (RANO-BM) criteria were used, and T1/T2 matching was analyzed for each follow-up. Logistic regression was performed relating the T1/T2 matching and susceptibility areas (susceptibility-weighted imaging (SWI)) for MRI-2 and MRI-3. Cross tables were created regarding treatment response and demographic characteristics according to Pearson's Chi-squared test.

RESULTS

The mean age was 56.7 years; 53.7% (n = 22) were female. Primary tumors included lung (29.3%, n = 12), breast (19.5%, n = 8), colon (12.2%, n = 5), and melanoma and kidney tumors (7.3%, n = 3). Post-SRS changes included transitions from solid to cystic lesions, reduced perilesional edema, size reduction, and increased areas of magnetic susceptibility. A mixed pattern (areas of T1/T2 match + mismatch) was noted at lesion margins during follow-ups (MRI-2: 70.7% (n = 29), MRI-3: 68.3% (n = 28)). Most patients exhibited a partial response at MRI-2 (43.9%, n = 18), while at MRI-3, disease progression occurred (43.9%, n = 18) due to an increase in lesion number. Logistic regression linking T1/T2 matching and SWI demonstrated a significantly central-peripheral SWI distribution for T1/T2 match during both follow-ups (MRI-2: p = 0.005, R2: 0.52; MRI-3: p = 0.002, R2: 0.56). SWI distribution was higher when a mixed T1/T2 matching was present. Significant associations were found with systemic treatment and response type at MRI-2 (p =0.001), predominantly showing a partial response for those receiving chemotherapy + targeted therapy.

CONCLUSIONS

SWI and T1/T2 mismatch are valuable tools reflecting changes in the tumor microenvironment postradiosurgery, aiding in treatment response monitoring. The appearance of susceptibility areas may precede changes in the enhancement of the lesion margin. Short-term follow-ups (2-3 months) are crucial due to prevalent progression, marked primarily by the appearance of new lesions in approximately 50% of patients.

Brain Metastases in Differentiated Thyroid Cancer: Clinical Presentation, Diagnosis, and Management

Background: Brain metastases (BM) are the most common intracranial neoplasms in adults and are a significant cause of morbidity and mortality. The brain is an unusual site for distant metastases of thyroid cancer; indeed, the most common sites are lungs and bones. In this narrative review, we discuss about the clinical characteristics, diagnosis, and treatment options for patients with BM from differentiated thyroid cancer (DTC). Summary: BM can be discovered before initial therapy due to symptoms, but in most patients, BM is diagnosed during follow-up because of imaging performed before starting tyrosine kinase inhibitors (TKI) or due to the onset of neurological symptoms. Older male patients with follicular thyroid cancer (FTC), poorly differentiated thyroid cancer (PDTC), and distant metastases may have an increased risk of developing BM. The gold standard for detection of BM is magnetic resonance imaging with contrast agent administration, which is superior to contrast-enhanced computed tomography. The treatment strategies for patients with BM from DTC remain controversial. Patients with poor performance status are candidates for palliative and supportive care. Neurosurgery is usually reserved for cases where symptoms persist despite medical treatment, especially in patients with favorable prognostic factors and larger lesions. It should also be considered for patients with a single BM in a surgically accessible location, particularly if the primary disease is controlled without other systemic metastases. Additionally, stereotactic radiosurgery (SRS) may be the preferred option for treating small lesions, especially those in inaccessible areas of the brain or when surgery is not advisable. Whole brain radiotherapy is less frequently used in treating these patients due to its potential side effects and the debated effectiveness. Therefore, it is typically reserved for cases involving multiple BM that are too large for SRS. TKIs are effective in patients with progressive radioiodine-refractory thyroid cancer and multiple metastases. Conclusions: Although routine screening for BM is not recommended, older male patients with FTC or PDTC and distant metastases may be at higher risk and should be carefully evaluated for BM. According to current data, patients who are suitable for neurosurgery seem to have the highest survival benefit, while SRS may be appropriate for selected patient.

Accelerating brain three-dimensional T2 fluid-attenuated inversion recovery using artificial intelligence-assisted compressed sensing: a comparison study with parallel imaging

Background

Shortening the acquisition time of brain three-dimensional T2 fluid-attenuated inversion recovery (3D T2 FLAIR) by using acceleration techniques has the potential to reduce motion artifacts in images and facilitate clinical application. This study aimed to assess the image quality of brain 3D T2 FLAIR accelerated by artificial intelligence-assisted compressed sensing (ACS) in comparison to 3D T2 FLAIR accelerated by parallel imaging (PI).

Methods

In this prospective cohort study, 102 consecutive participants, including both healthy individuals and those with suspected brain diseases, were recruited and underwent both ACS- and PI-3D T2 FLAIR scans with a 3.0-Tesla magnetic resonance imaging system from February 2023 to October 2023 in Beijing Tiantan Hospital, Capital Medical University. Quantitative assessment involved white matter (WM) and gray matter (GM) signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), whole-image sharpness, and tumor volume. Qualitative assessment included the scoring of overall image quality, GM-WM border sharpness, and diagnostic confidence in lesion detection.

Results

ACS-3D T2 FLAIR exhibited a shorter acquisition time compared to PI-3D T2 FLAIR (105 vs. 320 seconds). ACS-3D T2 FLAIR, compared to PI-3D T2 FLAIR, demonstrated a significantly higher mean SNRWM (25.922±6.811 vs. 22.544±5.853; P<0.001), SNRGM (18.324±7.137 vs. 17.102±6.659; P=0.049), CNRWM/GM (4.613±1.547 vs. 4.160±1.552; P<0.001), and sharpness (0.413±0.049 vs. 0.396±0.034; P<0.001), while no significant differences were found for the overall image quality ratings (P=0.063) or GM-WM border sharpness ratings (P=0.125). A good agreement on tumor volume was achieved between ACS-3D T2 FLAIR and PI-3D T2 FLAIR images (intraclass correlation coefficient =0.999; 0.998-1.000; P<0.001). Images acquired with ACS demonstrated nearly equivalent diagnostic confidence to those obtained with PI (P>0.05).

Conclusions

The ACS technique offers a substantial reduction in scanning time for brain 3D T2 FLAIR compared to PI while maintaining good image quality and equivalent diagnostic confidence.

Extracellular gadolinium-based contrast agents

Extracellular gadolinium-based contrast agents (GBCA) are commonly used in magnetic resonance imaging (MRI) because they increase the detection of alterations, improve tissue characterisation and enable a more precise differential diagnosis. GBCAs are considered to be safe but they are not risk-free. When using GBCAs, it is important to be aware of the risks and to know how to react in different situations (pregnancy, breastfeeding, kidney failure) including if complications occur (extravasations, adverse, allergic or anaphylactic reactions). The article describes the characteristics of the gadolinium molecule, the differences in the biochemical structure of these GBCA, their biodistribution and the effect on the MRI signal. It also reviews safety aspects and the most common clinical applications.

Multicenter privacy-preserving model training for deep learning brain metastases autosegmentation

OBJECTIVES

This work aims to explore the impact of multicenter data heterogeneity on deep learning brain metastases (BM) autosegmentation performance, and assess the efficacy of an incremental transfer learning technique, namely learning without forgetting (LWF), to improve model generalizability without sharing raw data.

MATERIALS AND METHODS

A total of six BM datasets from University Hospital Erlangen (UKER), University Hospital Zurich (USZ), Stanford, UCSF, New York University (NYU), and BraTS Challenge 2023 were used. First, the performance of the DeepMedic network for BM autosegmentation was established for exclusive single-center training and mixed multicenter training, respectively. Subsequently privacy-preserving bilateral collaboration was evaluated, where a pretrained model is shared to another center for further training using transfer learning (TL) either with or without LWF.

RESULTS

For single-center training, average F1 scores of BM detection range from 0.625 (NYU) to 0.876 (UKER) on respective single-center test data. Mixed multicenter training notably improves F1 scores at Stanford and NYU, with negligible improvement at other centers. When the UKER pretrained model is applied to USZ, LWF achieves a higher average F1 score (0.839) than naive TL (0.570) and single-center training (0.688) on combined UKER and USZ test data. Naive TL improves sensitivity and contouring accuracy, but compromises precision. Conversely, LWF demonstrates commendable sensitivity, precision and contouring accuracy. When applied to Stanford, similar performance was observed.

CONCLUSION

Data heterogeneity (e.g., variations in metastases density, spatial distribution, and image spatial resolution across centers) results in varying performance in BM autosegmentation, posing challenges to model generalizability. LWF is a promising approach to peer-to-peer privacy-preserving model training.

Imaging timing after surgery for glioblastoma: an evaluation of practice in Great Britain and Ireland (INTERVAL-GB)- a multi-centre, cohort study

PURPOSE

Post-operative MRI is used to assess extent of resection, monitor treatment response and detect progression in high-grade glioma. However, compliance with accepted guidelines for follow-up MRI, and impact on management/outcomes is unclear.

METHODS

Multi-center, retrospective observational cohort study of patients with confirmed WHO grade 4 glioma (August 2018-February 2019) receiving oncological treatment.

PRIMARY OBJECTIVE

investigate follow-up MRI surveillance practice and compliance with recommendations from NICE (Post-operative scan < 72h, MRI every 3-6 months) and EANO (Post-operative scan < 48h, MRI every 3 months).

RESULTS

There were 754 patients from 26 neuro-oncology centers with a median age of 63 years (IQR 54-70), yielding 10,100 (median, 12.5/person, IQR 5.2-19.4) person-months of follow-up. Of patients receiving debulking surgery, most patients had post-operative MRI within 72 h of surgery (78.0%, N = 407/522), and within 48 h of surgery (64.2%, N = 335/522). The median number of subsequent follow-up MRI scans was 1 (IQR 0-4). Compliance with NICE and EANO recommendations for follow-up MRI was 52.8% (N = 398/754) and 24.9% (N = 188/754), respectively. On multivariable Cox regression analysis, increased time spent in recommended follow-up according to NICE guidelines was associated with longer OS (HR 0.56, 95% CI 0.46-0.66, P < 0.001), but not PFS (HR 0.93, 95% CI 0.79-1.10, P = 0.349). Increased time spent in recommended follow-up according to EANO guidelines was associated with longer OS (HR 0.54, 95% CI 0.45-0.63, P < 0.001) but not PFS (HR 0.99, 95% CI 0.84-1.16, P = 0.874).

CONCLUSION

Regular surveillance follow-up for glioblastoma is associated with longer OS. Prospective trials are needed to determine whether regular or symptom-directed MRI influences outcomes.

Differentiation of glioblastoma from solitary brain metastasis using deep ensembles: Empirical estimation of uncertainty for clinical reliability

BACKGROUND AND OBJECTIVES

To develop a clinically reliable deep learning model to differentiate glioblastoma (GBM) from solitary brain metastasis (SBM) by providing predictive uncertainty estimates and interpretability.

METHODS

A total of 469 patients (300 GBM, 169 SBM) were enrolled in the institutional training set. Deep ensembles based on DenseNet121 were trained on multiparametric MRI. The model performance was validated in the external test set consisting of 143 patients (101 GBM, 42 SBM). Entropy values for each input were evaluated for uncertainty measurement; based on entropy values, the datasets were split to high- and low-uncertainty groups. In addition, entropy values of out-of-distribution (OOD) data from unknown class (257 patients with meningioma) were compared to assess uncertainty estimates of the model. The model interpretability was further evaluated by localization accuracy of the model.

RESULTS

On external test set, the area under the curve (AUC), accuracy, sensitivity and specificity of the deep ensembles were 0.83 (95 % confidence interval [CI] 0.76-0.90), 76.2 %, 54.8 % and 85.2 %, respectively. The performance was higher in the low-uncertainty group than in the high-uncertainty group, with AUCs of 0.91 (95 % CI 0.83-0.98) and 0.58 (95 % CI 0.44-0.71), indicating that assessment of uncertainty with entropy values ascertained reliable prediction in the low-uncertainty group. Further, deep ensembles classified a high proportion (90.7 %) of predictions on OOD data to be uncertain, showing robustness in dataset shift. Interpretability evaluated by localization accuracy provided further reliability in the "low-uncertainty and high-localization accuracy" subgroup, with an AUC of 0.98 (95 % CI 0.95-1.00).

CONCLUSIONS

Empirical assessment of uncertainty and interpretability in deep ensembles provides evidence for the robustness of prediction, offering a clinically reliable model in differentiating GBM from SBM.

Differentiating IDH-mutant astrocytomas and 1p19q-codeleted oligodendrogliomas using DSC-PWI: high performance through cerebral blood volume and percentage of signal recovery percentiles

OBJECTIVE

Presurgical differentiation between astrocytomas and oligodendrogliomas remains an unresolved challenge in neuro-oncology. This research aims to provide a comprehensive understanding of each tumor's DSC-PWI signatures, evaluate the discriminative capacity of cerebral blood volume (CBV) and percentage of signal recovery (PSR) percentile values, and explore the synergy of CBV and PSR combination for pre-surgical differentiation.

METHODS

Patients diagnosed with grade 2 and 3 IDH-mutant astrocytomas and IDH-mutant 1p19q-codeleted oligodendrogliomas were retrospectively retrieved (2010-2022). 3D segmentations of each tumor were conducted, and voxel-level CBV and PSR were extracted to compute mean, minimum, maximum, and percentile values. Statistical comparisons were performed using the Mann-Whitney U test and the area under the receiver operating characteristic curve (AUC-ROC). Lastly, the five most discriminative variables were combined for classification with internal cross-validation.

RESULTS

The study enrolled 52 patients (mean age 45-year-old, 28 men): 28 astrocytomas and 24 oligodendrogliomas. Oligodendrogliomas exhibited higher CBV and lower PSR than astrocytomas across all metrics (e.g., mean CBV = 2.05 and 1.55, PSR = 0.68 and 0.81 respectively). The highest AUC-ROCs and the smallest p values originated from CBV and PSR percentiles (e.g., PSRp70 AUC-ROC = 0.84 and p value = 0.0005, CBVp75 AUC-ROC = 0.8 and p value = 0.0006). The mean, minimum, and maximum values yielded lower results. Combining the best five variables (PSRp65, CBVp70, PSRp60, CBVp75, and PSRp40) achieved a mean AUC-ROC of 0.87 for differentiation.

CONCLUSIONS

Oligodendrogliomas exhibit higher CBV and lower PSR than astrocytomas, traits that are emphasized when considering percentiles rather than mean or extreme values. The combination of CBV and PSR percentiles results in promising classification outcomes.

CLINICAL RELEVANCE STATEMENT

The combination of histogram-derived percentile values of cerebral blood volume and percentage of signal recovery from DSC-PWI enhances the presurgical differentiation between astrocytomas and oligodendrogliomas, suggesting that incorporating these metrics into clinical practice could be beneficial.

KEY POINTS

• The unsupervised selection of percentile values for cerebral blood volume and percentage of signal recovery enhances presurgical differentiation of astrocytomas and oligodendrogliomas. • Oligodendrogliomas exhibit higher cerebral blood volume and lower percentage of signal recovery than astrocytomas. • Cerebral blood volume and percentage of signal recovery combined provide a broader perspective on tumor vasculature and yield promising results for this preoperative classification.

DSC-PWI presurgical differentiation of grade 4 astrocytoma and glioblastoma in young adults: rCBV percentile analysis across enhancing and non-enhancing regions

PURPOSE

The presurgical discrimination of IDH-mutant astrocytoma grade 4 from IDH-wildtype glioblastoma is crucial for patient management, especially in younger adults, aiding in prognostic assessment, guiding molecular diagnostics and surgical planning, and identifying candidates for IDH-targeted trials. Despite its potential, the full capabilities of DSC-PWI remain underexplored. This research evaluates the differentiation ability of relative-cerebral-blood-volume (rCBV) percentile values for the enhancing and non-enhancing tumor regions compared to the more commonly used mean or maximum preselected rCBV values.

METHODS

This retrospective study, spanning 2016-2023, included patients under 55 years (age threshold based on World Health Organization recommendations) with grade 4 astrocytic tumors and known IDH status, who underwent presurgical MR with DSC-PWI. Enhancing and non-enhancing regions were 3D-segmented to calculate voxel-level rCBV, deriving mean, maximum, and percentile values. Statistical analyses were conducted using the Mann-Whitney U test and AUC-ROC.

RESULTS

The cohort consisted of 59 patients (mean age 46; 34 male): 11 astrocytoma-4 and 48 glioblastoma. While glioblastoma showed higher rCBV in enhancing regions, the differences were not significant. However, non-enhancing astrocytoma-4 regions displayed notably higher rCBV, particularly in lower percentiles. The 30th rCBV percentile for non-enhancing regions was 0.705 in astrocytoma-4, compared to 0.458 in glioblastoma (p = 0.001, AUC-ROC = 0.811), outperforming standard mean and maximum values.

CONCLUSION

Employing an automated percentile-based approach for rCBV selection enhances differentiation capabilities, with non-enhancing regions providing more insightful data. Elevated rCBV in lower percentiles of non-enhancing astrocytoma-4 is the most distinguishable characteristic and may indicate lowly vascularized infiltrated edema, contrasting with glioblastoma's pure edema.

Imaging Recommendations for Diagnosis, Staging, and Management of Central Nervous System Neoplasms in Adults: CNS Metastases

Brain metastases (BMs) are the most common central nervous system (CNS) neoplasms, with an increasing incidence that is due in part to an overall increase in primary cancers, improved neuroimaging modalities leading to increased detection, better systemic therapies, and longer patient survival.

OBJECTIVE

To identify cancer patients at a higher risk of developing CNS metastases and to evaluate associated prognostic factors.

METHODS

Review of imaging referral guidelines, response criteria, interval imaging assessment, modality of choice, as well as the association of clinical, serological, and imaging findings as per various cancer societies.

RESULTS

Quantitative response assessment of target and non-target brain metastases as well as an interval imaging protocol set up based on primary histological diagnosis and therapy status are discussed as per various cancer societies and imaging programs.

CONCLUSION

Predictive factors in the primary tumor as well as independent variables of brain metastases like size, number, and response to therapy are necessary in management. The location of CNS metastases, symptomatic disease, as well as follow up imaging findings form a skeletal plan to prognosticate the disease, keeping in mind all the available new advanced therapy options of surgery, radiation, and immunotherapy that improve patient outcome significantly.

Challenges, limitations, and pitfalls of PET and advanced MRI in patients with brain tumors: A report of the PET/RANO group

Brain tumor diagnostics have significantly evolved with the use of positron emission tomography (PET) and advanced magnetic resonance imaging (MRI) techniques. In addition to anatomical MRI, these modalities may provide valuable information for several clinical applications such as differential diagnosis, delineation of tumor extent, prognostication, differentiation between tumor relapse and treatment-related changes, and the evaluation of response to anticancer therapy. In particular, joint recommendations of the Response Assessment in Neuro-Oncology (RANO) Group, the European Association of Neuro-oncology, and major European and American Nuclear Medicine societies highlighted that the additional clinical value of radiolabeled amino acids compared to anatomical MRI alone is outstanding and that its widespread clinical use should be supported. For advanced MRI and its steadily increasing use in clinical practice, the Standardization Subcommittee of the Jumpstarting Brain Tumor Drug Development Coalition provided more recently an updated acquisition protocol for the widely used dynamic susceptibility contrast perfusion MRI. Besides amino acid PET and perfusion MRI, other PET tracers and advanced MRI techniques (e.g. MR spectroscopy) are of considerable clinical interest and are increasingly integrated into everyday clinical practice. Nevertheless, these modalities have shortcomings which should be considered in clinical routine. This comprehensive review provides an overview of potential challenges, limitations, and pitfalls associated with PET imaging and advanced MRI techniques in patients with gliomas or brain metastases. Despite these issues, PET imaging and advanced MRI techniques continue to play an indispensable role in brain tumor management. Acknowledging and mitigating these challenges through interdisciplinary collaboration, standardized protocols, and continuous innovation will further enhance the utility of these modalities in guiding optimal patient care.

Interim FDG-PET improves treatment failure prediction in primary central nervous system lymphoma: An LOC network prospective multicentric study

BACKGROUND

The purpose of our study was to assess the predictive and prognostic role of 2-18F-fluoro-2-deoxy-d-glucose (FDG) positron emission tomography (PET)/MRI during high-dose methotrexate-based chemotherapy (HD-MBC) in de novo primary central nervous system lymphoma (PCNSL) patients aged 60 and above.

METHODS

This prospective multicentric ancillary study included 65 immunocompetent patients who received induction HD-MBC as part of the BLOCAGE01 phase III trial. FDG-PET/MRI were acquired at baseline, post 2 cycles (PET/MRI2), and posttreatment (PET/MRI3). FDG-PET response was dichotomized with "positive" indicating persistent tumor uptake higher than the contralateral mirroring brain region. Performances of FDG-PET and International PCNSL Collaborative Group criteria in predicting induction response, progression-free survival (PFS), and overall survival (OS) were compared.

RESULTS

Of the 48 PET2 scans performed, 9 were positive and aligned with a partial response (PR) on MRI2. Among these, 8 (89%) progressed by the end of the induction phase. In contrast, 35/39 (90%) of PET2-negative patients achieved complete response (CR). Among the 18 discordant responses at interim (PETCR/MRIPR), 83% ultimately achieved CR. Eighty-seven percent of the PET2-negative patients were disease free at 6 months versus 11% of the PET2-positive patients (P < .001). The MRI2 response did not significantly differentiate patients based on their PFS, regardless of whether they were in CR or PR. Both PET2 and MRI2 independently predicted OS in multivariate analysis, with PET2 showing a stronger association.

CONCLUSIONS

Our study highlights the potential of interim FDG-PET for early management of PCNSL patients. Response-driven treatment based on PET2 may guide future clinical trials.

TRIAL

LOCALYZE, NCT03582254, ancillary of phase III clinical trial BLOCAGE01, NCT02313389 (Registered July 10, 2018-retrospectively registered) https://clinicaltrials.gov/ct2/show/NCT03582254?term=LOCALYZE&draw=2&rank=1.

The Use of Apparent Diffusion Coefficient Values for Differentiating Bevacizumab-Related Cytotoxicity from Tumor Recurrence and Radiation Necrosis in Glioblastoma

OBJECTIVES

Glioblastomas (GBM) are the most common primary invasive neoplasms of the brain. Distinguishing between lesion recurrence and different types of treatment related changes in patients with GBM remains challenging using conventional MRI imaging techniques. Therefore, accurate and precise differentiation between true progression or pseudoresponse is crucial in deciding on the appropriate course of treatment. This retrospective study investigated the potential of apparent diffusion coefficient (ADC) map values derived from diffusion-weighted imaging (DWI) as a noninvasive method to increase diagnostic accuracy in treatment response.

METHODS

A cohort of 21 glioblastoma patients (mean age: 59.2 ± 11.8, 12 Male, 9 Female) that underwent treatment with bevacizumab were selected. The ADC values were calculated from the DWI images obtained from a standardized brain protocol across 1.5-T and 3-T MRI scanners. Ratios were calculated for rADC values. Lesions were classified as bevacizumab-induced cytotoxicity based on characteristic imaging features (well-defined regions of restricted diffusion with persistent diffusion restriction over the course of weeks without tissue volume loss and absence of contrast enhancement). The rADC value was compared to these values in radiation necrosis and recurrent lesions, which were concluded in our prior study. The nonparametric Wilcoxon signed rank test with p < 0.05 was used for significance.

RESULTS

The mean ± SD age of the selected patients was 59.2 ± 11.8. ADC values and corresponding mean rADC values for bevacizumab-induced cytotoxicity were 248.1 ± 67.2 and 0.39 ± 0.10, respectively. These results were compared to the ADC values and corresponding mean rADC values of tumor progression and radiation necrosis. Significant differences between rADC values were observed in all three groups (p < 0.001). Bevacizumab-induced cytotoxicity had statistically significant lower ADC values compared to both tumor recurrence and radiation necrosis.

CONCLUSION

The study demonstrates the potential of ADC values as noninvasive imaging biomarkers for differentiating recurrent glioblastoma from radiation necrosis and bevacizumab-induced cytotoxicity.

Role of artificial intelligence in brain tumour imaging

Artificial intelligence (AI) is a rapidly evolving field with many neuro-oncology applications. In this review, we discuss how AI can assist in brain tumour imaging, focusing on machine learning (ML) and deep learning (DL) techniques. We describe how AI can help in lesion detection, differential diagnosis, anatomic segmentation, molecular marker identification, prognostication, and pseudo-progression evaluation. We also cover AI applications in non-glioma brain tumours, such as brain metastasis, posterior fossa, and pituitary tumours. We highlight the challenges and limitations of AI implementation in radiology, such as data quality, standardization, and integration. Based on the findings in the aforementioned areas, we conclude that AI can potentially improve the diagnosis and treatment of brain tumours and provide a path towards personalized medicine and better patient outcomes.

Exploring the Impact of 3D Fast Spin Echo and Inversion Recovery Gradient Echo Sequences Magnetic Resonance Imaging Acquisition on Automated Brain Tumor Segmentation

Objective

To conduct a study comparing the performance of automated segmentation techniques using 2 different contrast-enhanced T1-weighted (CET1) magnetic resonance imaging (MRI) acquisition protocol.

Patients and Methods

We collected 100 preoperative glioblastoma (GBM) MRIs consisting of 50 IR-GRE and 50 3-dimensional fast spin echo (3D-FSE) image sets. Their gold-standard tumor segmentation mask was created based on the expert opinion of a neuroradiologist. Cases were randomly divided into training and test sets. We used the no new UNet (nnUNet) architecture pretrained on the 501-image public data set containing IR-GRE sequence image sets, followed by 2 training rounds with the IR-GRE and 3D-FSE images, respectively. For each patient, in the IR-GRE and 3D-FSE test sets, we had 2 prediction masks, one from the model fine-tuned with the IR-GRE training set and one with 3D-FSE. The dice similarity coefficients (DSCs) of the 2 sets of results for each case in the test sets were compared using the Wilcoxon tests.

Results

Models trained on 3D-FSE images outperformed IR-GRE models in lesion segmentation, with mean DSC differences of 0.057 and 0.022 in the respective test sets. For the 3D-FSE and IR-GRE test sets, the calculated P values comparing DSCs from 2 models were .02 and .61, respectively.

Conclusion

Including 3D-FSE MRI in the training data set improves segmentation performance when segmenting 3D-FSE images.

Where Does Auto-Segmentation for Brain Metastases Radiosurgery Stand Today?

Detection and segmentation of brain metastases (BMs) play a pivotal role in diagnosis, treatment planning, and follow-up evaluations for effective BM management. Given the rising prevalence of BM cases and its predominantly multiple onsets, automated segmentation is becoming necessary in stereotactic radiosurgery. It not only alleviates the clinician's manual workload and improves clinical workflow efficiency but also ensures treatment safety, ultimately improving patient care. Recent strides in machine learning, particularly in deep learning (DL), have revolutionized medical image segmentation, achieving state-of-the-art results. This review aims to analyze auto-segmentation strategies, characterize the utilized data, and assess the performance of cutting-edge BM segmentation methodologies. Additionally, we delve into the challenges confronting BM segmentation and share insights gleaned from our algorithmic and clinical implementation experiences.

Reducing false positives in deep learning-based brain metastasis detection by using both gradient-echo and spin-echo contrast-enhanced MRI: validation in a multi-center diagnostic cohort

OBJECTIVES

To develop a deep learning (DL) for detection of brain metastasis (BM) that incorporates both gradient- and turbo spin-echo contrast-enhanced MRI (dual-enhanced DL) and evaluate it in a clinical cohort in comparison with human readers and DL using gradient-echo-based imaging only (GRE DL).

MATERIALS AND METHODS

DL detection was developed using data from 200 patients with BM (training set) and tested in 62 (internal) and 48 (external) consecutive patients who underwent stereotactic radiosurgery and diagnostic dual-enhanced imaging (dual-enhanced DL) and later guide GRE imaging (GRE DL). The detection sensitivity and positive predictive value (PPV) were compared between two DLs. Two neuroradiologists independently analyzed BM and reference standards for BM were separately drawn by another neuroradiologist. The relative differences (RDs) from the reference standard BM numbers were compared between the DLs and neuroradiologists.

RESULTS

Sensitivity was similar between GRE DL (93%, 95% confidence interval [CI]: 90-96%) and dual-enhanced DL (92% [89-94%]). The PPV of the dual-enhanced DL was higher (89% [86-92%], p < .001) than that of GRE DL (76%, [72-80%]). GRE DL significantly overestimated the number of metastases (false positives; RD: 0.05, 95% CI: 0.00-0.58) compared with neuroradiologists (RD: 0.00, 95% CI: - 0.28, 0.15, p < .001), whereas dual-enhanced DL (RD: 0.00, 95% CI: 0.00-0.15) did not show a statistically significant difference from neuroradiologists (RD: 0.00, 95% CI: - 0.20-0.10, p = .913).

CONCLUSION

The dual-enhanced DL showed improved detection of BM and reduced overestimation compared with GRE DL, achieving similar performance to neuroradiologists.

CLINICAL RELEVANCE STATEMENT

The use of deep learning-based brain metastasis detection with turbo spin-echo imaging reduces false positive detections, aiding in the guidance of stereotactic radiosurgery when gradient-echo imaging alone is employed.

KEY POINTS

•Deep learning for brain metastasis detection improved by using both gradient- and turbo spin-echo contrast-enhanced MRI (dual-enhanced deep learning). •Dual-enhanced deep learning increased true positive detections and reduced overestimation. •Dual-enhanced deep learning achieved similar performance to neuroradiologists for brain metastasis counts.

Autoimmune encephalitis: what the radiologist needs to know

Autoimmune encephalitis is a relatively novel nosological entity characterized by an immune-mediated damage of the central nervous system. While originally described as a paraneoplastic inflammatory phenomenon affecting limbic structures, numerous instances of non-paraneoplastic pathogenesis, as well as extra-limbic involvement, have been characterized. Given the wide spectrum of insidious clinical presentations ranging from cognitive impairment to psychiatric symptoms or seizures, it is crucial to raise awareness about this disease category. In fact, an early diagnosis can be dramatically beneficial for the prognosis both to achieve an early therapeutic intervention and to detect a potential underlying malignancy. In this scenario, the radiologist can be the first to pose the hypothesis of autoimmune encephalitis and refer the patient to a comprehensive diagnostic work-up - including clinical, serological, and neurophysiological assessments.In this article, we illustrate the main radiological characteristics of autoimmune encephalitis and its subtypes, including the typical limbic presentation, the features of extra-limbic involvement, and also peculiar imaging findings. In addition, we review the most relevant alternative diagnoses that should be considered, ranging from other encephalitides to neoplasms, vascular conditions, and post-seizure alterations. Finally, we discuss the most appropriate imaging diagnostic work-up, also proposing a suggested MRI protocol.

Long-Term Results of Stereotactic Radiotherapy in Patients with at Least 10 Brain Metastases at Diagnosis

PURPOSE

to evaluate an SRT approach in patients with at least 10 lesions at the time of BM initial diagnosis.

METHODS

This is a monocentric prospective cohort of patients treated by SRT, followed by a brain MRI every two months. Subsequent SRT could be delivered in cases of new BMs during follow-up. The main endpoints were local control rate (LCR), overall survival (OS), and strategy success rate (SSR). Acute and late toxicity were evaluated.

RESULTS

Seventy patients were included from October 2014 to January 2019, and the most frequent primary diagnosis was non-small-cell lung cancer (N = 36, 51.4%). A total of 1174 BMs were treated at first treatment, corresponding to a median number of 14 BMs per patient. Most of the patients (N = 51, 72.6%) received a single fraction of 20-24 Gy. At 1 year, OS was 62.3%, with a median OS of 19.2 months, and SSR was 77.8%. A cumulative number of 1537 BM were treated over time, corresponding to a median cumulative number of 16 BM per patient. At 1-year, the LCR was 97.3%, with a cumulative incidence of radio-necrosis of 2.1% per lesion. Three patients (4.3%) presented Grade 2 toxicity, and there was no Grade ≥ 3 toxicity. The number of treated BMs and the treatment volume did not influence OS or SSR (p > 0.05).

CONCLUSIONS

SRT was highly efficient in controlling the BM, with minimal side effects. In this setting, an SRT treatment should be proposed even in patients with ≥10 BMs at diagnosis.

Compressed Sensitivity Encoding Artificial Intelligence Accelerates Brain Metastasis Imaging by Optimizing Image Quality and Reducing Scan Time

BACKGROUND AND PURPOSE

Accelerating the image acquisition speed of MR imaging without compromising the image quality is challenging. This study aimed to evaluate the feasibility of contrast-enhanced (CE) 3D T1WI and CE 3D-FLAIR sequences reconstructed with compressed sensitivity encoding artificial intelligence (CS-AI) for detecting brain metastases (BM) and explore the optimal acceleration factor (AF) for clinical BM imaging.

MATERIALS AND METHODS

Fifty-one patients with cancer with suspected BM were included. Fifty participants underwent different customized CE 3D-T1WI or CE 3D-FLAIR sequence scans. Compressed SENSE encoding acceleration 6 (CS6), a commercially available standard sequence, was used as the reference standard. Quantitative and qualitative methods were used to evaluate image quality. The SNR and contrast-to-noise ratio (CNR) were calculated, and qualitative evaluations were independently conducted by 2 neuroradiologists. After exploring the optimal AF, sample images were obtained from 1 patient by using both optimized sequences.

RESULTS

Quantitatively, the CNR of the CS-AI protocol for CE 3D-T1WI and CE 3D-FLAIR sequences was superior to that of the CS protocol under the same AF (P < .05). Compared with reference CS6, the CS-AI groups had higher CNR values (all P < .05), with the CS-AI10 scan having the highest value. The SNR of the CS-AI group was better than that of the reference for both CE 3D-T1WI and CE 3D-FLAIR sequences (all P < .05). Qualitatively, the CS-AI protocol produced higher image quality scores than did the CS protocol with the same AF (all P < .05). In contrast to the reference CS6, the CS-AI group showed good image quality scores until an AF of up to 10 (all P < .05). The CS-AI10 scan provided the optimal images, improving the delineation of normal gray-white matter boundaries and lesion areas (P < .05). Compared with the reference, CS-AI10 showed reductions in scan time of 39.25% and 39.93% for CE 3D-T1WI and CE 3D-FLAIR sequences, respectively.

CONCLUSIONS

CE 3D-T1WI and CE 3D-FLAIR sequences reconstructed with CS-AI for the detection of BM may provide a more effective alternative reconstruction approach than CS. CS-AI10 is suitable for clinical applications, providing optimal image quality and a shortened scan time.

Comparison between postcontrast thin-slice T1-weighted 2D spin echo and 3D T1-weighted SPACE sequences in the detection of brain metastases at 1.5 and 3 T

OBJECTIVES

Accurate detection of metastatic brain lesions (MBL) is critical due to advances in radiosurgery. We compared the results of three readers in detecting MBL using T1-weighted 2D spin echo (SE) and sampling perfection with application-optimized contrasts using different flip angle evolution (SPACE) sequences with whole-brain coverage at both 1.5 T and 3 T.

METHODS

Fifty-six patients evaluated for MBL were included and underwent a standard protocol (1.5 T, n = 37; 3 T, n = 19), including postcontrast T1-weighted SE and SPACE. The rating was performed by three raters in two sessions > six weeks apart. The true number of MBL was determined using all available imaging including follow-up. Intraclass correlations for intra-rater and inter-rater agreement were calculated. Signal intensity ratios (SIR; enhancing lesion, white matter) were determined on a subset of 46 MBL > 4 mm. A paired t-test was used to evaluate postcontrast sequence order and SIR. Reader accuracy was evaluated by the coefficient of determination.

RESULTS

A total of 135 MBL were identified (mean/subject 2.41, SD 6.4). The intra-rater agreement was excellent for all 3 raters (ICC = 0.97-0.992), as was the inter-rater agreement (ICC = 0.995 SE, 0.99 SPACE). Subjective qualitative ratings were lower for SE images; however, signal intensity ratios were higher in SE sequences. Accuracy was high in all readers for both SE (R2 0.95-0.96) and SPACE (R2 0.91-0.96) sequences.

CONCLUSIONS

Although SE sequences are superior to gradient echo sequences in the detection of small MBL, they have long acquisition times and frequent artifacts. We show that T1-weighted SPACE is not inferior to standard thin-slice SE sequences in the detection of MBL at both imaging fields.

CRITICAL RELEVANCE STATEMENT

Our results show the suitability of 3D T1-weighted turbo spin echo (TSE) sequences (SPACE, CUBE, VISTA) in the detection of brain metastases at both 1.5 T and 3 T.

KEY POINTS

• Accurate detection of brain metastases is critical due to advances in radiosurgery. • T1-weighted SE sequences are superior to gradient echo in detecting small metastases. • T1-weighted 3D-TSE sequences may achieve high resolution and relative insensitivity to artifacts. • T1-weighted 3D-TSE sequences have been recommended in imaging brain metastases at 3 T. • We found T1-weighted 3D-TSE equivalent to thin-slice SE at 1.5 T and 3 T.

Artificial intelligence-based MRI radiomics and radiogenomics in glioma

The specific genetic subtypes that gliomas exhibit result in variable clinical courses and the need to involve multidisciplinary teams of neurologists, epileptologists, neurooncologists and neurosurgeons. Currently, the diagnosis of gliomas pivots mainly around the preliminary radiological findings and the subsequent definitive surgical diagnosis (via surgical sampling). Radiomics and radiogenomics present a potential to precisely diagnose and predict survival and treatment responses, via morphological, textural, and functional features derived from MRI data, as well as genomic data. In spite of their advantages, it is still lacking standardized processes of feature extraction and analysis methodology among different research groups, which have made external validations infeasible. Radiomics and radiogenomics can be used to better understand the genomic basis of gliomas, such as tumor spatial heterogeneity, treatment response, molecular classifications and tumor microenvironment immune infiltration. These novel techniques have also been used to predict histological features, grade or even overall survival in gliomas. In this review, workflows of radiomics and radiogenomics are elucidated, with recent research on machine learning or artificial intelligence in glioma.

Integrating multi-modal imaging in radiation treatments for glioblastoma

Advances in diagnostic and treatment technology along with rapid developments in translational research may now allow the realization of precision radiotherapy. Integration of biologically informed multimodality imaging to address the spatial and temporal heterogeneity underlying treatment resistance in glioblastoma is now possible for patient care, with evidence of safety and potential benefit. Beyond their diagnostic utility, several candidate imaging biomarkers have emerged in recent early-phase clinical trials of biologically based radiotherapy, and their definitive assessment in multicenter prospective trials is already in development. In this review, the rationale for clinical implementation of candidate advanced magnetic resonance imaging and positron emission tomography imaging biomarkers to guide personalized radiotherapy, the current landscape, and future directions for integrating imaging biomarkers into radiotherapy for glioblastoma are summarized. Moving forward, response-adaptive radiotherapy using biologically informed imaging biomarkers to address emerging treatment resistance in rational combination with novel systemic therapies may ultimately permit improvements in glioblastoma outcomes and true individualization of patient care.

Deep learning-based metastasis detection in patients with lung cancer to enhance reproducibility and reduce workload in brain metastasis screening with MRI: a multi-center study

OBJECTIVES

To assess whether a deep learning-based system (DLS) with black-blood imaging for brain metastasis (BM) improves the diagnostic workflow in a multi-center setting.

MATERIALS AND METHODS

In this retrospective study, a DLS was developed in 101 patients and validated on 264 consecutive patients (with lung cancer) having newly developed BM from two tertiary university hospitals, which performed black-blood imaging between January 2020 and April 2021. Four neuroradiologists independently evaluated BM either with segmented masks and BM counts provided (with DLS) or not provided (without DLS) on a clinical trial imaging management system (CTIMS). To assess reading reproducibility, BM count agreement between the readers and the reference standard were calculated using limits of agreement (LoA). Readers' workload was assessed with reading time, which was automatically measured on CTIMS, and were compared between with and without DLS using linear mixed models considering the imaging center.

RESULTS

In the validation cohort, the detection sensitivity and positive predictive value of the DLS were 90.2% (95% confidence interval [CI]: 88.1-92.2) and 88.2% (95% CI: 85.7-90.4), respectively. The difference between the readers and the reference counts was larger without DLS (LoA: -0.281, 95% CI: -2.888, 2.325) than with DLS (LoA: -0.163, 95% CI: -2.692, 2.367). The reading time was reduced from mean 66.9 s (interquartile range: 43.2-90.6) to 57.3 s (interquartile range: 33.6-81.0) (P <.001) in the with DLS group, regardless of the imaging center.

CONCLUSION

Deep learning-based BM detection and counting with black-blood imaging improved reproducibility and reduced reading time, on multi-center validation.

Commissioning and implementing a Quality Assurance program for dedicated radiation oncology MRI scanners

PURPOSE

ACR and AAPM task group's guidelines addressing commissioning for dedicated MR simulators were recently published. The goal of the current paper is to present the authors' 2-year experience regarding the commissioning and introduction of a QA program based on these guidelines and an associated automated workflow.

METHODS

All mandatory commissioning tests suggested by AAPM report 284 were performed and results are reported for two MRI scanners (MAGNETOM Sola and Aera). Visual inspection, vendor clinical or service platform, third-party software, or in-house python-based code were used. Automated QA and data analysis was performed via vendor, in-house or third-party software. QATrack+ was used for QA data logging and storage. 3D geometric distortion, B0 inhomogeneity, EPI, and parallel imaging performance were evaluated.

RESULTS

Contrasting with AAPM report 284 recommendations, homogeneity and RF tests were performed monthly. The QA program allowed us to detect major failures over time (shimming, gradient calibration and RF interference). Automated QA, data analysis, and logging allowed fast ACR analysis daily and monthly QA to be performed in 3 h. On the Sola, the average distortion is 1 mm for imaging radii of 250 mm or less. For radii of up to 200 mm, the maximum, average (standard deviation) distortion is 1.2  and 0.4 mm (0.3 mm). Aera values are roughly double the Sola for radii up to 200 mm. EPI geometric distortion, ghosting ratio, and long-term stability were found to be under the maximum recommended values. Parallel imaging SNR ratio was stable and close to the theoretical value (ideal g-factor). No major failures were detected during commissioning.

CONCLUSION

An automated workflow and enhanced QA program allowed to automatically track machine and environmental changes over time and to detect periodic failures and errors that might otherwise have gone unnoticed. The Sola is more geometrically accurate, with a more homogenous B0 field than the Aera.

Deep learning-based detection and quantification of brain metastases on black-blood imaging can provide treatment suggestions: a clinical cohort study

OBJECTIVES

We aimed to evaluate whether deep learning-based detection and quantification of brain metastasis (BM) may suggest treatment options for patients with BMs.

METHODS

The deep learning system (DLS) for detection and quantification of BM was developed in 193 patients and applied to 112 patients that were newly detected on black-blood contrast-enhanced T1-weighted imaging. Patients were assigned to one of 3 treatment suggestion groups according to the European Association of Neuro-Oncology (EANO)-European Society for Medical Oncology (ESMO) recommendations using number and volume of the BMs detected by the DLS: short-term imaging follow-up without treatment (group A), surgery or stereotactic radiosurgery (limited BM, group B), or whole-brain radiotherapy or systemic chemotherapy (extensive BM, group C). The concordance between the DLS-based groups and clinical decisions was analyzed with or without consideration of targeted agents. The performance of distinguishing high-risk (B + C) was calculated.

RESULTS

Among 112 patients (mean age 64.3 years, 63 men), group C had the largest number and volume of BM, followed by group B (4.4 and 851.6 mm3) and A (1.5 and 15.5 mm3). The DLS-based groups were concordant with the actual clinical decisions, with an accuracy of 76.8% (86 of 112). Modified accuracy considering targeted agents was 81.3% (91 of 112). The DLS showed 95% (82/86) sensitivity and 81% (21/26) specificity for distinguishing the high risk.

CONCLUSION

DLS-based detection and quantification of BM have the potential to be helpful in the determination of treatment options for both low- and high-risk groups of limited and extensive BMs.

CLINICAL RELEVANCE STATEMENT

For patients with newly diagnosed brain metastasis, deep learning-based detection and quantification may be used in clinical settings where prompt and accurate treatment decisions are required, which can lead to better patient outcomes.

KEY POINTS

• Deep learning-based brain metastasis detection and quantification showed excellent agreement with ground-truth classifications. • By setting an algorithm to suggest treatment based on the number and volume of brain metastases detected by the deep learning system, the concordance was 81.3%. • When dividing patients into low- and high-risk groups, the sensitivity for detecting the latter was 95%.

Incidence and imaging characteristics of difficult to detect retrospectively identified brain metastases in patients receiving repeat courses of stereotactic radiosurgery

PURPOSE

During stereotactic radiosurgery (SRS) planning for brain metastases (BM), brain MRIs are reviewed to select appropriate targets based on radiographic characteristics. Some BM are difficult to detect and/or definitively identify and may go untreated initially, only to become apparent on future imaging. We hypothesized that in patients receiving multiple courses of SRS, reviewing the initial planning MRI would reveal early evidence of lesions that developed into metastases requiring SRS.

METHODS

Patients undergoing two or more courses of SRS to BM within 6 months between 2016 and 2018 were included in this single-institution, retrospective study. Brain MRIs from the initial course were reviewed for lesions at the same location as subsequently treated metastases; if present, this lesion was classified as a "retrospectively identified metastasis" or RIM. RIMs were subcategorized as meeting or not meeting diagnostic imaging criteria for BM (+ DC or -DC, respectively).

RESULTS

Among 683 patients undergoing 923 SRS courses, 98 patients met inclusion criteria. There were 115 repeat courses of SRS, with 345 treated metastases in the subsequent course, 128 of which were associated with RIMs found in a prior MRI. 58% of RIMs were + DC. 17 (15%) of subsequent courses consisted solely of metastases associated with + DC RIMs.

CONCLUSION

Radiographic evidence of brain metastases requiring future treatment was occasionally present on brain MRIs from prior SRS treatments. Most RIMs were + DC, and some subsequent SRS courses treated only + DC RIMs. These findings suggest enhanced BM detection might enable earlier treatment and reduce the need for additional SRS.