Multi-class glioma segmentation on real-world data with missing MRI sequences: comparison of three deep learning algorithms

This study tests the generalisability of three Brain Tumor Segmentation (BraTS) challenge models using a multi-center dataset of varying image quality and incomplete MRI datasets. In this retrospective study, DeepMedic, no-new-Unet (nn-Unet), and NVIDIA-net (nv-Net) were trained and tested using manual segmentations from preoperative MRI of glioblastoma (GBM) and low-grade gliomas (LGG) from the BraTS 2021 dataset (1251 in total), in addition to 275 GBM and 205 LGG acquired clinically across 12 hospitals worldwide. Data was split into 80% training, 5% validation, and 15% internal test data. An additional external test-set of 158 GBM and 69 LGG was used to assess generalisability to other hospitals' data. All models' median Dice similarity coefficient (DSC) for both test sets were within, or higher than, previously reported human inter-rater agreement (range of 0.74-0.85). For both test sets, nn-Unet achieved the highest DSC (internal = 0.86, external = 0.93) and the lowest Hausdorff distances (10.07, 13.87 mm, respectively) for all tumor classes (p < 0.001). By applying Sparsified training, missing MRI sequences did not statistically affect the performance. nn-Unet achieves accurate segmentations in clinical settings even in the presence of incomplete MRI datasets. This facilitates future clinical adoption of automated glioma segmentation, which could help inform treatment planning and glioma monitoring.

Segmentation of glioblastomas in early post-operative multi-modal MRI with deep neural networks

Extent of resection after surgery is one of the main prognostic factors for patients diagnosed with glioblastoma. To achieve this, accurate segmentation and classification of residual tumor from post-operative MR images is essential. The current standard method for estimating it is subject to high inter- and intra-rater variability, and an automated method for segmentation of residual tumor in early post-operative MRI could lead to a more accurate estimation of extent of resection. In this study, two state-of-the-art neural network architectures for pre-operative segmentation were trained for the task. The models were extensively validated on a multicenter dataset with nearly 1000 patients, from 12 hospitals in Europe and the United States. The best performance achieved was a 61% Dice score, and the best classification performance was about 80% balanced accuracy, with a demonstrated ability to generalize across hospitals. In addition, the segmentation performance of the best models was on par with human expert raters. The predicted segmentations can be used to accurately classify the patients into those with residual tumor, and those with gross total resection.

P11.37.B When to resect or biopsy for patients with supratentorial glioblastoma: a multivariable prediction model

Background

The prospects of a patient with suspected glioblastoma may rely heavily on the indication for surgical resection versus biopsy only. Biopsy percentages vary considerably across hospitals and guidelines for treatment of glioblastoma lack criteria for surgical decision-making. To identify patient and tumor characteristics associated with the decision to resect or biopsy a glioblastoma and to develop and validate a prediction model for decision support.

Material and Methods

Clinical data and pre-operative MRI scans were collected for adults who underwent first-time surgery for supratentorial glioblastoma from a registry-based cohort study of 12 hospitals from the Netherlands, Germany, France, Italy, and the United States between 1st of January 2007 and 31st of December 2011. The main outcome was the type of surgical procedure: surgical resection or biopsy only. Predictors were patient- and tumor-related characteristics. Radiological factors were extracted from MRI using an automated tumor segmentation method. A prediction model was constructed using multivariable logistic regression analysis. The model was cross-validated and externally validated with a leave-one-hospital-out approach.

Results

Out of 1053 patients treated for glioblastoma, 28% underwent biopsy only. Biopsy rates varied from 15-40% across hospitals. The prediction model showed excellent discrimination with an average area under the curve of 0.86. Of the patient-related characteristics, younger age was associated more with resection and Karnofsky Performance Score of 60 or less with biopsy. Of the tumor-related characteristics, a location in the right hemisphere, unifocality, no tumor midline crossing, and no involvement of the cortical spinal tract, were associated with resection, as well as a high expected resectability index, a location in the right occipital lobe, and a higher percentage of tumor in Schaefer’s dorsal or ventral attention, limbic, and default networks. External validation proved acceptable to outstanding discrimination with areas under the curve ranging between 0.79 and 0.92 for hospitals.

Conclusion

A prediction model is presented and validated to support the decision to resect or to biopsy a patient with a suspected supratentorial glioblastoma. In this prediction model, tumor-related characteristics were more informative than patient-related factors. This may support surgical decision-making for individual patients, or facilitate comparisons of patient cohorts between surgeons or institutions.

Preoperative Brain Tumor Imaging: Models and Software for Segmentation and Standardized Reporting

For patients suffering from brain tumor, prognosis estimation and treatment decisions are made by a multidisciplinary team based on a set of preoperative MR scans. Currently, the lack of standardized and automatic methods for tumor detection and generation of clinical reports, incorporating a wide range of tumor characteristics, represents a major hurdle. In this study, we investigate the most occurring brain tumor types: glioblastomas, lower grade gliomas, meningiomas, and metastases, through four cohorts of up to 4,000 patients. Tumor segmentation models were trained using the AGU-Net architecture with different preprocessing steps and protocols. Segmentation performances were assessed in-depth using a wide-range of voxel and patient-wise metrics covering volume, distance, and probabilistic aspects. Finally, two software solutions have been developed, enabling an easy use of the trained models and standardized generation of clinical reports: Raidionics and Raidionics-Slicer. Segmentation performances were quite homogeneous across the four different brain tumor types, with an average true positive Dice ranging between 80 and 90%, patient-wise recall between 88 and 98%, and patient-wise precision around 95%. In conjunction to Dice, the identified most relevant other metrics were the relative absolute volume difference, the variation of information, and the Hausdorff, Mahalanobis, and object average symmetric surface distances. With our Raidionics software, running on a desktop computer with CPU support, tumor segmentation can be performed in 16–54 s depending on the dimensions of the MRI volume. For the generation of a standardized clinical report, including the tumor segmentation and features computation, 5–15 min are necessary. All trained models have been made open-access together with the source code for both software solutions and validation metrics computation. In the future, a method to convert results from a set of metrics into a final single score would be highly desirable for easier ranking across trained models. In addition, an automatic classification of the brain tumor type would be necessary to replace manual user input. Finally, the inclusion of post-operative segmentation in both software solutions will be key for generating complete post-operative standardized clinical reports.

Glioblastoma Surgery Imaging-Reporting and Data System: Validation and Performance of the Automated Segmentation Task

For patients with presumed glioblastoma, essential tumor characteristics are determined from preoperative MR images to optimize the treatment strategy. This procedure is time-consuming and subjective, if performed by crude eyeballing or manually. The standardized GSI-RADS aims to provide neurosurgeons with automatic tumor segmentations to extract tumor features rapidly and objectively. In this study, we improved automatic tumor segmentation and compared the agreement with manual raters, describe the technical details of the different components of GSI-RADS, and determined their speed. Two recent neural network architectures were considered for the segmentation task: nnU-Net and AGU-Net. Two preprocessing schemes were introduced to investigate the tradeoff between performance and processing speed. A summarized description of the tumor feature extraction and standardized reporting process is included. The trained architectures for automatic segmentation and the code for computing the standardized report are distributed as open-source and as open-access software. Validation studies were performed on a dataset of 1594 gadolinium-enhanced T1-weighted MRI volumes from 13 hospitals and 293 T1-weighted MRI volumes from the BraTS challenge. The glioblastoma tumor core segmentation reached a Dice score slightly below 90%, a patientwise F1-score close to 99%, and a 95th percentile Hausdorff distance slightly below 4.0 mm on average with either architecture and the heavy preprocessing scheme. A patient MRI volume can be segmented in less than one minute, and a standardized report can be generated in up to five minutes. The proposed GSI-RADS software showed robust performance on a large collection of MRI volumes from various hospitals and generated results within a reasonable runtime.

On the cutting edge of glioblastoma surgery: where neurosurgeons agree and disagree on surgical decisions

OBJECTIVE

The aim of glioblastoma surgery is to maximize the extent of resection while preserving functional integrity. Standards are lacking for surgical decision-making, and previous studies indicate treatment variations. These shortcomings reflect the need to evaluate larger populations from different care teams. In this study, the authors used probability maps to quantify and compare surgical decision-making throughout the brain by 12 neurosurgical teams for patients with glioblastoma.

METHODS

The study included all adult patients who underwent first-time glioblastoma surgery in 2012-2013 and were treated by 1 of the 12 participating neurosurgical teams. Voxel-wise probability maps of tumor location, biopsy, and resection were constructed for each team to identify and compare patient treatment variations. Brain regions with different biopsy and resection results between teams were identified and analyzed for patient functional outcome and survival.

RESULTS

The study cohort consisted of 1087 patients, of whom 363 underwent a biopsy and 724 a resection. Biopsy and resection decisions were generally comparable between teams, providing benchmarks for probability maps of resections and biopsies for glioblastoma. Differences in biopsy rates were identified for the right superior frontal gyrus and indicated variation in biopsy decisions. Differences in resection rates were identified for the left superior parietal lobule, indicating variations in resection decisions.

CONCLUSIONS

Probability maps of glioblastoma surgery enabled capture of clinical practice decisions and indicated that teams generally agreed on which region to biopsy or to resect. However, treatment variations reflecting clinical dilemmas were observed and pinpointed by using the probability maps, which could therefore be useful for quality-of-care discussions between surgical teams for patients with glioblastoma.

Glioblastoma Surgery Imaging-Reporting and Data System: Standardized Reporting of Tumor Volume, Location, and Resectability Based on Automated Segmentations

Treatment decisions for patients with presumed glioblastoma are based on tumor characteristics available from a preoperative MR scan. Tumor characteristics, including volume, location, and resectability, are often estimated or manually delineated. This process is time consuming and subjective. Hence, comparison across cohorts, trials, or registries are subject to assessment bias. In this study, we propose a standardized Glioblastoma Surgery Imaging Reporting and Data System (GSI-RADS) based on an automated method of tumor segmentation that provides standard reports on tumor features that are potentially relevant for glioblastoma surgery. As clinical validation, we determine the agreement in extracted tumor features between the automated method and the current standard of manual segmentations from routine clinical MR scans before treatment. In an observational consecutive cohort of 1596 adult patients with a first time surgery of a glioblastoma from 13 institutions, we segmented gadolinium-enhanced tumor parts both by a human rater and by an automated algorithm. Tumor features were extracted from segmentations of both methods and compared to assess differences, concordance, and equivalence. The laterality, contralateral infiltration, and the laterality indices were in excellent agreement. The native and normalized tumor volumes had excellent agreement, consistency, and equivalence. Multifocality, but not the number of foci, had good agreement and equivalence. The location profiles of cortical and subcortical structures were in excellent agreement. The expected residual tumor volumes and resectability indices had excellent agreement, consistency, and equivalence. Tumor probability maps were in good agreement. In conclusion, automated segmentations are in excellent agreement with manual segmentations and practically equivalent regarding tumor features that are potentially relevant for neurosurgical purposes. Standard GSI-RADS reports can be generated by open access software.

Timing of glioblastoma surgery and patient outcomes: a multicenter cohort study

Background

The impact of time-to-surgery on clinical outcome for patients with glioblastoma has not been determined. Any delay in treatment is perceived as detrimental, but guidelines do not specify acceptable timings. In this study, we relate the time to glioblastoma surgery with the extent of resection and residual tumor volume, performance change, and survival, and we explore the identification of patients for urgent surgery.

Methods

Adults with first-time surgery in 2012–2013 treated by 12 neuro-oncological teams were included in this study. We defined time-to-surgery as the number of days between the diagnostic MR scan and surgery. The relation between time-to-surgery and patient and tumor characteristics was explored in time-to-event analysis and proportional hazard models. Outcome according to time-to-surgery was analyzed by volumetric measurements, changes in performance status, and survival analysis with patient and tumor characteristics as modifiers.

Results

Included were 1033 patients of whom 729 had a resection and 304 a biopsy. The overall median time-to-surgery was 13 days. Surgery was within 3 days for 235 (23%) patients, and within a month for 889 (86%). The median volumetric doubling time was 22 days. Lower performance status (hazard ratio [HR] 0.942, 95% confidence interval [CI] 0.893–0.994) and larger tumor volume (HR 1.012, 95% CI 1.010–1.014) were independently associated with a shorter time-to-surgery. Extent of resection, residual tumor volume, postoperative performance change, and overall survival were not associated with time-to-surgery.

Conclusions

With current decision-making for urgent surgery in selected patients with glioblastoma and surgery typically within 1 month, we found equal extent of resection, residual tumor volume, performance status, and survival after longer times-to-surgery.

Quantifying eloquent locations for glioblastoma surgery using resection probability maps

OBJECTIVE

Decisions in glioblastoma surgery are often guided by presumed eloquence of the tumor location. The authors introduce the "expected residual tumor volume" (eRV) and the "expected resectability index" (eRI) based on previous decisions aggregated in resection probability maps. The diagnostic accuracy of eRV and eRI to predict biopsy decisions, resectability, functional outcome, and survival was determined.

METHODS

Consecutive patients with first-time glioblastoma surgery in 2012-2013 were included from 12 hospitals. The eRV was calculated from the preoperative MR images of each patient using a resection probability map, and the eRI was derived from the tumor volume. As reference, Sawaya's tumor location eloquence grades (EGs) were classified. Resectability was measured as observed extent of resection (EOR) and residual volume, and functional outcome as change in Karnofsky Performance Scale score. Receiver operating characteristic curves and multivariable logistic regression were applied.

RESULTS

Of 915 patients, 674 (74%) underwent a resection with a median EOR of 97%, functional improvement in 71 (8%), functional decline in 78 (9%), and median survival of 12.8 months. The eRI and eRV identified biopsies and EORs of at least 80%, 90%, or 98% better than EG. The eRV and eRI predicted observed residual volumes under 10, 5, and 1 ml better than EG. The eRV, eRI, and EG had low diagnostic accuracy for functional outcome changes. Higher eRV and lower eRI were strongly associated with shorter survival, independent of known prognostic factors.

CONCLUSIONS

The eRV and eRI predict biopsy decisions, resectability, and survival better than eloquence grading and may be useful preoperative indices to support surgical decisions.

Between-hospital variation in mortality and survival after glioblastoma surgery in the Dutch Quality Registry for Neuro Surgery

Purpose

Standards for surgical decisions are unavailable, hence treatment decisions can be personalized, but also introduce variation in treatment and outcome. National registrations seek to monitor healthcare quality. The goal of the study is to measure between-hospital variation in risk-standardized survival outcome after glioblastoma surgery and to explore the association between survival and hospital characteristics in conjunction with patient-related risk factors.

Methods

Data of 2,409 adults with first-time glioblastoma surgery at 14 hospitals were obtained from a comprehensive, prospective population-based Quality Registry Neuro Surgery in The Netherlands between 2011 and 2014. We compared the observed survival with patient-specific risk-standardized expected early (30-day) mortality and late (2-year) survival, based on age, performance, and treatment year. We analyzed funnel plots, logistic regression and proportional hazards models.

Results

Overall 30-day mortality was 5.2% and overall 2-year survival was 13.5%. Median survival varied between 4.8 and 14.9 months among hospitals, and biopsy percentages ranged between 16 and 73%. One hospital had lower than expected early mortality, and four hospitals had lower than expected late survival. Higher case volume was related with lower early mortality (P = 0.031). Patient-related risk factors (lower age; better performance; more recent years of treatment) were significantly associated with longer overall survival. Of the hospital characteristics, longer overall survival was associated with lower biopsy percentage (HR 2.09, 1.34–3.26, P = 0.001), and not with academic setting, nor with case volume.

Conclusions

Hospitals vary more in late survival than early mortality after glioblastoma surgery. Widely varying biopsy percentages indicate treatment variation. Patient-related factors have a stronger association with overall survival than hospital-related factors.

Electronic supplementary material

The online version of this article (10.1007/s11060-019-03229-5) contains supplementary material, which is available to authorized users.

Observer variability of reference tissue selection for relativecerebral blood volume measurements in glioma patients

Objectives

To assess observer variability of different reference tissues used for relative CBV (rCBV) measurements in DSC-MRI of glioma patients.

Methods

In this retrospective study, three observers measured rCBV in DSC-MR images of 44 glioma patients on two occasions. rCBV is calculated by the CBV in the tumour hotspot/the CBV of a reference tissue at the contralateral side for normalization. One observer annotated the tumour hotspot that was kept constant for all measurements. All observers annotated eight reference tissues of normal white and grey matter. Observer variability was evaluated using the intraclass correlation coefficient (ICC), coefficient of variation (CV) and Bland-Altman analyses.

Results

For intra-observer, the ICC ranged from 0.50–0.97 (fair–excellent) for all reference tissues. The CV ranged from 5.1–22.1 % for all reference tissues and observers. For inter-observer, the ICC for all pairwise observer combinations ranged from 0.44–0.92 (poor–excellent). The CV ranged from 8.1–31.1 %. Centrum semiovale was the only reference tissue that showed excellent intra- and inter-observer agreement (ICC>0.85) and lowest CVs (<12.5 %). Bland-Altman analyses showed that mean differences for centrum semiovale were close to zero.

Conclusion

Selecting contralateral centrum semiovale as reference tissue for rCBV provides the lowest observer variability.

Key Points

• Reference tissue selection for rCBV measurements adds variability to rCBV measurements.

• rCBV measurements vary depending on the choice of reference tissue.

• Observer variability of reference tissue selection varies between poor and excellent.

• Centrum semiovale as reference tissue for rCBV provides the lowest observer variability.

Quantitative short echo time 1H MRSI of the peripheral edematous region of human brain tumors in the differentiation between glioblastoma, metastasis, and meningioma

PURPOSE

To assess metabolite levels in peritumoral edematous (PO) and surrounding apparently normal (SAN) brain regions of glioblastoma, metastasis, and meningioma in humans with (1)H-MRSI to find biomarkers that can discriminate between tumors and characterize infiltrative tumor growth.

MATERIALS AND METHODS

Magnetic resonance (MR) spectra (semi-LASER MRSI, 30 msec echo time, 3T) were selected from regions of interest (ROIs) under MRI guidance, and after quality control of MR spectra. Statistical testing between patient groups was performed for mean metabolite ratios of an entire ROI and for the highest value within that ROI.

RESULTS

The highest ratios of the level of choline compounds and the sum of myo-inositol and glycine over N-acetylaspartate and creatine compounds were significantly increased in PO regions of glioblastoma versus that of metastasis and meningioma. In the SAN region of glioblastoma some of these ratios were increased. Differences were less prominent for metabolite levels averaged over entire ROIs.

CONCLUSION

Specific metabolite ratios in PO and SAN regions can be used to discriminate glioblastoma from metastasis and meningioma. An analysis of these ratios averaged over entire ROIs and those with most abnormal values indicates that infiltrative tumor growth in glioblastoma is inhomogeneous and extends into the SAN region.

In vivo 13C magnetic resonance spectroscopy of a human brain tumor after application of 13C-1-enriched glucose

OBJECTIVES

As a unique tool to assess metabolic fluxes noninvasively, (13)C magnetic resonance spectroscopy (MRS) could help to characterize and understand malignancy in human tumors. However, its low sensitivity has hampered applications in patients. The aim of this study was to demonstrate that with sensitivity-optimized localized (13)C MRS and intravenous infusion of [1-(13)C]glucose under euglycemia, it is possible to assess the dynamic conversion of glucose into its metabolic products in vivo in human glioma tissue.

MATERIALS AND METHODS

Measurements were done at 3 T with a broadband single RF channel and a quadrature (13)C surface coil inserted in a (1)H volume coil. A (1)H/(13)C polarization transfer sequence was applied, modified for localized acquisition, alternatively in two (50 ml) voxels, one encompassing the tumor and the other normal brain tissue.

RESULTS

After about 20 min of [1-(13)C]glucose infusion, a [3-(13)C]lactate signal appeared among several resonances of metabolic products of glucose in MR spectra of the tumor voxel. The resonance of [3-(13)C]lactate was absent in MR spectra from contralateral tissue. In addition, the intensity of [1-(13)C]glucose signals in the tumor area was about 50% higher than that in normal tissue, likely reflecting more glucose in extracellular space due to a defective blood-brain barrier. The signal intensity for metabolites produced in or via the tricarboxylic acid (TCA) cycle was lower in the tumor than in the contralateral area, albeit that the ratios of isotopomer signals were comparable.

CONCLUSION

With an improved (13)C MRS approach, the uptake of glucose and its conversion into metabolites such as lactate can be monitored noninvasively in vivo in human brain tumors. This opens the way to assessing metabolic activity in human tumor tissue.

Fibrodysplasia ossificans progressiva and corneal keloid

PURPOSE

To report on the formation of a corneal keloid in a patient with fibrodysplasia ossificans progressiva after excision of a pterygium-like lesion.

METHODS

Clinical and pathophysiological observations and hypothesis concerning pathophysiological mechanisms.

RESULTS

An 11-year-old boy with fibrodysplasia ossificans progressiva presented with progressive development of a growing white scar in the central part of the left cornea after excision of a pterygium-like lesion.

DISCUSSION

To the best of our knowledge, this is the first case report of a corneal keloid formation in a patient with fibrodysplasia ossificans progressiva. Pathophysiological mechanisms are considered. Therapeutic options are discussed.