A Deep Learning Model for Preoperative Differentiation of Glioblastoma, Brain Metastasis and Primary Central Nervous System Lymphoma: A Pilot Study

Deep learning models for rapid discrimination of high-grade gliomas from solitary brain metastases using multi-plane T1-weighted contrast-enhanced (T1CE) images

Background

High-grade gliomas (HGG) and solitary brain metastases (SBM) are two common types of brain tumors in middle-aged and elderly patients. HGG and SBM display a high degree of similarity on magnetic resonance imaging (MRI) images. Consequently, differential diagnosis using preoperative MRI remains challenging. This study developed deep learning models that used pre-operative T1-weighted contrast-enhanced (T1CE) MRI images to differentiate between HGG and SBM before surgery.

Methods

By comparing various convolutional neural network models using T1CE image data from The First Medical Center of the Chinese PLA General Hospital and The Second People's Hospital of Yibin (Data collection for this study spanned from January 2016 to December 2023), it was confirmed that the GoogLeNet model exhibited the highest discriminative performance. Additionally, we evaluated the individual impact of the tumoral core and peritumoral edema regions on the network's predictive performance. Finally, we adopted a slice-based voting method to assess the accuracy of the validation dataset and evaluated patient prediction performance on an additional test dataset.

Results

The GoogLeNet model, in a five-fold cross-validation using multi-plane T1CE slices (axial, coronal, and sagittal) from 180 patients, achieved an average patient accuracy of 92.78%, a sensitivity of 95.56%, and a specificity of 90.00%. Moreover, on an external test set of 29 patients, the model achieved an accuracy of 89.66%, a sensitivity of 90.91%, and a specificity of 83.33%, with an area under the curve of 0.939 [95% confidence interval (CI): 0.842-1.000].

Conclusions

GoogLeNet performed better than previous methods at differentiating HGG from SBM, even for core and peritumoral edema in both. HGG and SBM could be fast screened using this end-to-end approach, improving workflow for both tumor treatments.

Predictors for the Differentiation between Glioblastoma, Primary Central Nervous System Lymphoma, and Metastasis in Patients with a Solitary Enhancing Intracranial Mass

Introduction  Differentiation between glioblastoma (GBM), primary central nervous system lymphoma (PCNSL), and metastasis is important in decision-making before surgery. However, these malignant brain tumors have overlapping features. This study aimed to identify predictors differentiating between GBM, PCNSL, and metastasis. Materials and Methods  Patients with a solitary intracranial enhancing tumor and a histopathological diagnosis of GBM, PCNSL, or metastasis were investigated. All patients with intracranial lymphoma had PCNSL without extracranial involvement. Demographic, clinical, and radiographic data were analyzed to determine their associations with the tumor types. Results  The predictors associated with GBM were functional impairment ( p  = 0.001), large tumor size ( p  < 0.001), irregular tumor margin ( p  < 0.001), heterogeneous contrast enhancement ( p  < 0.001), central necrosis ( p  < 0.001), intratumoral hemorrhage ( p  = 0.018), abnormal flow void ( p  < 0.001), and hypodensity component on noncontrast cranial computed tomography (CT) scan ( p  < 0.001). The predictors associated with PCNSL comprised functional impairment ( p  = 0.005), deep-seated tumor location ( p  = 0.006), homogeneous contrast enhancement ( p  < 0.001), absence of cystic appearance ( p  = 0.008), presence of hypointensity component on precontrast cranial T1-weighted magnetic resonance imaging (MRI; p  = 0.027), and presence of isodensity component on noncontrast cranial CT ( p  < 0.008). Finally, the predictors for metastasis were an infratentorial ( p  < 0.001) or extra-axial tumor location ( p  = 0.035), smooth tumor margin ( p  < 0.001), and presence of isointensity component on cranial fluid-attenuated inversion recovery MRI ( p  = 0.047). Conclusion  These predictors may be used to differentiate between GBM, PCNSL, and metastasis, and they are useful in clinical management.

Deep learning aided preoperative diagnosis of primary central nervous system lymphoma

The preoperative distinction between glioblastoma (GBM) and primary central nervous system lymphoma (PCNSL) can be difficult, even for experts, but is highly relevant. We aimed to develop an easy-to-use algorithm, based on a convolutional neural network (CNN) to preoperatively discern PCNSL from GBM and systematically compare its performance to experienced neurosurgeons and radiologists. To this end, a CNN-based on DenseNet169 was trained with the magnetic resonance (MR)-imaging data of 68 PCNSL and 69 GBM patients and its performance compared to six trained experts on an external test set of 10 PCNSL and 10 GBM. Our neural network predicted PCNSL with an accuracy of 80% and a negative predictive value (NPV) of 0.8, exceeding the accuracy achieved by clinicians (73%, NPV 0.77). Combining expert rating with automated diagnosis in those cases where experts dissented yielded an accuracy of 95%. Our approach has the potential to significantly augment the preoperative radiological diagnosis of PCNSL.

Artificial intelligence in neuroimaging of brain tumors: reality or still promise?

PURPOSE OF REVIEW

To provide an updated overview of artificial intelligence (AI) applications in neuro-oncologic imaging and discuss current barriers to wider clinical adoption.

RECENT FINDINGS

A wide variety of AI applications in neuro-oncologic imaging have been developed and researched, spanning tasks from pretreatment brain tumor classification and segmentation, preoperative planning, radiogenomics, prognostication and survival prediction, posttreatment surveillance, and differentiating between pseudoprogression and true disease progression. While earlier studies were largely based on data from a single institution, more recent studies have demonstrated that the performance of these algorithms are also effective on external data from other institutions. Nevertheless, most of these algorithms have yet to see widespread clinical adoption, given the lack of prospective studies demonstrating their efficacy and the logistical difficulties involved in clinical implementation.

SUMMARY

While there has been significant progress in AI and neuro-oncologic imaging, clinical utility remains to be demonstrated. The next wave of progress in this area will be driven by prospective studies measuring outcomes relevant to clinical practice and go beyond retrospective studies which primarily aim to demonstrate high performance.

Deep learning for image classification between primary central nervous system lymphoma and glioblastoma in corpus callosal tumors

Objectives

It can be challenging in some situations to distinguish primary central nervous system lymphoma (PCNSL) from glioblastoma (GBM) based on magnetic resonance imaging (MRI) scans, especially those involving the corpus callosum. The objective of this study was to assess the diagnostic performance of deep learning (DL) models between PCNSLs and GBMs in corpus callosal tumors.

Materials and Methods

The axial T1-weighted gadolinium-enhanced MRI scans of 274 individuals with pathologically confirmed PCNSL (n = 94) and GBM (n = 180) were examined. After image pooling, pre-operative MRI scans were randomly split with an 80/20 procedure into a training dataset (n = 709) and a testing dataset (n = 177) for DL model development. Therefore, the DL model was deployed as a web application and validated with the unseen images (n = 114) and area under the receiver operating characteristic curve (AUC); other outcomes were calculated to assess the discrimination performance.

Results

The first baseline DL model had an AUC of 0.77 for PCNSL when evaluated with unseen images. The 2nd model with ridge regression regularization and the 3rd model with drop-out regularization increased an AUC of 0.83 and 0.84. In addition, the last model with data augmentation yielded an AUC of 0.57.

Conclusion

DL with regularization may provide useful diagnostic information to help doctors distinguish PCNSL from GBM.

Combining amide proton transfer-weighted and arterial spin labeling imaging to differentiate solitary brain metastases from glioblastomas

PURPOSE

To evaluate the clinical utility of amide proton transfer-weighted imaging (APTw) and arterial spin labeling (ASL) in differentiating solitary brain metastases (SBMs) from glioblastomas (GBMs).

METHODS

Forty-eight patients diagnosed with brain tumors were enrolled. All patients underwent conventional MRI, APTw, and ASL scans on a 3.0 T MRI system. The mean APTw value and mean cerebral blood flow (CBF) value were measured. The differences in various parameters between GBMs and SBMs were assessed using the independent-samples t-test. The quantitative performance of these MRI parameters in distinguishing between GBMs and SBMs was evaluated using receiver operating characteristic (ROC) curve analysis.

RESULTS

GBMs exhibited significantly higher APTw and CBF values in peritumoral regions compared with SBMs (P < 0.05). There was no significant difference between SBMs and GBMs in tumor cores. APTw MRI had a higher diagnostic efficiency in differentiating SBMs from GBMs (area under the curve [AUC]: 0.864; 75.0% sensitivity and 81.8% specificity). Combined use of APTw and CBF value increased the AUC to 0.927.

CONCLUSION

APTw may be superior to ASL for distinguishing between SBMs and GBMs. Combination of APTw and ASL showed better discrimination and a superior diagnostic performance.

Attention Deep Feature Extraction from Brain MRIs in Explainable Mode: DGXAINet

Artificial intelligence models do not provide information about exactly how the predictions are reached. This lack of transparency is a major drawback. Particularly in medical applications, interest in explainable artificial intelligence (XAI), which helps to develop methods of visualizing, explaining, and analyzing deep learning models, has increased recently. With explainable artificial intelligence, it is possible to understand whether the solutions offered by deep learning techniques are safe. This paper aims to diagnose a fatal disease such as a brain tumor faster and more accurately using XAI methods. In this study, we preferred datasets that are widely used in the literature, such as the four-class kaggle brain tumor dataset (Dataset I) and the three-class figshare brain tumor dataset (Dataset II). To extract features, a pre-trained deep learning model is chosen. DenseNet201 is used as the feature extractor in this case. The proposed automated brain tumor detection model includes five stages. First, training of brain MR images with DenseNet201, the tumor area was segmented with GradCAM. The features were extracted from DenseNet201 trained using the exemplar method. Extracted features were selected with iterative neighborhood component (INCA) feature selector. Finally, the selected features were classified using support vector machine (SVM) with 10-fold cross-validation. An accuracy of 98.65% and 99.97%, were obtained for Datasets I and II, respectively. The proposed model obtained higher performance than the state-of-the-art methods and can be used to aid radiologists in their diagnosis.

Artificial intelligence applications in pediatric oncology diagnosis

Artificial intelligence (AI) algorithms have been applied in abundant medical tasks with high accuracy and efficiency. Physicians can improve their diagnostic efficiency with the assistance of AI techniques for improving the subsequent personalized treatment and surveillance. AI algorithms fundamentally capture data, identify underlying patterns, achieve preset endpoints, and provide decisions and predictions about real-world events with working principles of machine learning and deep learning. AI algorithms with sufficient graphic processing unit power have been demonstrated to provide timely diagnostic references based on preliminary training of large amounts of clinical and imaging data. The sample size issue is an inevitable challenge for pediatric oncology considering its low morbidity and individual heterogeneity. However, this problem may be solved in the near future considering the exponential advancements of AI algorithms technically to decrease the dependence of AI operation on the amount of data sets and the efficiency of computing power. For instance, it could be a feasible solution by shifting convolutional neural networks (CNNs) from adults and sharing CNN algorithms across multiple institutions besides original data. The present review provides important insights into emerging AI applications for the diagnosis of pediatric oncology by systematically overviewing of up-to-date literature.

Differentiation of Glioblastoma and Brain Metastases by MRI-Based Oxygen Metabolomic Radiomics and Deep Learning

Glioblastoma (GB) and brain metastasis (BM) are the most frequent types of brain tumors in adults. Their therapeutic management is quite different and a quick and reliable initial characterization has a significant impact on clinical outcomes. However, the differentiation of GB and BM remains a major challenge in today's clinical neurooncology due to their very similar appearance in conventional magnetic resonance imaging (MRI). Novel metabolic neuroimaging has proven useful for improving diagnostic performance but requires artificial intelligence for implementation in clinical routines. Here; we investigated whether the combination of radiomic features from MR-based oxygen metabolism ("oxygen metabolic radiomics") and deep convolutional neural networks (CNNs) can support reliably pre-therapeutic differentiation of GB and BM in a clinical setting. A self-developed one-dimensional CNN combined with radiomic features from the cerebral metabolic rate of oxygen (CMRO₂) was clearly superior to human reading in all parameters for classification performance. The radiomic features for tissue oxygen saturation (mitoPO₂; i.e., tissue hypoxia) also showed better diagnostic performance compared to the radiologists. Interestingly, both the mean and median values for quantitative CMRO₂ and mitoPO₂ values did not differ significantly between GB and BM. This demonstrates that the combination of radiomic features and DL algorithms is more efficient for class differentiation than the comparison of mean or median values. Oxygen metabolic radiomics and deep neural networks provide insights into brain tumor phenotype that may have important diagnostic implications and helpful in clinical routine diagnosis.

Progress of radiological‑pathological workflows in the differential diagnosis between primary central nervous system lymphoma and high‑grade glioma (Review)

Primary central nervous system lymphoma (PCNSL) and high‑grade glioma (HGG) are distinct entities of the CNS with completely distinct treatments. The treatment of PCNSL is chemotherapy‑based, while surgery is the first choice for HGG. However, the clinical features of the two entities often overlap, and a clear pathological diagnosis is important for subsequent management, especially for the management of PCNSL. Stereotactic biopsy is recognized as one of the minimally invasive alternatives for evaluating the involvement of the CNS. However, in the case of limited tissue materials, the differential diagnosis between the two entities is still difficult. In addition, some patients are too ill to tolerate a needle biopsy. Therefore, combining imaging, histopathology and laboratory examinations is essential in order to make a clear diagnosis as soon as possible. The present study reviews the progress of comparative research on both imaging and laboratory tests based on the pathophysiological changes of the two entities, and proposes an integrative and optimized diagnostic process, with the purpose of building a better understanding for neurologists, hematologists, radiologists and pathologists.

Use of 18F-FDG-PET/CT in differential diagnosis of primary central nervous system lymphoma and high-grade gliomas: A meta-analysis

Background

Primary central nervous system lymphoma (PCNSL) and high-grade glioma (HGG) appear similar under imaging. However, since the two tumors vary in their treatment methods, their differential diagnosis is crucial. The use of 18F-fluorodeoxyglucose positron emission tomography computed tomography (18F-FDG-PET/CT) imaging to effectively distinguish between the two tumors is not clear; therefore, a meta-analysis was carried out to determine its effectiveness.

Materials and methods

The databases PubMed, EMBASE, Cochrane, Web of Science, China National Knowledge Infrastructure (CNKI), Wanfang, China Science, and Technology Journal Database (CQVIP) were exhaustively searched using stringent inclusion and exclusion criteria to select high-quality literature. The Quality Assessment Tool for Diagnostic Accuracy Studies (QUADAS-2) was used for the qualitative assessment of the included literature. The bivariate effect model was used to combine statistics such as sensitivity (SEN) and specificity (SPE), positive likelihood ratio (PLR), negative likelihood ratio (NLR), and diagnostic odds ratio (DOR) [95% confidence intervals (CI)], plot summary receiver operating characteristic (SROC) curve, and calculate the area under the curve (AUC) value. Sensitivity analysis was used to evaluate the stability of the results, and Deek's test was used to assess publication bias. Meta-regression and subgroup analysis was used to determine the sources of heterogeneity.

Results

A total of nine studies were included in this study. For differential diagnosis of PCNSL and HGG, the combined SEN was 0.91 (95% CI: 0.80–0.96; I² = 46.73%), combined SPE was 0.88 (95% CI: 0.82–0.93; I² = 56.30%), the combined PLR was 7.83 (95% CI: 4.96–12.37; I² = 15.57%), combined NLR was 0.10 (95% CI: 0.05–0.23; I² = 31.99%), combined DOR was 77.36 (95% CI: 32.74–182.77; I² = 70.70%). The AUC of SROC was 0.95 (95% CI: 0.93–0.97). No publication bias was found and the sample size and different parameters were the primary reason for heterogeneity.

Conclusion

The 18F-FDG-PET/CT imaging technique has a high diagnostic accuracy in the differential diagnosis of PNCSL and HGG. Patients suspected to have the above two tumors are suggested to be examined by 18F-FDG-PET / CT to help in the clinical distinction and further treatment modalities.

Convolutional Neural Network Techniques for Brain Tumor Classification (from 2015 to 2022): Review, Challenges, and Future Perspectives

Convolutional neural networks (CNNs) constitute a widely used deep learning approach that has frequently been applied to the problem of brain tumor diagnosis. Such techniques still face some critical challenges in moving towards clinic application. The main objective of this work is to present a comprehensive review of studies using CNN architectures to classify brain tumors using MR images with the aim of identifying useful strategies for and possible impediments in the development of this technology. Relevant articles were identified using a predefined, systematic procedure. For each article, data were extracted regarding training data, target problems, the network architecture, validation methods, and the reported quantitative performance criteria. The clinical relevance of the studies was then evaluated to identify limitations by considering the merits of convolutional neural networks and the remaining challenges that need to be solved to promote the clinical application and development of CNN algorithms. Finally, possible directions for future research are discussed for researchers in the biomedical and machine learning communities. A total of 83 studies were identified and reviewed. They differed in terms of the precise classification problem targeted and the strategies used to construct and train the chosen CNN. Consequently, the reported performance varied widely, with accuracies of 91.63–100% in differentiating meningiomas, gliomas, and pituitary tumors (26 articles) and of 60.0–99.46% in distinguishing low-grade from high-grade gliomas (13 articles). The review provides a survey of the state of the art in CNN-based deep learning methods for brain tumor classification. Many networks demonstrated good performance, and it is not evident that any specific methodological choice greatly outperforms the alternatives, especially given the inconsistencies in the reporting of validation methods, performance metrics, and training data encountered. Few studies have focused on clinical usability.

The Role of [68Ga]Ga-DOTA-SSTR PET Radiotracers in Brain Tumors: A Systematic Review of the Literature and Ongoing Clinical Trials

Simple Summary

[⁶⁸Ga]Ga-DOTA-SSTR PET imaging has recently been introduced in the management of patients with brain tumors, mostly meningiomas and pituitary adenomas or carcinomas. The current literature demonstrated the superior diagnostic accuracy of this imaging modality, especially for lesions difficult to be detected or characterized on conventional imaging protocols, such as skull base or transosseous meningiomas. [⁶⁸Ga]Ga-DOTA-SSTR PET tracers also seem to provide superior volume contouring for radiotherapy planning and may also be used to evaluate the tumor’s overexpression of somatostatin receptors for devising patient-tailored peptide receptor radionuclide therapy. In this review, we comprehensively analyzed the current literature discussing the implementation of [⁶⁸Ga]Ga-DOTA-SSTR PET imaging in brain tumors, further presenting ongoing clinical trials and suggesting potential future applications.

Abstract

Background: The development of [⁶⁸Ga]Ga-DOTA-SSTR PET tracers has garnered interest in neuro-oncology, to increase accuracy in diagnostic, radiation planning, and neurotheranostics protocols. We systematically reviewed the literature on the current uses of [⁶⁸Ga]Ga-DOTA-SSTR PET in brain tumors. Methods: PubMed, Scopus, Web of Science, and Cochrane were searched in accordance with the PRISMA guidelines to include published studies and ongoing trials utilizing [⁶⁸Ga]Ga-DOTA-SSTR PET in patients with brain tumors. Results: We included 63 published studies comprising 1030 patients with 1277 lesions, and 4 ongoing trials. [⁶⁸Ga]Ga-DOTA-SSTR PET was mostly used for diagnostic purposes (62.5%), followed by treatment planning (32.7%), and neurotheranostics (4.8%). Most lesions were meningiomas (93.6%), followed by pituitary adenomas (2.8%), and the DOTATOC tracer (53.2%) was used more frequently than DOTATATE (39.1%) and DOTANOC (5.7%), except for diagnostic purposes (DOTATATE 51.1%). [⁶⁸Ga]Ga-DOTA-SSTR PET studies were mostly required to confirm the diagnosis of meningiomas (owing to their high SSTR2 expression and tracer uptake) or evaluate their extent of bone invasion, and improve volume contouring for better radiotherapy planning. Some studies reported the uncommon occurrence of SSTR2-positive brain pathology challenging the diagnostic accuracy of [⁶⁸Ga]Ga-DOTA-SSTR PET for meningiomas. Pre-treatment assessment of tracer uptake rates has been used to confirm patient eligibility (high somatostatin receptor-2 expression) for peptide receptor radionuclide therapy (PRRT) (i.e., neurotheranostics) for recurrent meningiomas and pituitary carcinomas. Conclusion: [⁶⁸Ga]Ga-DOTA-SSTR PET studies may revolutionize the routine neuro-oncology practice, especially in meningiomas, by improving diagnostic accuracy, delineation of radiotherapy targets, and patient eligibility for radionuclide therapies.