Matthias W. Wagner

Khashayar Namdar, Matthias W. Wagner, Birgit B. Ertl-Wagner et al. · utoronto

Background: As an important branch of machine learning pipelines in medical imaging, radiomics faces two major challenges namely reproducibility and accessibility. In this work, we introduce open-radiomics, a set of radiomics datasets along with a comprehensive radiomics pipeline based on our proposed technical protocol to improve the reproducibility of the results. Methods: We curated large-scale radiomics datasets based on three open-source datasets; BraTS 2020 for high-grade glioma (HGG) versus low-grade glioma (LGG) classification and survival analysis, BraTS 2023 for O6-methylguanine-DNA methyltransferase classification, and non-small cell lung cancer survival analysis from the Cancer Imaging Archive. Using BraTS 2020 Magnetic Resonance Imaging (MRI) dataset, we applied our protocol to 369 brain tumor patients (76 LGG, 293 HGG). Leveraging PyRadiomics for LGG vs. HGG classification, we generated 288 datasets from 4 MRI sequences, 3 binWidths, 6 normalization methods, and 4 tumor subregions. Random Forest classifiers were trained and validated (60%,20%,20%) across 100 different data splits (28,800 test results), evaluating Area Under the Receiver Operating Characteristic Curve (AUROC). Results: Unlike binWidth and image normalization, tumor subregion and imaging sequence significantly affected performance of the models. T1 contrast-enhanced sequence and the union of Necrotic and the non-enhancing tumor core subregions resulted in the highest AUROCs (average test AUROC 0.951, 95% confidence interval of (0.949, 0.952)). Although several settings and data splits (28 out of 28800) yielded test AUROC of 1, they were irreproducible. Conclusion: Our experiments demonstrate the sources of variability in radiomics pipelines (e.g., tumor subregion) can have a significant impact on the results, which may lead to superficial perfect performances that are irreproducible.

Khashayar Namdar, Matthias W. Wagner, Kareem Kudus et al. · utoronto

Background and Purpose: Pediatric low-grade glioma (pLGG) is the most common type of brain tumor in children, and identification of molecular markers for pLGG is crucial for successful treatment planning. Convolutional Neural Network (CNN) models for pLGG subtype identification rely on tumor segmentation. We hypothesize tumor segmentations are suboptimal and thus, we propose to augment the CNN models using tumor location probability in MRI data. Materials and Methods: Our REB-approved retrospective study included MRI Fluid-Attenuated Inversion Recovery (FLAIR) sequences of 143 BRAF fused and 71 BRAF V600E mutated tumors. Tumor segmentations (regions of interest (ROIs)) were provided by a pediatric neuroradiology fellow and verified by a senior pediatric neuroradiologist. In each experiment, we randomly split the data into development and test with an 80/20 ratio. We combined the 3D binary ROI masks for each class in the development dataset to derive the probability density functions (PDF) of tumor location, and developed three pipelines: location-based, CNN-based, and hybrid. Results: We repeated the experiment with different model initializations and data splits 100 times and calculated the Area Under Receiver Operating Characteristic Curve (AUC). The location-based classifier achieved an AUC of 77.90, 95% confidence interval (CI) (76.76, 79.03). CNN-based classifiers achieved AUC of 86.11, CI (84.96, 87.25), while the tumor-location-guided CNNs outperformed the formers with an average AUC of 88.64 CI (87.57, 89.72), which was statistically significant (Student's t-test p-value 0.0018). Conclusion: We achieved statistically significant improvements by incorporating tumor location into the CNN models. Our results suggest that manually segmented ROIs may not be optimal.

Jay J. Yoo, Khashayar Namdar, Matthias W. Wagner et al. · utoronto

Segmentation of regions of interest (ROIs) for identifying abnormalities is a leading problem in medical imaging. Using machine learning for this problem generally requires manually annotated ground-truth segmentations, demanding extensive time and resources from radiologists. This work presents a weakly supervised approach that utilizes binary image-level labels, which are much simpler to acquire, to effectively segment anomalies in 2D magnetic resonance images without ground truth annotations. We train a generative adversarial network (GAN) that converts cancerous images to healthy variants, which are used along with localization seeds as priors to generate improved weakly supervised segmentations. The non-cancerous variants can also be used to evaluate the segmentations in a weakly supervised fashion, which allows for the most effective segmentations to be identified and then applied to downstream clinical classification tasks. On the Multimodal Brain Tumor Segmentation (BraTS) 2020 dataset, our proposed method generates and identifies segmentations that achieve test Dice coefficients of 83.91%. Using these segmentations for pathology classification results with a test AUC of 93.32% which is comparable to the test AUC of 95.80% achieved when using true segmentations.

Sara Ketabi, Matthias W. Wagner, Birgit Betina Ertl-Wagner et al.

In spite of the strong performance of machine learning (ML) models in radiology, they have not been widely accepted by radiologists, limiting clinical integration. A key reason is the lack of explainability, which ensures that model predictions are understandable and verifiable by clinicians. Several methods and tools have been proposed to improve explainability, but most reflect developers' perspectives and lack systematic clinical validation. In this work, we gathered insights from radiologists with varying experience and specialties into explainable ML requirements through a structured questionnaire. They also highlighted key clinical tasks where ML could be most beneficial and how it might be deployed. Based on their input, we propose guidelines for designing and developing explainable ML models in radiology. These guidelines can help researchers develop clinically useful models, facilitating integration into radiology practice as a supportive tool.

Khashayar Namdar, Matthias W. Wagner, Cynthia Hawkins et al.

Pediatric Low-Grade Neuroepithelial Tumors (PLGNT) are the most common pediatric cancer type, accounting for 40% of brain tumors in children, and identifying PLGNT molecular subtype is crucial for treatment planning. However, the gold standard to determine the PLGNT subtype is biopsy, which can be impractical or dangerous for patients. This research improves the performance of Convolutional Neural Networks (CNNs) in classifying PLGNT subtypes through MRI scans by introducing a loss function that specifically improves the model's Area Under the Receiver Operating Characteristic (ROC) Curve (AUROC), offering a non-invasive diagnostic alternative. In this study, a retrospective dataset of 339 children with PLGNT (143 BRAF fusion, 71 with BRAF V600E mutation, and 125 non-BRAF) was curated. We employed a CNN model with Monte Carlo random data splitting. The baseline model was trained using binary cross entropy (BCE), and achieved an AUROC of 86.11% for differentiating BRAF fusion and BRAF V600E mutations, which was improved to 87.71% using our proposed AUROC loss function (p-value 0.045). With multiclass classification, the AUROC improved from 74.42% to 76. 59% (p-value 0.0016).

Sara Ketabi, Matthias W. Wagner, Cynthia Hawkins et al.

Despite the promising performance of convolutional neural networks (CNNs) in brain tumor diagnosis from magnetic resonance imaging (MRI), their integration into the clinical workflow has been limited. That is mainly due to the fact that the features contributing to a model's prediction are unclear to radiologists and hence, clinically irrelevant, i.e., lack of explainability. As the invaluable sources of radiologists' knowledge and expertise, radiology reports can be integrated with MRI in a contrastive learning (CL) framework, enabling learning from image-report associations, to improve CNN explainability. In this work, we train a multimodal CL architecture on 3D brain MRI scans and radiology reports to learn informative MRI representations. Furthermore, we integrate tumor location, salient to several brain tumor analysis tasks, into this framework to improve its generalizability. We then apply the learnt image representations to improve explainability and performance of genetic marker classification of pediatric Low-grade Glioma, the most prevalent brain tumor in children, as a downstream task. Our results indicate a Dice score of 31.1% between the model's attention maps and manual tumor segmentation (as an explainability measure) with test classification performance of 87.7%, significantly outperforming the baselines. These enhancements can build trust in our model among radiologists, facilitating its integration into clinical practices for more efficient tumor diagnosis.

Partoo Vafaeikia, Matthias W. Wagner, Uri Tabori et al.

Brain tumor segmentation is a critical task for tumor volumetric analyses and AI algorithms. However, it is a time-consuming process and requires neuroradiology expertise. While there has been extensive research focused on optimizing brain tumor segmentation in the adult population, studies on AI guided pediatric tumor segmentation are scarce. Furthermore, MRI signal characteristics of pediatric and adult brain tumors differ, necessitating the development of segmentation algorithms specifically designed for pediatric brain tumors. We developed a segmentation model trained on magnetic resonance imaging (MRI) of pediatric patients with low-grade gliomas (pLGGs) from The Hospital for Sick Children (Toronto, Ontario, Canada). The proposed model utilizes deep Multitask Learning (dMTL) by adding tumor's genetic alteration classifier as an auxiliary task to the main network, ultimately improving the accuracy of the segmentation results.