Deep Learning Can Differentiate IDH-Mutant from IDH-Wild GBM

Identification of IDH and TERTp mutations using dynamic susceptibility contrast MRI with deep learning in 162 gliomas

PURPOSE

Isocitrate dehydrogenase (IDH) and telomerase reverse transcriptase gene promoter (TERTp) mutations play crucial roles in glioma biology. Such genetic information is typically obtained invasively from excised tumor tissue; however, these mutations need to be identified preoperatively for better treatment planning. The relative cerebral blood volume (rCBV) information derived from dynamic susceptibility contrast MRI (DSC-MRI) has been demonstrated to correlate with tumor vascularity, functionality, and biology, and might provide some information about the genetic alterations in gliomas before surgery. Therefore, this study aims to predict IDH and TERTp mutational subgroups in gliomas using deep learning applied to rCBV images.

METHOD

After the generation of rCBV images from DSC-MRI data, classical machine learning algorithms were applied to the features obtained from the segmented tumor volumes to classify IDH and TERTp mutation subgroups. Furthermore, pre-trained convolutional neural networks (CNNs) and CNNs enhanced with attention gates were trained using rCBV images or a combination of rCBV and anatomical images to classify the mutational subgroups.

RESULTS

The best accuracies obtained with classical machine learning algorithms were 83 %, 68 %, and 76 % for the identification of IDH mutational, TERTp mutational, and TERTp-only subgroups, respectively. On the other hand, the best-performing CNN model achieved 88 % accuracy (86 % sensitivity, 91 % specificity) for the IDH-mutational subgroups, 70 % accuracy (73 % sensitivity and 67 % specificity) for the TERTp-mutational subgroups, and 84 % accuracy (86 % sensitivity, 81 % specificity) for the TERTp-only subgroup using attention gates.

CONCLUSIONS

DSC-MRI can be utilized to noninvasively classify IDH- and TERTp-based molecular subgroups of gliomas, facilitating preoperative identification of these genetic alterations.

Machine Learning Algorithm: Texture Analysis in CNO and Application in Distinguishing CNO and Bone Marrow Growth-Related Changes on Whole-Body MRI

OBJECTIVE

The purpose of this study is to analyze the texture characteristics of chronic non-bacterial osteomyelitis (CNO) bone lesions, identified as areas of altered signal intensity on short tau inversion recovery (STIR) sequences, and to distinguish them from bone marrow growth-related changes through Machine Learning (ML) and Deep Learning (DL) analysis.

MATERIALS AND METHODS

We included a group of 66 patients with confirmed diagnosis of CNO and a group of 28 patients with suspected extra-skeletal systemic disease. All examinations were performed on a 1.5 T MRI scanner. Using the opensource 3D Slicer software version 4.10.2, the ROIs on CNO lesions and on the red bone marrow were sampled. Texture analysis (TA) was carried out using Pyradiomics. We applied an optimization search grid algorithm on nine classic ML classifiers and a Deep Learning (DL) Neural Network (NN). The model's performance was evaluated using Accuracy (ACC), AUC-ROC curves, F1-score, Positive Predictive Value (PPV), Mean Absolute Error (MAE) and Root-Mean-Square Error (RMSE). Furthermore, we used Shapley additive explanations to gain insight into the behavior of the prediction model.

RESULTS

Most predictive characteristics were selected by Boruta algorithm for each combination of ROI sequences for the characterization and classification of the two types of signal hyperintensity. The overall best classification result was obtained by the NN with ACC = 0.91, AUC = 0.93 with 95% CI 0.91-0.94, F1-score = 0.94 and PPV = 93.8%. Between classic ML methods, ensemble learners showed high model performance; specifically, the best-performing classifier was the Stack (ST) with ACC = 0.85, AUC = 0.81 with 95% CI 0.8-0.84, F1-score = 0.9, PPV = 90%.

CONCLUSIONS

Our results show the potential of ML methods in discerning edema-like lesions, in particular by distinguishing CNO lesions from hematopoietic bone marrow changes in a pediatric population. The Neural Network showed the overall best results, while a Stacking classifier, based on Gradient Boosting and Random Forest as principal estimators and Logistic Regressor as final estimator, achieved the best results between the other ML methods.

Comparison of MRI radiomics-based machine learning survival models in predicting prognosis of glioblastoma multiforme

Objective

To compare the performance of radiomics-based machine learning survival models in predicting the prognosis of glioblastoma multiforme (GBM) patients.

Methods

131 GBM patients were included in our study. The traditional Cox proportional-hazards (CoxPH) model and four machine learning models (SurvivalTree, Random survival forest (RSF), DeepSurv, DeepHit) were constructed, and the performance of the five models was evaluated using the C-index.

Results

After the screening, 1792 radiomics features were obtained. Seven radiomics features with the strongest relationship with prognosis were obtained following the application of the least absolute shrinkage and selection operator (LASSO) regression. The CoxPH model demonstrated that age (HR = 1.576, p = 0.037), Karnofsky performance status (KPS) score (HR = 1.890, p = 0.006), radiomics risk score (HR = 3.497, p = 0.001), and radiomics risk level (HR = 1.572, p = 0.043) were associated with poorer prognosis. The DeepSurv model performed the best among the five models, obtaining C-index of 0.882 and 0.732 for the training and test set, respectively. The performances of the other four models were lower: CoxPH (0.663 training set / 0.635 test set), SurvivalTree (0.702/0.655), RSF (0.735/0.667), DeepHit (0.608/0.560).

Conclusion

This study confirmed the superior performance of deep learning algorithms based on radiomics relative to the traditional method in predicting the overall survival of GBM patients; specifically, the DeepSurv model showed the best predictive ability.

MRI-Based Deep Learning Method for Classification of IDH Mutation Status

Isocitrate dehydrogenase (IDH) mutation status has emerged as an important prognostic marker in gliomas. This study sought to develop deep learning networks for non-invasive IDH classification using T2w MR images while comparing their performance to a multi-contrast network. Methods: Multi-contrast brain tumor MRI and genomic data were obtained from The Cancer Imaging Archive (TCIA) and The Erasmus Glioma Database (EGD). Two separate 2D networks were developed using nnU-Net, a T2w-image-only network (T2-net) and a multi-contrast network (MC-net). Each network was separately trained using TCIA (227 subjects) or TCIA + EGD data (683 subjects combined). The networks were trained to classify IDH mutation status and implement single-label tumor segmentation simultaneously. The trained networks were tested on over 1100 held-out datasets including 360 cases from UT Southwestern Medical Center, 136 cases from New York University, 175 cases from the University of Wisconsin-Madison, 456 cases from EGD (for the TCIA-trained network), and 495 cases from the University of California, San Francisco public database. A receiver operating characteristic curve (ROC) was drawn to calculate the AUC value to determine classifier performance. Results: T2-net trained on TCIA and TCIA + EGD datasets achieved an overall accuracy of 85.4% and 87.6% with AUCs of 0.86 and 0.89, respectively. MC-net trained on TCIA and TCIA + EGD datasets achieved an overall accuracy of 91.0% and 92.8% with AUCs of 0.94 and 0.96, respectively. We developed reliable, high-performing deep learning algorithms for IDH classification using both a T2-image-only and a multi-contrast approach. The networks were tested on more than 1100 subjects from diverse databases, making this the largest study on image-based IDH classification to date.

Functional MRI in Neuro-Oncology: State of the Art and Future Directions

Since its discovery in the early 1990s, functional MRI (fMRI) has been used to study human brain function. One well-established application of fMRI in the clinical setting is the neurosurgical planning of patients with brain tumors near eloquent cortical areas. Clinical fMRI aims to preoperatively identify eloquent cortices that serve essential functions in daily life, such as hand movement and language. The primary goal of neurosurgery is to maximize tumor resection while sparing eloquent cortices adjacent to the tumor. When a lesion presents in the vicinity of an eloquent cortex, surgeons may use fMRI to plan their best surgical approach by determining the proximity of the lesion to regions of activation, providing guidance for awake brain surgery and intraoperative brain mapping. The acquisition of fMRI requires patient preparation prior to imaging, determination of functional paradigms, monitoring of patient performance, and both processing and analysis of images. Interpretation of fMRI maps requires a strong understanding of functional neuroanatomy and familiarity with the technical limitations frequently present in brain tumor imaging, including neurovascular uncoupling, patient compliance, and data analysis. This review discusses clinical fMRI in neuro-oncology, relevant ongoing research topics, and prospective future developments in this exciting discipline.

Deep learning in precision medicine and focus on glioma

Deep learning (DL) has been successfully applied to different fields for a range of tasks. In medicine, DL methods have been also used to improve the efficiency of disease diagnosis. In this review, we first summarize the history of the development of artificial intelligence models, demonstrate the features of the subtypes of machine learning and different DL networks, and then explore their application in the different fields of precision medicine, such as cardiology, gastroenterology, ophthalmology, dermatology, and oncology. By digging more information and extracting multilevel features from medical data, we found that DL helps doctors assess diseases automatically and monitor patients' physical health. In gliomas, research regarding application prospect of DL was mainly shown through magnetic resonance imaging and then by pathological slides. However, multi-omics data, such as whole exome sequence, RNA sequence, proteomics, and epigenomics, have not been covered thus far. In general, the quality and quantity of DL datasets still need further improvements, and more fruitful multi-omics characteristics will bring more comprehensive and accurate diagnosis in precision medicine and glioma.

MRI-based deep learning techniques for the prediction of isocitrate dehydrogenase and 1p/19q status in grade 2-4 adult gliomas

Molecular biomarkers are becoming increasingly important in the classification of intracranial gliomas. While tissue sampling remains the gold standard, there is growing interest in the use of deep learning (DL) techniques to predict these markers. This narrative review with a systematic approach identifies and synthesises the current published data on DL techniques using conventional MRI sequences for predicting isocitrate dehydrogenase (IDH) and 1p/19q-codeletion status in World Health Organisation grade 2-4 gliomas. Three databases were searched for relevant studies. In all, 13 studies met the inclusion criteria after exclusions. Key results, limitations and discrepancies between studies were synthesised. High accuracy has been reported in some studies, but the existing literature has several limitations, including generally small cohort sizes, a paucity of studies with independent testing cohorts and a lack of studies assessing IDH and 1p/19q together. While DL shows promise as a non-invasive means of predicting glioma genotype, addressing these limitations in future research will be important for facilitating clinical translation.

One Step Forward-The Current Role of Artificial Intelligence in Glioblastoma Imaging

Artificial intelligence (AI) is rapidly integrating into diagnostic methods across many branches of medicine. Significant progress has been made in tumor assessment using AI algorithms, and research is underway on how image manipulation can provide information with diagnostic, prognostic and treatment impacts. Glioblastoma (GB) remains the most common primary malignant brain tumor, with a median survival of 15 months. This paper presents literature data on GB imaging and the contribution of AI to the characterization and tracking of GB, as well as recurrence. Furthermore, from an imaging point of view, the differential diagnosis of these tumors can be problematic. How can an AI algorithm help with differential diagnosis? The integration of clinical, radiomics and molecular markers via AI holds great potential as a tool for enhancing patient outcomes by distinguishing brain tumors from mimicking lesions, classifying and grading tumors, and evaluating them before and after treatment. Additionally, AI can aid in differentiating between tumor recurrence and post-treatment alterations, which can be challenging with conventional imaging methods. Overall, the integration of AI into GB imaging has the potential to significantly improve patient outcomes by enabling more accurate diagnosis, precise treatment planning and better monitoring of treatment response.

The promise of metabolic imaging in diffuse midline glioma

Recent insights into histopathological and molecular subgroups of glioma have revolutionized the field of neuro-oncology by refining diagnostic categories. An emblematic example in pediatric neuro-oncology is the newly defined diffuse midline glioma (DMG), H3 K27-altered. DMG represents a rare tumor with a dismal prognosis. The diagnosis of DMG is largely based on clinical presentation and characteristic features on conventional magnetic resonance imaging (MRI), with biopsy limited by its delicate neuroanatomic location. Standard MRI remains limited in its ability to characterize tumor biology. Advanced MRI and positron emission tomography (PET) imaging offer additional value as they enable non-invasive evaluation of molecular and metabolic features of brain tumors. These techniques have been widely used for tumor detection, metabolic characterization and treatment response monitoring of brain tumors. However, their role in the realm of pediatric DMG is nascent. By summarizing DMG metabolic pathways in conjunction with their imaging surrogates, we aim to elucidate the untapped potential of such imaging techniques in this devastating disease.

3D CT-Inclusive Deep-Learning Model to Predict Mortality, ICU Admittance, and Intubation in COVID-19 Patients

Chest CT is a useful initial exam in patients with coronavirus disease 2019 (COVID-19) for assessing lung damage. AI-powered predictive models could be useful to better allocate resources in the midst of the pandemic. Our aim was to build a deep-learning (DL) model for COVID-19 outcome prediction inclusive of 3D chest CT images acquired at hospital admission. This retrospective multicentric study included 1051 patients (mean age 69, SD = 15) who presented to the emergency department of three different institutions between 20th March 2020 and 20th January 2021 with COVID-19 confirmed by real-time reverse transcriptase polymerase chain reaction (RT-PCR). Chest CT at hospital admission were evaluated by a 3D residual neural network algorithm. Training, internal validation, and external validation groups included 608, 153, and 290 patients, respectively. Images, clinical, and laboratory data were fed into different customizations of a dense neural network to choose the best performing architecture for the prediction of mortality, intubation, and intensive care unit (ICU) admission. The AI model tested on CT and clinical features displayed accuracy, sensitivity, specificity, and ROC-AUC, respectively, of 91.7%, 90.5%, 92.4%, and 95% for the prediction of patient's mortality; 91.3%, 91.5%, 89.8%, and 95% for intubation; and 89.6%, 90.2%, 86.5%, and 94% for ICU admission (internal validation) in the testing cohort. The performance was lower in the validation cohort for mortality (71.7%, 55.6%, 74.8%, 72%), intubation (72.6%, 74.7%, 45.7%, 64%), and ICU admission (74.7%, 77%, 46%, 70%) prediction. The addition of the available laboratory data led to an increase in sensitivity for patient's mortality (66%) and specificity for intubation and ICU admission (50%, 52%, respectively), while the other metrics maintained similar performance results. We present a deep-learning model to predict mortality, ICU admittance, and intubation in COVID-19 patients. KEY POINTS: • 3D CT-based deep learning model predicted the internal validation set with high accuracy, sensibility and specificity (> 90%) mortality, ICU admittance, and intubation in COVID-19 patients. • The model slightly increased prediction results when laboratory data were added to the analysis, despite data imbalance. However, the model accuracy dropped when CT images were not considered in the analysis, implying an important role of CT in predicting outcomes.

Magnetic Resonance Imaging of Primary Adult Brain Tumors: State of the Art and Future Perspectives

MRI is undoubtedly the cornerstone of brain tumor imaging, playing a key role in all phases of patient management, starting from diagnosis, through therapy planning, to treatment response and/or recurrence assessment. Currently, neuroimaging can describe morphologic and non-morphologic (functional, hemodynamic, metabolic, cellular, microstructural, and sometimes even genetic) characteristics of brain tumors, greatly contributing to diagnosis and follow-up. Knowing the technical aspects, strength and limits of each MR technique is crucial to correctly interpret MR brain studies and to address clinicians to the best treatment strategy. This article aimed to provide an overview of neuroimaging in the assessment of adult primary brain tumors. We started from the basilar role of conventional/morphological MR sequences, then analyzed, one by one, the non-morphological techniques, and finally highlighted future perspectives, such as radiomics and artificial intelligence.

Combining Multi-Shell Diffusion with Conventional MRI Improves Molecular Diagnosis of Diffuse Gliomas with Deep Learning

The WHO classification since 2016 confirms the importance of integrating molecular diagnosis for prognosis and treatment decisions of adult-type diffuse gliomas. This motivates the development of non-invasive diagnostic methods, in particular MRI, to predict molecular subtypes of gliomas before surgery. At present, this development has been focused on deep-learning (DL)-based predictive models, mainly with conventional MRI (cMRI), despite recent studies suggesting multi-shell diffusion MRI (dMRI) offers complementary information to cMRI for molecular subtyping. The aim of this work is to evaluate the potential benefit of combining cMRI and multi-shell dMRI in DL-based models. A model implemented with deep residual neural networks was chosen as an illustrative example. Using a dataset of 146 patients with gliomas (from grade 2 to 4), the model was trained and evaluated, with nested cross-validation, on pre-operative cMRI, multi-shell dMRI, and a combination of the two for the following classification tasks: (i) IDH-mutation; (ii) 1p/19q-codeletion; and (iii) three molecular subtypes according to WHO 2021. The results from a subset of 100 patients with lower grades gliomas (2 and 3 according to WHO 2016) demonstrated that combining cMRI and multi-shell dMRI enabled the best performance in predicting IDH mutation and 1p/19q codeletion, achieving an accuracy of 75 ± 9% in predicting the IDH-mutation status, higher than using cMRI and multi-shell dMRI separately (both 70 ± 7%). Similar findings were observed for predicting the 1p/19q-codeletion status, with the accuracy from combining cMRI and multi-shell dMRI (72 ± 4%) higher than from each modality used alone (cMRI: 65 ± 6%; multi-shell dMRI: 66 ± 9%). These findings remain when we considered all 146 patients for predicting the IDH status (combined: 81 ± 5% accuracy; cMRI: 74 ± 5%; multi-shell dMRI: 73 ± 6%) and for the diagnosis of the three molecular subtypes according to WHO 2021 (combined: 60 ± 5%; cMRI: 57 ± 8%; multi-shell dMRI: 56 ± 7%). Together, these findings suggest that combining cMRI and multi-shell dMRI can offer higher accuracy than using each modality alone for predicting the IDH and 1p/19q status and in diagnosing the three molecular subtypes with DL-based models.

Predicting methylation class from diffusely infiltrating adult gliomas using multimodality MRI data

Background

Radiogenomic studies of adult-type diffuse gliomas have used magnetic resonance imaging (MRI) data to infer tumor attributes, including abnormalities such as IDH-mutation status and 1p19q deletion. This approach is effective but does not generalize to tumor types that lack highly recurrent alterations. Tumors have intrinsic DNA methylation patterns and can be grouped into stable methylation classes even when lacking recurrent mutations or copy number changes. The purpose of this study was to prove the principle that a tumor's DNA-methylation class could be used as a predictive feature for radiogenomic modeling.

Methods

Using a custom DNA methylation-based classification model, molecular classes were assigned to diffuse gliomas in The Cancer Genome Atlas (TCGA) dataset. We then constructed and validated machine learning models to predict a tumor's methylation family or subclass from matched multisequence MRI data using either extracted radiomic features or directly from MRI images.

Results

For models using extracted radiomic features, we demonstrated top accuracies above 90% for predicting IDH-glioma and GBM-IDHwt methylation families, IDH-mutant tumor methylation subclasses, or GBM-IDHwt molecular subclasses. Classification models utilizing MRI images directly demonstrated average accuracies of 80.6% for predicting methylation families, compared to 87.2% and 89.0% for differentiating IDH-mutated astrocytomas from oligodendrogliomas and glioblastoma molecular subclasses, respectively.

Conclusions

These findings demonstrate that MRI-based machine learning models can effectively predict the methylation class of brain tumors. Given appropriate datasets, this approach could generalize to most brain tumor types, expanding the number and types of tumors that could be used to develop radiomic or radiogenomic models.

Synthetic Post-Contrast Imaging through Artificial Intelligence: Clinical Applications of Virtual and Augmented Contrast Media

Contrast media are widely diffused in biomedical imaging, due to their relevance in the diagnosis of numerous disorders. However, the risk of adverse reactions, the concern of potential damage to sensitive organs, and the recently described brain deposition of gadolinium salts, limit the use of contrast media in clinical practice. In recent years, the application of artificial intelligence (AI) techniques to biomedical imaging has led to the development of 'virtual' and 'augmented' contrasts. The idea behind these applications is to generate synthetic post-contrast images through AI computational modeling starting from the information available on other images acquired during the same scan. In these AI models, non-contrast images (virtual contrast) or low-dose post-contrast images (augmented contrast) are used as input data to generate synthetic post-contrast images, which are often undistinguishable from the native ones. In this review, we discuss the most recent advances of AI applications to biomedical imaging relative to synthetic contrast media.

Predicting IDH subtype of grade 4 astrocytoma and glioblastoma from tumor radiomic patterns extracted from multiparametric magnetic resonance images using a machine learning approach

Background and purpose

Semantic imaging features have been used for molecular subclassification of high-grade gliomas. Radiomics-based prediction of molecular subgroups has the potential to strategize and individualize therapy. Using MRI texture features, we propose to distinguish between IDH wild type and IDH mutant type high grade gliomas.

Methods

Between 2013 and 2020, 100 patients were retrospectively analyzed for the radiomics study. Immunohistochemistry of the pathological specimen was used to initially identify patients for the IDH mutant/wild phenotype and was then confirmed by Sanger’s sequencing. Image texture analysis was performed on contrast-enhanced T1 (T1C) and T2 weighted (T2W) MR images. Manual segmentation was performed on MR image slices followed by single-slice multiple sampling image augmentation. Both whole tumor multislice segmentation and single-slice multiple sampling approaches were used to arrive at the best model. Radiomic features were extracted, which included first-order features, second-order (GLCM—Grey level co-occurrence matrix), and shape features. Feature enrichment was done using LASSO (Least Absolute Shrinkage and Selection Operator) regression, followed by radiomic classification using Support Vector Machine (SVM) and a 10-fold cross-validation strategy for model development. The area under the Receiver Operator Characteristic (ROC) curve and predictive accuracy were used as diagnostic metrics to evaluate the model to classify IDH mutant and wild-type subgroups.

Results

Multislice analysis resulted in a better model compared to the single-slice multiple-sampling approach. A total of 164 MR-based texture features were extracted, out of which LASSO regression identified 14 distinctive GLCM features for the endpoint, which were used for further model development. The best model was achieved by using combined T1C and T2W MR images using a Quadratic Support Vector Machine Classifier and a 10-fold internal cross-validation approach, which demonstrated a predictive accuracy of 89% with an AUC of 0.89 for each IDH mutant and IDH wild subgroup.

Conclusion

A machine learning classifier of radiomic features extracted from multiparametric MRI images (T1C and T2w) provides important diagnostic information for the non-invasive prediction of the IDH mutant or wild-type phenotype of high-grade gliomas and may have potential use in either escalating or de-escalating adjuvant therapy for gliomas or for using targeted agents in the future.

Swin Transformer Improves the IDH Mutation Status Prediction of Gliomas Free of MRI-Based Tumor Segmentation

Background: Deep learning (DL) could predict isocitrate dehydrogenase (IDH) mutation status from MRIs. Yet, previous work focused on CNNs with refined tumor segmentation. To bridge the gap, this study aimed to evaluate the feasibility of developing a Transformer-based network to predict the IDH mutation status free of refined tumor segmentation. Methods: A total of 493 glioma patients were recruited from two independent institutions for model development (TCIA; N = 259) and external test (AHXZ; N = 234). IDH mutation status was predicted directly from T2 images with a Swin Transformer and conventional ResNet. Furthermore, to investigate the necessity of refined tumor segmentation, seven strategies for the model input image were explored: (i) whole tumor slice; (ii-iii) tumor mask and/or not edema; (iv-vii) tumor bounding box of 0.8, 1.0, 1.2, 1.5 times. Performance comparison was made among the networks of different architectures along with different image input strategies, using area under the curve (AUC) and accuracy (ACC). Finally, to further boost the performance, a hybrid model was built by incorporating the images with clinical features. Results: With the seven proposed input strategies, seven Swin Transformer models and seven ResNet models were built, respectively. Based on the seven Swin Transformer models, an averaged AUC of 0.965 (internal test) and 0.842 (external test) were achieved, outperforming 0.922 and 0.805 resulting from the seven ResNet models, respectively. When a bounding box of 1.0 times was used, Swin Transformer (AUC = 0.868, ACC = 80.7%), achieved the best results against the one that used tumor segmentation (Tumor + Edema, AUC = 0.862, ACC = 78.5%). The hybrid model that integrated age and location features into images yielded improved performance (AUC = 0.878, Accuracy = 82.0%) over the model that used images only. Conclusions: Swin Transformer outperforms the CNN-based ResNet in IDH prediction. Using bounding box input images benefits the DL networks in IDH prediction and makes the IDH prediction free of refined glioma segmentation feasible.

A multimodal domain adaptive segmentation framework for IDH genotype prediction

PURPOSE

The gene mutation status of isocitrate dehydrogenase (IDH) in gliomas leads to a different prognosis. It is challenging to perform automated tumor segmentation and genotype prediction directly using label-deprived multimodal magnetic resonance (MR) images. We propose a novel framework that employs a domain adaptive mechanism to address this issue.

METHODS

Multimodal domain adaptive segmentation (MDAS) framework was proposed to solve the gap issue in cross dataset model transfer. Image translation was used to adaptively align the multimodal data from two domains at the image level, and segmentation consistency loss was proposed to retain more pathological information through semantic constraints. The data distribution between the labeled public dataset and label-free target dataset was learned to achieve better unsupervised segmentation results on the target dataset. Then, the segmented tumor foci were used as a mask to extract the radiomics and deep features. And the subsequent prediction of IDH gene mutation status was conducted by training a random forest classifier. The prediction model does not need any expert segmented labels.

RESULTS

We implemented our method on the public BraTS 2019 dataset and 110 astrocytoma cases of grade II-IV brain tumors from our hospital. We obtained a Dice score of 77.41% for unsupervised tumor segmentation, a genotype prediction accuracy (ACC) of 0.7639 and an area under curve (AUC) of 0.8600. Experimental results demonstrate that our domain adaptive approach outperforms the methods utilizing direct transfer learning. The model using hybrid features gives better results than the model using radiomics or deep features alone.

CONCLUSIONS

Domain adaptation enables the segmentation network to achieve better performance, and the extraction of mixed features at multiple levels on the segmented region of interest ensures effective prediction of the IDH gene mutation status.

The Progression Related Gene RAB42 Affects the Prognosis of Glioblastoma Patients

BACKGROUND

Glioblastoma (GBM) represents the most malignant glioma among astrocytomas and is a lethal form of brain cancer. Many RAB genes are involved in different cancers but RAB42 (Ras-associated binding 42) is seldom studied in GBM. Our study aimed to explore the role of RAB42 expression in the development and prognosis of GBM.

METHODS

All GBM patient data were obtained from The Cancer Genome Atlas (TCGA) and Chinese Glioma Genome Atlas (CGGA) databases. The relevance of RAB42 expression to the clinicopathologic characteristics of GBM patients was analyzed. The overall survival (OS) significance was determined using log-rank. Significantly enriched KEGG pathways were screened using gene set enrichment analysis (GSEA).

RESULTS

High expression of RAB42 was observed in GBM specimens compared with normal samples, which was also verified in cell lines and tissue samples. Elevated RAB42 expression was correlated with higher GBM histological grade. The prognosis of GBM patients with high RAB42 expression was worse than those with lower RAB42. A total of 35 pathways, such as the P53 pathway, were significantly activated in highly RAB42-expressed GBM samples.

CONCLUSIONS

High RAB42 expression is related to the development of GBM, and RAB42 is a probable prognostic marker for GBM.

Validation Study for Non-Invasive Prediction of IDH Mutation Status in Patients with Glioma Using In Vivo 1H-Magnetic Resonance Spectroscopy and Machine Learning

Simple Summary

The enzyme isocitrate dehydrogenase (IDH) affects glioma cell metabolism in multiple ways. Mutation of IDH is not only indicative of the presence of astrocytoma or oligodendroglioma but it also comes with a better prognosis and constitutes a promising therapeutic target. Therefore, determination of IDH mutation status is essential in clinical practice. In most patients, tissue can be obtained by resection or biopsy to determine IDH status histologically. However, in some cases, this is not possible for technical reasons. We recently showed in a small cohort of patients that non-invasive determination of IDH mutation status using proton magnetic resonance spectroscopy (¹H-MRS) at 3.0 Tesla (T) together with machine learning techniques is feasible in a standard clinical setting and with acceptable effort. Here, we demonstrate that our approach showed comparably good results in sensitivity (82.6%) and specificity (72.7%) in a larger validation cohort employing ¹H-MRS at 1.5 T in a retrospective, distinct setting. We concluded that our method works well regardless of the magnetic field strength and scanner used, and thus, may improve patient care.

Abstract

The isocitrate dehydrogenase (IDH) mutation status is an indispensable prerequisite for diagnosis of glioma (astrocytoma and oligodendroglioma) according to the WHO classification of brain tumors 2021 and is a potential therapeutic target. Usually, immunohistochemistry followed by sequencing of tumor tissue is performed for this purpose. In clinical routine, however, non-invasive determination of IDH mutation status is desirable in cases where tumor biopsy is not possible and for monitoring neuro-oncological therapies. In a previous publication, we presented reliable prediction of IDH mutation status employing proton magnetic resonance spectroscopy (¹H-MRS) on a 3.0 Tesla (T) scanner and machine learning in a prospective cohort of 34 glioma patients. Here, we validated this approach in an independent cohort of 67 patients, for which ¹H-MR spectra were acquired at 1.5 T between 2002 and 2007, using the same data analysis approach. Despite different technical conditions, a sensitivity of 82.6% (95% CI, 61.2–95.1%) and a specificity of 72.7% (95% CI, 57.2–85.0%) could be achieved. We concluded that our ¹H-MRS based approach can be established in a routine clinical setting with affordable effort and time, independent of technical conditions employed. Therefore, the method provides a non-invasive tool for determining IDH status that is well-applicable in an everyday clinical setting.

External Validation of a Convolutional Neural Network for IDH Mutation Prediction

Background and Objectives: The IDH (isocitrate dehydrogenase) status represents one of the main prognosis factors for gliomas. However, determining it requires invasive procedures and specialized surgical skills. Medical imaging such as MRI is essential in glioma diagnosis and management. Lately, fields such as Radiomics and Radiogenomics emerged as pertinent prediction tools for extracting molecular information out of medical images. These fields are based on Artificial Intelligence algorithms that require external validation in order to evaluate their general performance. The aim of this study was to provide an external validation for the algorithm formulated by Yoon Choi et al. of IDH status prediction using preoperative common MRI sequences and patient age. Material and Methods: We applied Choi’s IDH status prediction algorithm on T1c, T2 and FLAIR preoperative MRI images of gliomas (grades WHO II-IV) of 21 operated adult patients from the Neurosurgery clinic of the Cluj County Emergency Clinical Hospital (CCECH), Cluj-Napoca Romania. We created a script to automate the testing process with DICOM format MRI sequences as input and IDH predicted status as output. Results: In terms of patient characteristics, the mean age was 48.6 ± 15.6; 57% were female and 43% male; 43% were IDH positive and 57% IDH negative. The proportions of WHO grades were 24%, 14% and 62% for II, III and IV, respectively. The validation test achieved a relative accuracy of 76% with 95% CI of (53%, 92%) and an Area Under the Curve (AUC) through DeLong et al. method of 0.74 with 95% CI of (0.53, 0.91) and a p of 0.021. Sensitivity and Specificity were 0.78 with 95% CI of (0.45, 0.96) and 0.75 with 95% CI of (0.47, 0.91), respectively. Conclusions: Although our results match the external test the author made on The Cancer Imaging Archive (TCIA) online dataset, performance of the algorithm on external data is still not high enough for clinical application. Radiogenomic approaches remain a high interest research field that may provide a rapid and accurate diagnosis and prognosis of patients with intracranial glioma.

AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?

Radiomic models outperform clinical data for outcome prediction in high-grade gliomas (HGG). However, lack of parameter standardization limits clinical applications. Many machine learning (ML) radiomic models employ single classifiers rather than ensemble learning, which is known to boost performance, and comparative analyses are lacking in the literature. We aimed to compare ML classifiers to predict clinically relevant tasks for HGG: overall survival (OS), isocitrate dehydrogenase (IDH) mutation, O-6-methylguanine-DNA-methyltransferase (MGMT) promoter methylation, epidermal growth factor receptor vIII (EGFR) amplification, and Ki-67 expression, based on radiomic features from conventional and advanced magnetic resonance imaging (MRI). Our objective was to identify the best algorithm for each task. One hundred fifty-six adult patients with pathologic diagnosis of HGG were included. Three tumoral regions were manually segmented: contrast-enhancing tumor, necrosis, and non-enhancing tumor. Radiomic features were extracted with a custom version of Pyradiomics and selected through Boruta algorithm. A Grid Search algorithm was applied when computing ten times K-fold cross-validation (K=10) to get the highest mean and lowest spread of accuracy. Model performance was assessed as AUC-ROC curve mean values with 95% confidence intervals (CI). Extreme Gradient Boosting (xGB) obtained highest accuracy for OS (74,5%), Adaboost (AB) for IDH mutation (87.5%), MGMT methylation (70,8%), Ki-67 expression (86%), and EGFR amplification (81%). Ensemble classifiers showed the best performance across tasks. High-scoring radiomic features shed light on possible correlations between MRI and tumor histology.

COVID Mortality Prediction with Machine Learning Methods: A Systematic Review and Critical Appraisal

More than a year has passed since the report of the first case of coronavirus disease 2019 (COVID), and increasing deaths continue to occur. Minimizing the time required for resource allocation and clinical decision making, such as triage, choice of ventilation modes and admission to the intensive care unit is important. Machine learning techniques are acquiring an increasingly sought-after role in predicting the outcome of COVID patients. Particularly, the use of baseline machine learning techniques is rapidly developing in COVID mortality prediction, since a mortality prediction model could rapidly and effectively help clinical decision-making for COVID patients at imminent risk of death. Recent studies reviewed predictive models for SARS-CoV-2 diagnosis, severity, length of hospital stay, intensive care unit admission or mechanical ventilation modes outcomes; however, systematic reviews focused on prediction of COVID mortality outcome with machine learning methods are lacking in the literature. The present review looked into the studies that implemented machine learning, including deep learning, methods in COVID mortality prediction thus trying to present the existing published literature and to provide possible explanations of the best results that the studies obtained. The study also discussed challenging aspects of current studies, providing suggestions for future developments.

In Vivo Brain GSH: MRS Methods and Clinical Applications

Glutathione (GSH) is an important antioxidant implicated in several physiological functions, including the oxidation−reduction reaction balance and brain antioxidant defense against endogenous and exogenous toxic agents. Altered brain GSH levels may reflect inflammatory processes associated with several neurologic disorders. An accurate and reliable estimation of cerebral GSH concentrations could give a clear and thorough understanding of its metabolism within the brain, thus providing a valuable benchmark for clinical applications. In this context, we aimed to provide an overview of the different magnetic resonance spectroscopy (MRS) technologies introduced for in vivo human brain GSH quantification both in healthy control (HC) volunteers and in subjects affected by different neurological disorders (e.g., brain tumors, and psychiatric and degenerative disorders). Additionally, we aimed to provide an exhaustive list of normal GSH concentrations within different brain areas. The definition of standard reference values for different brain areas could lead to a better interpretation of the altered GSH levels recorded in subjects with neurological disorders, with insights into the possible role of GSH as a biomarker and therapeutic target.

Glioblastoma radiomics to predict survival: Diffusion characteristics of surrounding nonenhancing tissue to select patients for extensive resection

BACKGROUND AND PURPOSE

Glioblastoma (GBM) is an aggressive primary CNS neoplasm with poor overall survival (OS) despite standard of care. On MRI, GBM is usually characterized by an enhancing portion (CET) (surgery target) and a nonenhancing surrounding (NET). Extent of resection is a long debated issue in GBM, with recent evidence suggesting that both CET and NET should be resected in <65 years old patients, regardless of other risk factors (i.e., molecular biomarkers). Our aim was to test a radiomic model for patient survival stratification in <65 years old patients, by analyzing MRI features of NET, to aid tumor resection.

METHODS

Sixty-eight <65 years old GBM patients, with extensive CET resection, were selected. Resection was evaluated by manually segmenting CET on volumetric T1-weighted MRI pre and postsurgery (within 72 h). All patients underwent the same treatment protocol including chemoradiation. NET radiomic features were extracted with a custom version of Pyradiomics. Feature selection was performed with principal component analysis (PCA) and its effect on survival tested with Cox regression model. Twelve months OS discrimination was tested by t-test followed by logistic regression. Statistical significance was set at p<0.05. The most relevant features were identified from the component matrix.

RESULTS

Five PCA components (PC1-5) explained 90% of the variance. PC5 resulted significant in the Cox model (p = 0.002; exp(B) = 0.686), at t-test (p = 0.002) and logistic regression analysis (p = 0.006). Apparent diffusion coefficient (ADC)-based features were the most significant for patient survival stratification.

CONCLUSIONS

ADC radiomic features on NET predict survival after standard therapy and could be used to improve patient selection for more extensive surgery.