Spectrum of qualitative and quantitative imaging of pilomyxoid, intermediate pilomyxoid and pilocytic astrocytomas in relation to their genetic alterations.

PURPOSE: Pilomyxoid astrocytomas (PMA) are pediatric brain tumors predominantly located in the suprasellar region, third ventricle and posterior fossa, which are considered to be more clinically aggressive than pilocytic astrocytomas (PA). Another entity, intermediate pilomyxoid tumors (IPT), exists within the spectrum of pilocytic/pilomyxoid astrocytomas. The 2021 WHO CNS classification refrained from assigning grade 1 or 2 status to PMA, thereby reflecting the need to further elucidate their clinical and imaging characteristics. METHODS: We included a total of 15 patients with PMA, IPT and suprasellar PA. We retrospectively evaluated immunohistochemistry, imaging findings and diffusion characteristics within these tumors as well as whole exome sequencing for three of the cases. RESULTS: 87% of the tumors were supratentorial with 11 cases suprasellar in location, 1 case located in the frontal white matter and 1 in the hippocampus. 6 cases demonstrated intraventricular extension. ADC values were higher in PMA and IPT than PA. 3 cases demonstrated KIAA1549-BRAF-fusion, 2 had BRAF[Formula: see text]-mutation and 6 were BRAF-wildtype. All cases had recurrence/progression on follow-up. CONCLUSION: PMA and IPT do not demonstrate aggressive imaging characteristics in respect to their diffusion imaging with ADC values being higher than PA. Lack of BRAF-alteration in PMA corresponded to atypical location of tumors with atypical driver mutations and mechanisms.

Clinical implications of Peri-hematomal edema microperfusion fraction in intracerebral hemorrhage intravoxel incoherent motion imaging - A pilot study.

BACKGROUND AND PURPOSE: Perihematomal edema (PHE) represents the secondary brain injury after intracerebral hemorrhage (ICH). However, neurobiological characteristics of post-ICH parenchymal injury other than PHE volume have not been fully characterized. Using intravoxel incoherent motion imaging (IVIM), we explored the clinical correlates of PHE diffusion and (micro)perfusion metrics in subacute ICH. MATERIALS AND METHODS: In 41 consecutive patients scanned 1-to-7 days after supratentorial ICH, we determined the mean diffusion (D), pseudo-diffusion (D*), and perfusion fraction (F) within manually segmented PHE. Using univariable and multivariable statistics, we evaluated the relationship of these IVIM metrics with 3-month outcome based on the modified Rankin Scale (mRS). RESULTS: In our cohort, the average (± standard deviation) age of patients was 68.6±15.6 years, median (interquartile) baseline National Institute of Health Stroke Scale (NIHSS) was 7 (3-13), 11 (27 %) patients had poor outcomes (mRS>3), and 4 (10 %) deceased during the follow-up period. In univariable analyses, admission NIHSS (p < 0.001), ICH volume (p = 0.019), ICH+PHE volume (p = 0.016), and average F of the PHE (p = 0.005) had significant correlation with 3-month mRS. In multivariable model, the admission NIHSS (p = 0.006) and average F perfusion fraction of the PHE (p = 0.003) were predictors of 3-month mRS. CONCLUSION: The IVIM perfusion fraction (F) maps represent the blood flow within microvasculature. Our pilot study shows that higher PHE microperfusion in subacute ICH is associated with worse outcomes. Once validated in larger cohorts, IVIM metrics may provide insight into neurobiology of post-ICH secondary brain injury and identify at-risk patients who may benefit from neuroprotective therapy.

Incidence and characteristics of metastatic intracranial lesions in stage III and IV melanoma: a single institute retrospective analysis.

INTRODUCTION: The study aimed to describe the brain metastases (BM) incidence, at diagnosis and follow-up, in patients initially presenting with stage III or IV melanoma and characterize their metastatic brain lesions. We also sought to describe the association of common genetic mutations and immunotherapy with BM development in advanced melanoma. METHODS: Using our institution's tumor registry, we identified patients with initial diagnoses of stage III and stage IV melanoma. In this cohort, we obtained BM incidence at diagnosis and follow-up, characterized the metastatic brain lesions and primary tumor's genetic profile. RESULTS: During the follow-up period, 22.9% of patients with an initial diagnosis of stage III developed BM. In this cohort, the median time for BM occurrence was 20 months; [95% CI (14-29)]. Likewise, 37.7% of patients with Stage IV melanoma presented with BM at the time of diagnosis, and 22.7% of remaining patients developed BM at follow-up over a median duration of 6 months [95% CI (4-11)]. Therefore, suggesting an overall incidence of 51.9% in stage IV melanoma. Next, we observed that the incidence of BM development during the follow-up period significantly decreased from 2012 to 2017 (p < 0.001). Lastly, we found a significantly higher frequency of mutational BRAF in the primary tumor of patients with BM (68.7% vs. 31.2%; p = 0.02). CONCLUSIONS: While the overall incidence of BM remains high, the decreasing incidence of BM over the follow-up period is promising. Similar BM incidence in patients with an initial diagnosis of stage III or stage IV warrants appropriate imaging surveillance regimen for stage III patients.

Type of bony involvement predicts genomic subgroup in sphenoid wing meningiomas.

PURPOSE: As sphenoid wing meningiomas (SWMs) are associated with varying degrees of bony involvement, we sought to understand potential relationships between genomic subgroup and this feature. METHODS: Patients treated at Yale-New Haven Hospital for SWM were reviewed. Genomic subgroup was determined via whole exome sequencing, while the extent of bony involvement was radiographically classified as no bone invasion (Type I), hyperostosis only (Type II), tumor invasion only (Type III), or both hyperostosis and tumor invasion (Type IV). Among additional clinical variables collected, a subset of tumors was identified as spheno-orbital meningiomas (SOMs). Machine-learning approaches were used to predict genomic subgroups based on pre-operative clinical features. RESULTS: Among 64 SWMs, 53% had Type-II, 9% had Type-III, and 14% had Type-IV bone involvement; nine SOMs were identified. Tumors with invasion (i.e., Type III or IV) were more likely to be WHO grade II (p: 0.028). Additionally, tumors with invasion were nearly 30 times more likely to harbor NF2 mutations (OR 27.6; p: 0.004), while hyperostosis only were over 4 times more likely to have a TRAF7 mutation (OR 4.5; p: 0.023). SOMs were a significant predictor of underlying TRAF7 mutation (OR 10.21; p: 0.004). CONCLUSIONS: SWMs with invasion into bone tend to be higher grade and are more likely to be NF2 mutated, while SOMs and those with hyperostosis are associated with TRAF7 variants. Pre-operative prediction of molecular subtypes based on radiographic bony characteristics may have significant biological and clinical implications based on known recurrence patterns associated with genomic drivers and grade.

A Deep Learning Model to Predict a Diagnosis of Alzheimer Disease by Using 18F-FDG PET of the Brain.

Purpose To develop and validate a deep learning algorithm that predicts the final diagnosis of Alzheimer disease (AD), mild cognitive impairment, or neither at fluorine 18 (18F) fluorodeoxyglucose (FDG) PET of the brain and compare its performance to that of radiologic readers. Materials and Methods Prospective 18F-FDG PET brain images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) (2109 imaging studies from 2005 to 2017, 1002 patients) and retrospective independent test set (40 imaging studies from 2006 to 2016, 40 patients) were collected. Final clinical diagnosis at follow-up was recorded. Convolutional neural network of InceptionV3 architecture was trained on 90% of ADNI data set and tested on the remaining 10%, as well as the independent test set, with performance compared to radiologic readers. Model was analyzed with sensitivity, specificity, receiver operating characteristic (ROC), saliency map, and t-distributed stochastic neighbor embedding. Results The algorithm achieved area under the ROC curve of 0.98 (95% confidence interval: 0.94, 1.00) when evaluated on predicting the final clinical diagnosis of AD in the independent test set (82% specificity at 100% sensitivity), an average of 75.8 months prior to the final diagnosis, which in ROC space outperformed reader performance (57% [four of seven] sensitivity, 91% [30 of 33] specificity; P < .05). Saliency map demonstrated attention to known areas of interest but with focus on the entire brain. Conclusion By using fluorine 18 fluorodeoxyglucose PET of the brain, a deep learning algorithm developed for early prediction of Alzheimer disease achieved 82% specificity at 100% sensitivity, an average of 75.8 months prior to the final diagnosis. © RSNA, 2018 Online supplemental material is available for this article. See also the editorial by Larvie in this issue.

Case 245: Erdheim-Chester Disease.

History A 53-year-old man experienced headache and double vision that progressed over 1 year. After a traumatic fall, he was hospitalized, and proptosis was identified at physical examination. Laboratory tests were remarkable for leukocytosis. Hematocrit level, thyroid stimulating hormone level, autoimmune antibody level, erythrocyte sedimentation rate, and C-reactive protein level were normal. Computed tomography (CT) of the head revealed bilateral intraconal masses, for which magnetic resonance (MR) imaging of the orbits was subsequently performed ( Fig 1 ). CT imaging of the chest and abdomen ( Fig 2 ) revealed periaortic and retroperitoneal stranding. Perinephric biopsy was performed, and a diagnosis of immunoglobulin G4 (IgG4)-related disease was made based on identification of a few plasma cells per high-power field that were positive for IgG4. Orbital biopsy was then performed, but the results were inconclusive for IgG4-related disease. The patient was discharged and given steroid therapy for presumed IgG4-related disease. [Figure: see text][Figure: see text][Figure: see text][Figure: see text][Figure: see text][Figure: see text] Several months later, the patient returned to our institution with progressive symptoms despite ongoing steroid treatment. His case was reviewed by several specialists to develop alternative treatments for IgG4-related disease. After review of the available images, a neuroradiology fellow (M.D.M.) performed history taking and a physical examination and subsequently recommended radiography of the lower extremities ( Fig 3 ). [Figure: see text][Figure: see text].

In vitro clearance of doxorubicin with a DNA-based filtration device designed for intravascular use with intra-arterial chemotherapy.

To report a novel method using immobilized DNA within mesh to sequester drugs that have intrinsic DNA binding characteristics directly from flowing blood. DNA binding experiments were carried out in vitro with doxorubicin in saline (PBS solution), porcine serum, and porcine blood. Genomic DNA was used to identify the concentration of DNA that shows optimum binding clearance of doxorubicin from solution. Doxorubicin binding kinetics by DNA enclosed within porous mesh bags was evaluated. Flow model simulating blood flow in the inferior vena cava was used to determine in vitro binding kinetics between doxorubicin and DNA. The kinetics of doxorubicin binding to free DNA is dose-dependent and rapid, with 82-96 % decrease in drug concentration from physiologic solutions within 1 min of reaction time. DNA demonstrates faster binding kinetics by doxorubicin as compared to polystyrene resins that use an ion exchange mechanism. DNA contained within mesh yields an approximately 70 % decrease in doxorubicin concentration from solution within 5 min. In the IVC flow model, there is a 70 % drop in doxorubicin concentration at 60 min. A DNA-containing ChemoFilter device can rapidly clear clinical doses of doxorubicin from a flow model in simple and complex physiological solutions, thereby suggesting a novel approach to reduce the toxicity of DNA-binding drugs.

Comparison of Volumetric and 2D Measurements and Longitudinal Trajectories in the Response Assessment of BRAF V600E-Mutant Pediatric Gliomas in the Pacific Pediatric Neuro-Oncology Consortium Clinical Trial.

BACKGROUND AND PURPOSE: Response on imaging is widely used to evaluate treatment efficacy in clinical trials of pediatric gliomas. While conventional criteria rely on 2D measurements, volumetric analysis may provide a more comprehensive response assessment. There is sparse research on the role of volumetrics in pediatric gliomas. Our purpose was to compare 2D and volumetric analysis with the assessment of neuroradiologists using the Brain Tumor Reporting and Data System (BT-RADS) in BRAF V600E-mutant pediatric gliomas. MATERIALS AND METHODS: Manual volumetric segmentations of whole and solid tumors were compared with 2D measurements in 31 participants (292 follow-up studies) in the Pacific Pediatric Neuro-Oncology Consortium 002 trial (NCT01748149). Two neuroradiologists evaluated responses using BT-RADS. Receiver operating characteristic analysis compared classification performance of 2D and volumetrics for partial response. Agreement between volumetric and 2D mathematically modeled longitudinal trajectories for 25 participants was determined using the model-estimated time to best response. RESULTS: Of 31 participants, 20 had partial responses according to BT-RADS criteria. Receiver operating characteristic curves for the classification of partial responders at the time of first detection (median = 2 months) yielded an area under the curve of 0.84 (95% CI, 0.69-0.99) for 2D area, 0.91 (95% CI, 0.80-1.00) for whole-volume, and 0.92 (95% CI, 0.82-1.00) for solid volume change. There was no significant difference in the area under the curve between 2D and solid (P = .34) or whole volume (P = .39). There was no significant correlation in model-estimated time to best response (ρ = 0.39, P >.05) between 2D and whole-volume trajectories. Eight of the 25 participants had a difference of ≥90 days in transition from partial response to stable disease between their 2D and whole-volume modeled trajectories. CONCLUSIONS: Although there was no overall difference between volumetrics and 2D in classifying partial response assessment using BT-RADS, further prospective studies will be critical to elucidate how the observed differences in tumor 2D and volumetric trajectories affect clinical decision-making and outcomes in some individuals.

A large open access dataset of brain metastasis 3D segmentations on MRI with clinical and imaging information.

Resection and whole brain radiotherapy (WBRT) are standard treatments for brain metastases (BM) but are associated with cognitive side effects. Stereotactic radiosurgery (SRS) uses a targeted approach with less side effects than WBRT. SRS requires precise identification and delineation of BM. While artificial intelligence (AI) algorithms have been developed for this, their clinical adoption is limited due to poor model performance in the clinical setting. The limitations of algorithms are often due to the quality of datasets used for training the AI network. The purpose of this study was to create a large, heterogenous, annotated BM dataset for training and validation of AI models. We present a BM dataset of 200 patients with pretreatment T1, T1 post-contrast, T2, and FLAIR MR images. The dataset includes contrast-enhancing and necrotic 3D segmentations on T1 post-contrast and peritumoral edema 3D segmentations on FLAIR. Our dataset contains 975 contrast-enhancing lesions, many of which are sub centimeter, along with clinical and imaging information. We used a streamlined approach to database-building through a PACS-integrated segmentation workflow.

Everolimus for Children With Recurrent or Progressive Low-Grade Glioma: Results From the Phase II PNOC001 Trial.

PURPOSE: The PNOC001 phase II single-arm trial sought to estimate progression-free survival (PFS) associated with everolimus therapy for progressive/recurrent pediatric low-grade glioma (pLGG) on the basis of phosphatidylinositol 3-kinase (PI3K)/AKT/mammalian target of rapamycin (mTOR) pathway activation as measured by phosphorylated-ribosomal protein S6 and to identify prognostic and predictive biomarkers. PATIENTS AND METHODS: Patients, age 3-21 years, with progressive/recurrent pLGG received everolimus orally, 5 mg/m2 once daily. Frequency of driver gene alterations was compared among independent pLGG cohorts of newly diagnosed and progressive/recurrent patients. PFS at 6 months (primary end point) and median PFS (secondary end point) were estimated for association with everolimus therapy. RESULTS: Between 2012 and 2019, 65 subjects with progressive/recurrent pLGG (median age, 9.6 years; range, 3.0-19.9; 46% female) were enrolled, with a median follow-up of 57.5 months. The 6-month PFS was 67.4% (95% CI, 60.0 to 80.0) and median PFS was 11.1 months (95% CI, 7.6 to 19.8). Hypertriglyceridemia was the most common grade ≥3 adverse event. PI3K/AKT/mTOR pathway activation did not correlate with clinical outcomes (6-month PFS, active 68.4% v nonactive 63.3%; median PFS, active 11.2 months v nonactive 11.1 months; P = .80). Rare/novel KIAA1549::BRAF fusion breakpoints were most frequent in supratentorial midline pilocytic astrocytomas, in patients with progressive/recurrent disease, and correlated with poor clinical outcomes (median PFS, rare/novel KIAA1549::BRAF fusion breakpoints 6.1 months v common KIAA1549::BRAF fusion breakpoints 16.7 months; P < .05). Multivariate analysis confirmed their independent risk factor status for disease progression in PNOC001 and other, independent cohorts. Additionally, rare pathogenic germline variants in homologous recombination genes were identified in 6.8% of PNOC001 patients. CONCLUSION: Everolimus is a well-tolerated therapy for progressive/recurrent pLGGs. Rare/novel KIAA1549::BRAF fusion breakpoints may define biomarkers for progressive disease and should be assessed in future clinical trials.

Comparison of volumetric and 2D-based response methods in the PNOC-001 pediatric low-grade glioma clinical trial.

Background: Although response in pediatric low-grade glioma (pLGG) includes volumetric assessment, more simplified 2D-based methods are often used in clinical trials. The study's purpose was to compare volumetric to 2D methods. Methods: An expert neuroradiologist performed solid and whole tumor (including cyst and edema) volumetric measurements on MR images using a PACS-based manual segmentation tool in 43 pLGG participants (213 total follow-up images) from the Pacific Pediatric Neuro-Oncology Consortium (PNOC-001) trial. Classification based on changes in volumetric and 2D measurements of solid tumor were compared to neuroradiologist visual response assessment using the Brain Tumor Reporting and Data System (BT-RADS) criteria for a subset of 65 images using receiver operating characteristic (ROC) analysis. Longitudinal modeling of solid tumor volume was used to predict BT-RADS classification in 54 of the 65 images. Results: There was a significant difference in ROC area under the curve between 3D solid tumor volume and 2D area (0.96 vs 0.78, P = .005) and between 3D solid and 3D whole volume (0.96 vs 0.84, P = .006) when classifying BT-RADS progressive disease (PD). Thresholds of 15-25% increase in 3D solid tumor volume had an 80% sensitivity in classifying BT-RADS PD included in their 95% confidence intervals. The longitudinal model of solid volume response had a sensitivity of 82% and a positive predictive value of 67% for detecting BT-RADS PD. Conclusions: Volumetric analysis of solid tumor was significantly better than 2D measurements in classifying tumor progression as determined by BT-RADS criteria and will enable more comprehensive clinical management.

Evolution and implementation of radiographic response criteria in neuro-oncology.

Radiographic response assessment in neuro-oncology is critical in clinical practice and trials. Conventional criteria, such as the MacDonald and response assessment in neuro-oncology (RANO) criteria, rely on bidimensional (2D) measurements of a single tumor cross-section. Although RANO criteria are established for response assessment in clinical trials, there is a critical need to address the complexity of brain tumor treatment response with multiple new approaches being proposed. These include volumetric analysis of tumor compartments, structured MRI reporting systems like the Brain Tumor Reporting and Data System, and standardized approaches to advanced imaging techniques to distinguish tumor response from treatment effects. In this review, we discuss the strengths and limitations of different neuro-oncology response criteria and summarize current research findings on the role of novel response methods in neuro-oncology clinical trials and practice.

A Large Open Access Dataset of Brain Metastasis 3D Segmentations with Clinical and Imaging Feature Information.

Resection and whole brain radiotherapy (WBRT) are the standards of care for the treatment of patients with brain metastases (BM) but are often associated with cognitive side effects. Stereotactic radiosurgery (SRS) involves a more targeted treatment approach and has been shown to avoid the side effects associated with WBRT. However, SRS requires precise identification and delineation of BM. While many AI algorithms have been developed for this purpose, their clinical adoption has been limited due to poor model performance in the clinical setting. Major reasons for non-generalizable algorithms are the limitations in the datasets used for training the AI network. The purpose of this study was to create a large, heterogenous, annotated BM dataset for training and validation of AI models to improve generalizability. We present a BM dataset of 200 patients with pretreatment T1, T1 post-contrast, T2, and FLAIR MR images. The dataset includes contrast-enhancing and necrotic 3D segmentations on T1 post-contrast and whole tumor (including peritumoral edema) 3D segmentations on FLAIR. Our dataset contains 975 contrast-enhancing lesions, many of which are sub centimeter, along with clinical and imaging feature information. We used a streamlined approach to database-building leveraging a PACS-integrated segmentation workflow.

Clinical Evaluation of Nuclear Imaging Agents in Breast Cancer.

Precision medicine is the customization of therapy for specific groups of patients using genetic or molecular profiling. Noninvasive imaging is one strategy for molecular profiling and is the focus of this review. The combination of imaging and therapy for precision medicine gave rise to the field of theranostics. In breast cancer, the detection and quantification of therapeutic targets can help assess their heterogeneity, especially in metastatic disease, and may help guide clinical decisions for targeted treatments. Positron emission tomography (PET) or single-photon emission tomography (SPECT) imaging has the potential to play an important role in the molecular profiling of therapeutic targets in vivo for the selection of patients who are likely to respond to corresponding targeted therapy. In this review, we discuss the state-of-the-art nuclear imaging agents in clinical research for breast cancer. We reviewed 17 clinical studies on PET or SPECT agents that target 10 different receptors in breast cancer. We also discuss the limitations of the study designs and of the imaging agents in these studies. Finally, we offer our perspective on which imaging agents have the highest potential to be used in clinical practice in the future.

Machine Learning Tools for Image-Based Glioma Grading and the Quality of Their Reporting: Challenges and Opportunities.

Technological innovation has enabled the development of machine learning (ML) tools that aim to improve the practice of radiologists. In the last decade, ML applications to neuro-oncology have expanded significantly, with the pre-operative prediction of glioma grade using medical imaging as a specific area of interest. We introduce the subject of ML models for glioma grade prediction by remarking upon the models reported in the literature as well as by describing their characteristic developmental workflow and widely used classifier algorithms. The challenges facing these models-including data sources, external validation, and glioma grade classification methods -are highlighted. We also discuss the quality of how these models are reported, explore the present and future of reporting guidelines and risk of bias tools, and provide suggestions for the reporting of prospective works. Finally, this review offers insights into next steps that the field of ML glioma grade prediction can take to facilitate clinical implementation.

Real-time PACS-integrated longitudinal brain metastasis tracking tool provides comprehensive assessment of treatment response to radiosurgery.

Background: Treatment of brain metastases can be tailored to individual lesions with treatments such as stereotactic radiosurgery. Accurate surveillance of lesions is a prerequisite but challenging in patients with multiple lesions and prior imaging studies, in a process that is laborious and time consuming. We aimed to longitudinally track several lesions using a PACS-integrated lesion tracking tool (LTT) to evaluate the efficiency of a PACS-integrated lesion tracking workflow, and characterize the prevalence of heterogenous response (HeR) to treatment after Gamma Knife (GK). Methods: We selected a group of brain metastases patients treated with GK at our institution. We used a PACS-integrated LTT to track the treatment response of each lesion after first GK intervention to maximally seven diagnostic follow-up scans. We evaluated the efficiency of this tool by comparing the number of clicks necessary to complete this task with and without the tool and examined the prevalence of HeR in treatment. Results: A cohort of eighty patients was selected and 494 lesions were measured and tracked longitudinally for a mean follow-up time of 374 days after first GK. Use of LTT significantly decreased number of necessary clicks. 81.7% of patients had HeR to treatment at the end of follow-up. The prevalence increased with increasing number of lesions. Conclusions: Lesions in a single patient often differ in their response to treatment, highlighting the importance of individual lesion size assessments for further treatment planning. PACS-integrated lesion tracking enables efficient lesion surveillance workflow and specific and objective result reports to treating clinicians.

Machine Learning Models for Classifying High- and Low-Grade Gliomas: A Systematic Review and Quality of Reporting Analysis.

Objectives: To systematically review, assess the reporting quality of, and discuss improvement opportunities for studies describing machine learning (ML) models for glioma grade prediction. Methods: This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses of Diagnostic Test Accuracy (PRISMA-DTA) statement. A systematic search was performed in September 2020, and repeated in January 2021, on four databases: Embase, Medline, CENTRAL, and Web of Science Core Collection. Publications were screened in Covidence, and reporting quality was measured against the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) Statement. Descriptive statistics were calculated using GraphPad Prism 9. Results: The search identified 11,727 candidate articles with 1,135 articles undergoing full text review and 85 included in analysis. 67 (79%) articles were published between 2018-2021. The mean prediction accuracy of the best performing model in each study was 0.89 ± 0.09. The most common algorithm for conventional machine learning studies was Support Vector Machine (mean accuracy: 0.90 ± 0.07) and for deep learning studies was Convolutional Neural Network (mean accuracy: 0.91 ± 0.10). Only one study used both a large training dataset (n>200) and external validation (accuracy: 0.72) for their model. The mean adherence rate to TRIPOD was 44.5% ± 11.1%, with poor reporting adherence for model performance (0%), abstracts (0%), and titles (0%). Conclusions: The application of ML to glioma grade prediction has grown substantially, with ML model studies reporting high predictive accuracies but lacking essential metrics and characteristics for assessing model performance. Several domains, including generalizability and reproducibility, warrant further attention to enable translation into clinical practice. Systematic Review Registration: PROSPERO, identifier CRD42020209938.

Machine Learning Applications for Differentiation of Glioma from Brain Metastasis-A Systematic Review.

Glioma and brain metastasis can be difficult to distinguish on conventional magnetic resonance imaging (MRI) due to the similarity of imaging features in specific clinical circumstances. Multiple studies have investigated the use of machine learning (ML) models for non-invasive differentiation of glioma from brain metastasis. Many of the studies report promising classification results, however, to date, none have been implemented into clinical practice. After a screening of 12,470 studies, we included 29 eligible studies in our systematic review. From each study, we aggregated data on model design, development, and best classifiers, as well as quality of reporting according to the TRIPOD statement. In a subset of eligible studies, we conducted a meta-analysis of the reported AUC. It was found that data predominantly originated from single-center institutions (n = 25/29) and only two studies performed external validation. The median TRIPOD adherence was 0.48, indicating insufficient quality of reporting among surveyed studies. Our findings illustrate that despite promising classification results, reliable model assessment is limited by poor reporting of study design and lack of algorithm validation and generalizability. Therefore, adherence to quality guidelines and validation on outside datasets is critical for the clinical translation of ML for the differentiation of glioma and brain metastasis.