AI and Cancer Research ── Brain ── 2026-11-02
Brain tumours: treatment, trials and AI
- Abstract
ABSTRACT Background Quantitative MRI may enable noninvasive glioma characterization, but optimal analytic strategies and prognostic utility remain uncertain.
Purpose: To compare volume‐of‐interest (VOI)‐mean, radiomics, and 2D, 2.5D, and 3D deep learning (DL) for preoperative molecular classification and assess the prognostic value of a 2.5D‐DL–derived isocitrate dehydrogenase wild‐type score (2.5D‐IDHwt score) in a postoperative clinicopathologic survival model. Study Type Retrospective multicenter study. Population One thousand four hundred and eight adults with diffuse glioma: training ( n = 409), temporal validation ( n = 300), and three external test cohorts ( n = 119, 310, and 270). Field Strength/Sequence 3 T; spin‐echo echo‐planar diffusion‐weighted imaging, 3D pseudo‐continuous arterial spin labeling, and multidynamic multiecho synthetic MRI. Assessment Apparent diffusion coefficient, cerebral blood flow, T1, T2, proton‐density, and contrast‐enhanced T1 maps were generated. Two neuroradiologists (5 and 20 years' experience) delineated whole‐tumor VOIs. Six map‐specific models were stacked separately for VOI‐mean, radiomics, 2D‐DL, 2.5D‐DL, and 3D‐DL. 2.5D‐qMR‐Clin combined the 2.5D‐IDHwt score with age, O 6 ‐methylguanine‐DNA methyltransferase promoter methylation, and extent of resection and was compared with PathoClin incorporating pathological IDH status. Statistical Tests Area under the curve (AUC), DeLong tests, Harrell C‐index, bootstrap resampling, calibration, decision‐curve analysis, Kaplan–Meier analysis, and Cox regression. p < 0.05 indicated significance.
Results: The 2.5D‐DL stacking model achieved the highest external AUCs for IDH mutation (0.89–0.93) and 1p/19q codeletion (0.92–0.95), significantly outperforming the other four stacking models. Reverse Kaplan–Meier median follow‐up was 1344 days (95% CI, 1316–1383); 27.5% died. External C‐indices for 2.5D‐qMR‐Clin were 0.750, 0.778, and 0.762 versus 0.704, 0.722, and 0.723 for PathoClin, with significant differences in test cohorts 2 and 3. Fixed‐cutoff risk stratification was significant across all external cohorts (hazard ratios, 5.30–5.80). A web‐based 2.5D‐qMR‐Clin explorer provided 1‐, 2‐, and 3‐year OS estimates. Data Conclusion Among evaluated approaches, 2.5D‐DL stacking performed best for molecular classification, and the 2.5D‐IDHwt score improved OS assessment in the postoperative clinicopathologic model. Evidence Level 3. Technical Efficacy Stage 2.
Journal IF-equivalent: 2.4 (OpenAlex 2-year mean citedness, value as of 2026-10-15, retrieved 2026-10-16; not the official Clarivate IF)Reference: Ge X, Shen Y, Han Y, Li S, Feng X, Wang C, et al. Multiparametric Quantitative MRI With 2. 5D Deep Learning for Molecular Classification and Overall Survival Assessment in Adult‐Type Diffuse Gliomas. Magnetic Resonance Imaging. 2026 Oct 31 [Epub ahead of print]. doi:10.1002/jmri.70599.Checked: Abstract only - Abstract
A brain tumor (BT) arises due to the abnormal growth of cells. It ranks among the primary fatal malignancies for adults globally. Tumor treatment relies on early tumor recognition. The diagnosis is usually performed using medical imaging techniques, which assist in the accurate evaluation of the patient's brain. Therefore, timely disease detection and rapid treatment result in improved quality of life and increased life expectancy in such patients. Recently, deep learning (DL) models have led to considerable advancements in medical imaging. In this study, an automated brain tumor classification and severity prediction using image segmentation and a convolutional deep learning (ABTCSP-ISCDL) model is proposed. This study aims to develop an intelligent system for precise BT detection and severity classification to help with timely diagnosis and effective treatment planning. Besides, image segmentation is performed using a Transformer-based U-Net (TransUNet) method, which accurately isolates the tumor region. Furthermore, the ABTCSP-ISCDL technique performs feature extraction using the RegNet002 model to capture local and global features of the BT. Finally, the root mean square propagation (RMSProp) optimizer is used with a convolutional long short-term memory (ConvLSTM) network to perform tumor classification and severity prediction, improving both accuracy and overall performance. A comprehensive set of experiments on the ABTCSP-ISCDL model is performed using the Figshare BT dataset. The experimental analysis indicated that the proposed framework achieved an accuracy of 96.25% compared with conventional approaches.
Journal IF-equivalent: 1.4 (OpenAlex 2-year mean citedness, value as of 2026-11-01, retrieved 2026-11-02; not the official Clarivate IF)Reference: Chanemougavel V, Jayanthi K. Brain tumor segmentation, classification, and severity prediction using deep transfer learning and convolutional long short-term memory. IJATEE. 2026 Oct 31;13(143):854-75. doi:10.19101/ijatee.2025.121221250.Checked: Abstract only