Junhyeok Lee

An interpretable multimodal deep learning model integrating MRI with clinical and molecular variables enabled survival prediction with time-dependent model interpretability in glioblastoma, with imaging-derived features providing complementary prognostic information.

Jang J, et al. Korean J Radiol. 2026 Sep;27(9):803-806. https://doi.org/10.3348/kjr.2026.0981

To explore how molecular glioblastoma (mol-GBM) differs from histological glioblastoma (hist-GBM) in tumor heterogeneity using multi-parametric physiologic MRI-based tumor habitat analysis. In this multi-institutional retrospective study, imaging data were collected from two tertiary centers: 13 mol-GBMs and 39 hist-GBMs from institution 1 (2007-2024) for habitat definition, model development, and internal validation and nine mol-GBMs were obtained from institution 2 (2020-2024) for external validation. The apparent diffusion coefficient (ADC; cellularity), relative cerebral blood volume (rCBV; vascularity), and volume transfer constant (K^trans; permeability) were binarized into high/low categories, which yielded eight spatial habitat clusters to visualize tumor heterogeneity. Habitat proportions and intratumoral heterogeneity features were compared between groups. Multivariable logistic regression models incorporating habitat-derived features were developed to distinguish mol-GBM from hist-GBM, and their performance was evaluated. Fifty-two patients (hist-GBM, n = 39; mol-GBM, n = 13; mean age, 60.0 ± 11.1 years; 22 male) were evaluated. Habitat analysis revealed that mol-GBM had a significantly lower proportion of the most malignant habitat (cluster 3: low-ADC, high rCBV, high K^trans; 0.80% vs. 7.0%, P < 0.001). The total proportion of high K^trans habitats was markedly lower in the mol-GBM group (5.7% vs. 21.3%, P < 0.001). A habitat-based multivariable model incorporating tumor volume, cluster 3 proportion, high K^trans proportion, and Shannon entropy achieved an area under receiver operating characteristic curve (AUC) of 0.87 (95% confidence interval [CI]: 0.73-0.97) at internal validation using a leave-one-out cross-validation, which was substantially greater, albeit nonsignificant, than the AUC for tumor size alone (0.70, 95% CI: 0.51-0.85, P = 0.057), with an external validation accuracy of 88.9% (8/9). Multiparametric physiologic MRI habitat analysis demonstrated differences in the tumor heterogeneity between mol-GBM and hist-GBM. Tumor permeability (K^trans) was the most discriminating parameter. Incorporating habitat-derived features alongside tumor volume may improve the discrimination of the two subtypes on MRI analysis. Further studies, including more rigorous external validation, are warranted.

Deep learning (DL) models are widely adopted in biomedical imaging, where image segmentation is increasingly recognized as a quantitative tool for extracting clinically meaningful information. However, model performance critically depends on dataset size and training configuration, including model capacity. Traditional sample size estimation methods are inadequate for DL due to its reliance on high-dimensional data and its nonlinear learning behavior. To address this gap, we propose a DL-specific framework to estimate the minimal dataset size required for stable segmentation performance. We validate this framework across two distinct clinical tasks: colorectal polyp segmentation from 2D endoscopic images (Kvasir-SEG) and glioma segmentation from 3D brain MRIs (BraTS 2020). We trained residual U-Nets—a simple, yet foundational architecture—across 200 configurations for Kvasir-SEG and 40 configurations for BraTS 2020, varying data subsets (2%–100% for the 2D task and 5%–100% for the 3D task). In both tasks, performance metrics such as the Dice Similarity Coefficient (DSC) consistently improved with increasing data and depth, but gains invariably plateaued beyond approximately 80% data usage. The best configuration for polyp segmentation (6 layers, 100% data) achieved a DSC of 0.86, while the best for brain tumor segmentation reached a DSC of 0.79. Critically, we introduce a surrogate modeling pipeline using Long Short-Term Memory (LSTM) networks to predict these performance curves. A simple uni-directional LSTM model accurately forecasted the final DSC, accurately forecasting the final DSC with low mean absolute error across both tasks. These findings demonstrate that segmentation performance can be reliably estimated with lightweight models, suggesting that collecting a moderate amount of high-quality data is often sufficient for developing clinically viable DL models. Our framework provides a practical, empirical method for optimizing resource allocation in medical AI development.

Precise delineation of meningiomas is crucial for effective radiotherapy (RT) planning, directly influencing treatment efficacy and preservation of adjacent healthy tissues. While automated deep learning approaches have demonstrated considerable potential, achieving consistently accurate clinical segmentation remains challenging due to tumor heterogeneity. Interactive Medical Image Segmentation (IMIS) addresses this challenge by integrating advanced AI techniques with clinical input. However, generic segmentation tools, despite widespread applicability, often lack the specificity required for clinically critical and disease-specific tasks like meningioma RT planning. To overcome these limitations, we introduce Interactive-MEN-RT, a dedicated IMIS tool specifically developed for clinician-assisted 3D meningioma segmentation in RT workflows. The system incorporates multiple clinically relevant interaction methods, including point annotations, bounding boxes, lasso tools, and scribbles, enhancing usability and clinical precision. In our evaluation involving 500 contrast-enhanced T1-weighted MRI scans from the BraTS 2025 Meningioma RT Segmentation Challenge, Interactive-MEN-RT demonstrated substantial improvement compared to other segmentation methods, achieving Dice similarity coefficients of up to 77.6\% and Intersection over Union scores of 64.8\%. These results emphasize the need for clinically tailored segmentation solutions in critical applications such as meningioma RT planning. The code is publicly available at: https://github.com/snuh-rad-aicon/Interactive-MEN-RT

An unsupervised learning framework detected blood-brain barrier leakage without pharmacokinetic models and arterial input function and distinguished between different tissue types within tumors, offering potential to improve diffuse glioma diagnosis.

MRI is an imaging technology that non-invasively obtains high-quality medical images for diagnosis. However, MRI has the major disadvantage of long scan times which cause patient discomfort and image artifacts. As one of the methods for reducing the long scan time of MRI, the parallel MRI method for reconstructing a high-fidelity MR image from under-sampled multi-coil k-space data is widely used. In this study, we propose a method to reconstruct a high-fidelity MR image from under-sampled multi-coil k-space data using deep-learning. The proposed multi-domain Neumann network with sensitivity maps (MDNNSM) is based on the Neumann network and uses a forward model including coil sensitivity maps for parallel MRI reconstruction. The MDNNSM consists of three main structures: the CNN-based sensitivity reconstruction block estimates coil sensitivity maps from multi-coil under-sampled k-space data; the recursive MR image reconstruction block reconstructs the MR image; and the skip connection accumulates each output and produces the final result. Experiments using the fastMRI T1-weighted brain image dataset were conducted at acceleration factors of 2, 4, and 8. Qualitative and quantitative experimental results show that the proposed MDNNSM method reconstructs MR images more accurately than other methods, including the generalized autocalibrating partially parallel acquisitions (GRAPPA) method and the original Neumann network.