Accepted for/Published in: JMIR Medical Informatics
Date Submitted: Sep 14, 2025
Date Accepted: Feb 28, 2026
Research on the Detection of Interpretable and Fine-Grained Brain Tumor MRI based on Progressive Pruning: Machine Learning Model Development and Validation Study
ABSTRACT
Background:
Brain tumor is one of the most malignant diseases of the central nervous system, and early accurate detection is of great significance for improving patient survival rate. However, the heterogeneity of brain tumors in terms of morphology, size, and location on MRI image, as well as their similarity to surrounding normal brain tissue, poses significant challenges for tumor detection.
Objective:
This study aims to develop a high-performance brain tumor detection framework that integrates feature enhancement, channel attention, and progressive pruning, achieving an optimal balance between detection accuracy, model efficiency, and interpretability for clinical application.
Methods:
This paper, based on YOLOv11, proposes a Convolution-Prewitt and Pooling-based Preprocessing (CSPP) which highlights important structural detail more effectively than traditional statistics. A Dynamic Convolution-based C3k2 (DCC) is integrated to more efficiently capture both local and global features. A Channel Prior Convolutional Attention (CPCA) is introduced before the detection head, enabling the network to specifically focus on information-rich channels and key spatial regions. Through Progressive Hybrid Pruning Strategy (PHPS) , the model is lightweighted for efficient inference. Furthermore, Eigen-CAM is utilized to interpret the prediction result, making them more transparent and clinically valuable.
Results:
Extensive experiments on three brain tumor MRI datasets demonstrate the superior performance of CDCP-YOLO:On Br35H, mAP50 increased by 2.6%, mAP50:95 increased by 5.9%, and GFLOPs were reduced by 47.7%. On YOLO, mAP50 increased by 19.5%, mAP50:95 increased by 7.7%, and GFLOPs were reduced by 47.7%. On Capstone, mAP50 increased by 6.9%, mAP50:95 increased by 5.8%, and GFLOPs were reduced by 47.7%.
Conclusions:
The proposed CDCP-YOLO framework achieves an optimal balance between accuracy, efficiency, and interpretability, providing a lightweight yet reliable solution for brain tumor detection and clinical decision support.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.