Related Experiment Videos
YOLO-LS: a novel deep learning framework for brain tumor segmentation in Magnetic Resonance Imaging.
Jinghui Chen1, Yan Hu1, Tao Yang1
1The First Clinical Medical College, The Affiliated People's Hospital of Fujian University of Traditional Chinese Medicine, Fuzhou, Fujian, China.
Scientific Reports
|May 13, 2026
Summary
This study introduces YOLO-LS, a lightweight AI model for precise brain tumor segmentation in MRI scans. It balances high accuracy with computational efficiency, aiding diagnosis in resource-limited settings.
Area of Science:
- Medical Imaging and Artificial Intelligence
- Neuro-oncology and Computational Pathology
Background:
- Brain tumor segmentation in MRI is crucial but challenging due to tumor heterogeneity and limitations of manual annotation.
- Existing automated methods often struggle with a precision-efficiency trade-off, especially at infiltrative tumor boundaries, hindering clinical deployment.
Purpose of the Study:
- To develop an efficient and highly accurate framework, YOLO-LS (Lightweight Segmentation), for brain tumor detection and segmentation in MRI scans.
- To address the limitations of current automated segmentation techniques by improving boundary adherence and computational efficiency.
Main Methods:
- Proposed YOLO-LS framework based on YOLO11n-seg, integrating ShuffleNet V1 as a lightweight backbone for reduced complexity.
- Incorporated DySample dynamic upsampling to preserve fine-grained details and improve boundary recovery.
- Optimized the neck network with a C3k2-PoolingFormer module for efficient multi-scale feature fusion and global context aggregation.
Main Results:
- YOLO-LS achieved high performance on internal datasets: mAP50 of 0.953 ± 0.011, Dice coefficient of 0.91 ± 0.01, and HD95 of 4.35 ± 0.34 mm.
- Demonstrated significant computational cost reduction (15.6% decrease to 8.1 GFLOPs) with improved mAP50 (2.9% increase) compared to baseline YOLO11n-seg.
- Showcased strong generalization on an external dataset (Dice coefficient of 0.895) and outperformed state-of-the-art models like U-Net and Swin-UNet.
Conclusions:
- YOLO-LS effectively balances segmentation precision, computational efficiency, and interpretability for brain tumor analysis.
- The proposed framework shows significant potential for assisting diagnostic workflows, particularly in resource-constrained clinical environments.
- Innovations in backbone, upsampling, and neck network design contribute to superior performance in brain tumor segmentation.
Related Concept Videos
01:24Magnetic Resonance Imaging
Magnetic resonance imaging (MRI) is a noninvasive medical imaging technique based on a phenomenon of nuclear physics discovered in the 1930s, in which matter exposed to magnetic fields and radio waves was found to emit radio signals. In 1970, a physician and researcher named Raymond Damadian noticed that malignant (cancerous) tissue gave off different signals than normal body tissue. He applied for a patent for the first MRI scanning device in clinical use by the early 1980s. The early MRI...
01:27Imaging Studies IV: Magnetic Resonance Imaging
Introduction:Magnetic Resonance Imaging, or MRI, can include a specialized imaging technique of the urinary system known as Magnetic Resonance Urography (MRU). This radiation-free technique uses strong magnetic fields and radio waves to produce detailed images with the help of a computer. MRU is particularly effective for visualizing fluid-filled structures like the kidneys, ureters, and bladder.Applications of MRI in the Genitourinary SystemKidneys and Ureters: MRI detects tumors, cysts,...