CMRA-DETR: A Lightweight and High-Accuracy Detection Framework for MRI-Based Brain Tumor Identification
-
By
-
Weng, Cai
-
Huang, Bowei
-
Chen, Jinghui
-
Hu, Wei
-
Huang, Zhiqing
-
Weng, Punan
-
Zhao, Hongjia
-
Zheng, Minqin
-
April 21, 2026
-
Clinical Scorecard: CMRA-DETR: An Efficient and Precise Framework for Identifying Brain Tumors in MRI Scans
At a Glance
| Category | Detail |
|---|
| Condition | Brain Tumors |
| Key Mechanisms | CSP-MambaOut backbone, AIFI-MALA module, RetBlockC3 module |
| Target Population | Patients undergoing MRI scans for brain tumor evaluation |
| Care Setting | Clinical settings with MRI capabilities |
Key Highlights
- Achieves P = 95.5% and R = 95.7% on internal test dataset
- Records mAP@50 = 97.9% and mAP@50-95 = 82.6%
- Reduces parameters and GFLOPs by 37.7% and 30.9% respectively
- Demonstrates strong cross-dataset generalization
- Applicable for AI-assisted detection on resource-limited devices
Guideline-Based Recommendations
Diagnosis
- Utilize CMRA-DETR for automated identification of brain tumors in MRI scans.
Management
- Implement AI-assisted detection frameworks in clinical workflows for enhanced accuracy.
Monitoring & Follow-up
- Regularly evaluate the performance of AI models against new datasets.
Risks
- Consider limitations of automated systems in complex cases with low-contrast lesions.
Patient & Prescribing Data
Individuals with suspected brain tumors requiring MRI imaging.
AI frameworks can improve diagnostic accuracy and efficiency in identifying brain tumors.
Clinical Best Practices
- Integrate AI detection tools into routine MRI analysis.
- Ensure continuous training and validation of AI models with diverse datasets.
References
