Related Experiment Video
Updated: May 5, 2026
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Deep learning-based mobile application for efficient eyelid tumor recognition in clinical images.
Shiqi Hui1, Jing Xie2, Li Dong1
1Beijing Tongren Eye Center, Beijing Key Laboratory of Intraocular Tumor Diagnosis and Treatment, Beijing Ophthalmology & Visual Sciences Key Lab, Medical Artificial Intelligence Research and Verification Key Laboratory of the Ministry of Industry and Information Technology, Beijing Tongren Hospital, Capital Medical University, Beijing, China.
A new mobile app uses deep learning for self-diagnosing eyelid tumors, offering accurate detection and monitoring. This technology improves patient care by reducing hospital visits for those with poor medical conditions.
Area of Science:
- Ophthalmology
- Artificial Intelligence
- Medical Imaging
Background:
- Eyelid tumor detection and monitoring are critical but often require frequent hospital visits.
- Frequent hospital visits pose a significant burden for patients with poor medical conditions.
Purpose of the Study:
- To validate a novel deep learning-based mobile application for self-diagnosis of eyelid tumors.
- To develop an improved health support system for patients with eyelid tumors.
Main Methods:
- Utilized YOLOv5 and Efficient-Net v2-B architectures for model development.
- Trained the model on 1195 preprocessed clinical ocular photographs and biopsy results.
- Converted the best-performing model into a smartphone application for external validation.
Main Results:
- Achieved 0.921 accuracy in triple classification (benign/malignant eyelid tumors or normal eye).
- Demonstrated performance superior to general physicians, resident doctors, and ophthalmology specialists.
- The application features a user-friendly interface and treatment recommendations.
Conclusions:
- The Intelligent Eyelid Tumor Screening application provides preliminary evidence for recognizing eyelid tumors.
- The app can be utilized by healthcare professionals, patients, and caregivers for detection and monitoring.
- This technology offers a more accessible and efficient approach to eyelid tumor management.
