Related Experiment Video
Updated: Sep 13, 2025
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
251
Real-time detection of premalignant cervical lesion using Artificial Intelligence (AI) model in multispectral imaging
A Keerthana1, Arpitha Anantharaju2, P T Mohammed Ansar1
1Department of Electronics and Communication Engineering, Indian Institute of Information Technology, Design and Manufacturing, Kancheepuram, Chennai, India.
Summary
This study introduces an AI model for cervical cancer screening, improving detection of abnormal blood vessels and iodine uptake. The AI-assisted system aims to reduce inter-clinician variability and unnecessary biopsies.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Oncology
Background:
- Cervical cancer is a leading global cancer in women.
- Early diagnosis and prevention are crucial for improved prognosis.
- Current cervical examination accuracy is limited by inter-clinician variability.
Purpose of the Study:
- Develop a real-time AI model integrated with a multispectral imaging system (GynoSight).
- Assist clinicians in detecting abnormal blood vessels, acetowhite uptake, and iodine-negative regions during cervix screening.
- Reduce inter-clinician variability and unnecessary biopsies.
Main Methods:
- Utilized a dataset of 609 colposcopy images (normal saline, 3% acetic acid, Lugol's Iodine).
- Implemented an EfficientNet architecture AI model for identifying atypical blood vessels and abnormal uptake regions.
- Employed a transvaginal imaging probe (GynoSight) with a trained TensorFlow model for real-time detection.
Main Results:
- Achieved 86.67% accuracy for Iodine-negative region identification (F1 score 90%, mAP 91.1%).
- Improved acetic acid uptake detection accuracy from 57% to 85.71% with a segmentation algorithm.
- AI system's detection of atypical blood vessels, acetowhite regions, and Iodine-negative uptake compared with clinician interpretation and biopsy results.
Conclusions:
- An AI-assisted real-time detection system can aid clinicians during cervical screening.
- The system has the potential to decrease the number of unnecessary biopsies.
- AI integration can help standardize cervical examination interpretation and reduce variability.
