The Eurasian Patent Office has granted patent No. 050968 for an innovative method for assessing the reliability of medical image analysis using eye-tracking technology and neural networks.
Patent owner: Innopolis University (Russia)
What is the invention about?
The system is designed to automatically assess the physicians’ performance when analyzing medical images such as X-rays, CT scans, and MRIs.
Its main objective is to eliminate diagnostic errors caused by human factors, including fatigue, lack of experience, or loss of concentration.
How does it work?
- The eye tracker monitors the doctor’s eye movements in real time.
- A heat map highlights areas of greatest attention.
- A neural network processes both the image and gaze map.
- AI predicts the likelihood of error and the realibility of the diagnosis.
Technical Result
- Fully automation of diagnostic quality control
- Increased accuracy and objectivity of assessments
- Reduction of false alarms
- Option to obtain a “second opinion” if necessary
Practical Application
The invention will help improve the reliability of diagnostics and optimize the work of radiology departments in medical institutions.
