Houston Daily

A new artificial intelligence system, MedGaze, developed by Hien Van Nguyen, associate professor of electrical and computer engineering at the University of Houston, is set to revolutionize radiology training and diagnostic accuracy. This system emulates the visual interpretation methods used by radiologists when examining chest X-rays.

Nguyen's research, published in Nature Scientific Reports, aims to enhance how machines and future radiologists think by replicating expert attention. "We’re not just trying to guess what a radiologist will do next; we’re helping teach machines and future radiologists how to think more like experts by seeing the world as they do," Nguyen explained.

MedGaze increases hospital efficiency by identifying which cases require more time and effort while improving AI-powered diagnostic systems' accuracy. "MedGaze is a non-invasive, non-interfering software system trained to mimic how expert doctors visually examine chest X-rays," said Nguyen. It analyzes images alongside radiology reports to create a 'Digital Gaze Twin.'

The system learns from thousands of previous eye-tracking sessions where radiologists’ gaze paths were recorded during X-ray interpretations. This enables it to predict where a radiologist is likely to focus when reviewing new images.

In medical imaging, predicting scan paths is crucial for enhancing diagnostic accuracy and efficiency. MedGaze addresses this by modeling fixation sequences significantly longer than those managed by current methods. "Unlike previous computer vision efforts that focus on predicting scan paths based on specific objects or categories, our approach addresses a broader context," Nguyen stated.

The methodology behind MedGaze is adaptable for various imaging modalities beyond chest X-rays, such as MRI and CT scans. "This opens the door to a unified, AI-driven approach for understanding and replicating clinical expertise across the full spectrum of medical imaging," Nguyen noted.

The research team includes UH graduate students Akash Awasthi and Mai-Anh Vu along with professors Carol Wu and Rishi Agrawal from MD Anderson Cancer Center.