Association between Visual Gaze Patterns and Adenoma Detection Rate

Project: Research project

Project Details


Association between Visual Gaze Patterns and Adenoma Detection Rate Association between Visual Gaze Patterns and Adenoma Detection Rate Colonoscopy screening programs are the hallmark in the diagnosis and prevention of colorectal cancer (CRC). However, colonoscopy is far from being a perfect technique given that polyps are often missed. The risk of interval cancer has been linked to the endoscopists adenoma detection rate (ADR). Therefore, the ADR has been validated as a measure of colonoscopy quality. The ADR varies widely among endoscopists reflecting individual differences in the colonoscopy performance. It has been demonstrated that certain technical factors such as slowing the withdrawal time and careful inspection behind of folds and flexures can be modified to increase the quality of the colonoscopy. We hypothesize that physician variables such as their visual gaze patterns may be fundamental to high quality colonoscopy and indirectly to effective colon cancer prevention. A preliminary study with 11 endoscopists watching offline colonoscopy videos has suggested that visual gaze tracking data might be useful for identifying such factors. AIMS: Our specific aim is to capture and analyze visual gaze tracking data from a large sample of endoscopists to develop metrics that will allow us to identify specific visual gaze patterns that correlate with higher ADRs. METHODS: We will invite all endoscopists from the Gastroenterology Division of Mayo Clinic Rochester, Arizona and Florida with previously known ADRs (n=60). All participants will watch 5 standardized videos while their visual gaze pattern is recorded with the Tobii 1750 desktop eye tracker. The videos will consist of approximately five, 2-3 minute uncompressed, high definition segments extracted from normal and abnormal colonoscopies which containing multiple polyps of different size and shape, including cofounding factors such as areas of deficient prep. The endoscopists will be instructed to verbalize any abnormality that they recognize in the videos. For each individual we will assess: gaze pattern (location and characteristics of the viewer scan path along the screen), visual attention (time and number of fixations, changes in pupil size, blinking, length and number of saccades for each abnormality), time of reaction (time since the abnormality appeared at the screen until the gaze is fixed on it), time until diagnosis (time since the gaze is fixed on the abnormality until the diagnosis is verbalized); and diagnostic rate (number of polyps detected out of the total). These parameters will be compared among endoscopists of different experience and correlated with the ADRs.
Effective start/end date8/1/127/31/13


  • Mayo Clinic Jacksonville: $20,000.00


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.