Quality Assessment for Health Applications (QAH)
The QAH group focuses on the quality assessment and visual attention studies for both medical contents and health applications. The main goals are as follows:
- Assemble all the existing publicly accessible databases on medical image/video quality
- Define suitable experimental methodologies/recommendations/guidelines with medical experts on diagnostic/surgery tasks
- Evaluate/develop existing/new quality metrics/models for medical images/videos
- Study the quality requirements and Quality of Experience (QoE) in the context of telemedicine and other telehealth services
- Assemble all the existing publicly accessible eye-tracking databases, related to health applications.
- Provide methodologies, recommendations, and guidelines for subjective tests including different eye tracking studies, related to health applications.
- Evaluate/develop existing/new visual attention prediction models.
In this project we propose to focus on health application. Nowadays, we understand that psychological well-being is as important as physical well being. Moreover substantial studies show the influence of mental health on physical health. Through this project we want to address two issues related to health. The first one aims to help medical experts to gain a better medical (diagnostic, surgery, etc.) task performance or to spend less time in the acquisition and interpretation step with the same clinical reliability. The second aims to help psychological specialists to understand and provide solutions to people, especially neuroatypical ones. Indeed, the learning and work environments have to support neurodiversity.
Medical field: Task-based approaches have been popular in the medical image quality assessment domain since 1993, named “model observer.” The underlying paradigm is to quantify the quality of a particular image by its effectiveness with respect to its intended task. A system or algorithm that enables medical experts to perform the diagnostic task in a reliable way in less time. The gain could be to shorten the visualization and interpretation of the results as well as the duration of the examination and thus the radiation exposure. Because of the methodology limitations, only detection and localization tasks have been addressed in the literature. Nowadays, with the development of deep learning and image-text retrieval technologies, the characterization task may be modeled. Thus in this project, we want to promote the task-based approaches (make a bridge between VQEG and the medical image quality community) for medical images and explore this diagnostic task.
Psychological scope: Visual attention study through eye-tracking devices allows learning and understanding human attention and some aspects of human behavior. Whether the subject is neurotypical or neuroatypical, the eye tracking devices are suitable for use without disturbing the environment. Eye-trackers can measure the eye positions and eye movements, the captured data can also serve as the ground truth for saliency prediction models development. However, to the best of our knowledge there are no guidelines or recommendations for the use of eye-trackers in experiments with neuroatypical subjects. Thus we propose to expand our projects for the benefit of mental health and neurodiversity by sharing studies and feedback of experiments conducted using eye-trackers.
Neurodiversity describes the idea that people experience and interact with the world around them in many different ways; there is no one "right" way of thinking, learning, and behaving, and differences are not viewed as deficits. Taken from "What is Neurodiversity?"
Want to join us?
1) Subscribe to QAH : firstname.lastname@example.org
2) And fill this Google Doc
A summary of existing subjective quality assessment for medical images (with discussion on merits and drawbacks of the methodologies, as well as recommendations) is done, disseminated as a publication:
- Lévêque, M. Outtas, H. Liu, L. Zhang. “Comparative study of the methodologies used for subjective medical image quality assessment”. Physics in Medicine and Biology; July 2021; 66(15), 15TR02.
A summary of existing objective quality assessment approaches (including both task-based methods and visual quality-based methods) is done, will be disseminated as a publication:
- Rodrigues, L. Lévêque, J. Gutiérrez, H. Jebbari, M. Outtas, L Zhang, A. Chetouani, S. Al-Juboori, M. G. Martini, A. M.G. Pinheiro, “Objective Quality Assessment of Medical Images and Videos: Review and Challenges”.
Questions should be addressed to the QAH Chair Lu Zhang (email@example.com), Vice Chair Meriem Outtas (Meriem.Outtas@insa-rennes.fr).