University developing AI-enhanced imaging system to aid domestic violence victims
George Mason University is developing a system that utilizes different light sources, artificial intelligence and imaging to spot bruises, soft tissue injury and contusions more easily. The research is aimed at helping victims of interpersonal and domestic violence by making it easier for nurses, social workers and law enforcement to identify injuries.
The new tool works five times better than traditional white light detection, and it can be used to spot bruises on all skin tones. The system is part of multidisciplinary research being conducted at the university, focusing on the utilization of alternate light sources to enhance forensic and medical imaging.
In a statement, George Mason University said the research recently earned an anonymous $4.85 million donation after garnering national media attention and praise from domestic violence survivor advocacy groups.
"By improving documentation of physical trauma, we hope to address disparities in clinical care, which can lead to improved healthcare for all as well as legal recourse for all survivors of violence," Katherine Scafide, PhD, RN, associate professor of nursing at George Mason's College of Public Health and co-lead researcher on the project said in the statement. "This gift expands Mason's leadership in developing new technologies to empower vulnerable communities."
For medical and criminal cases alike, developing a remedy requires data. Among domestic violence victims, bruises and soft tissue trauma are the most commonly seen injuries. For some victims, such as the elderly with dementia, assessing damage from an assault can only be done with a physical examination. Accurate documentation improves the likelihood of interventions, such as criminal prosecution.
Development of the system is ongoing, with researchers continuing to gather data to perfect injury detection. The automated AI component, in particular, will rely heavily on robust datasets from diverse skin types and a variety of injuries.
"A vital piece to success is ensuring that any AI tool has access to in-depth data from all skin tones and skin characteristics alongside human expert analysis," added co-lead researcher David Lattanzi, PhD, associate professor in the George Mason College of Engineering.
No timeline has been set for when the tools will be ready for use by the public. For more information on the research, click here.
Chad is an award-winning writer and editor with over 15 years of experience working in media. He has a decade-long professional background in healthcare, working as a writer and in public relations.