Researchers in the US have developed an AI system to identify patients at risk of intimate partner violence years before they officially seek help or enter intervention programs. The system was trained primarily on hospital data, including medical history and written clinical notes.
The dynamic combination of both system results achieved an 88% accuracy rate in identifying risk. The technology detects subtle patterns of physical trauma and medical history that resemble confirmed abuse cases, allowing it to flag potential victims who may not disclose abuse due to fear or stigma.
The study published in Nature, found that the tool could flag potential abuse more than three years before patients typically entered domestic abuse support programs. The tool is intended to act as a clinical decision support system rather than a diagnostic tool. It provides a signal to healthcare professionals so they can approach the topic sensitively and offer support, but it does not replace a doctor’s judgement.
In this connection Qi Duan, the program director of the division of health informatics technologies at the US National Institute of Health’s Biomedical Imaging and Bioengineering said: “This clinical decision support tool could make a significant impact on prediction and prevention of intimate partner violence."
Nonetheless, researchers plan to integrate the AI into electronic medical record systems, providing hospitals with real-time risk assessments during routine patient care.