An artificial intelligence tool is now capable of identifying individuals at high risk of intimate partner violence (IPV) years before they seek help, offering a potential breakthrough in early intervention and prevention. Developed by researchers in the United States, the AI leverages hospital data to detect patterns indicative of abuse, even when victims remain silent.
The Problem with Current Screening
Traditional domestic abuse screening relies on direct questioning by healthcare professionals. However, many victims never disclose abuse due to fear, stigma, or safety concerns, leading to underreporting and delayed intervention. According to the European Commission, 18% of women with partners have experienced physical or sexual violence at some point in their lives. This demonstrates the scale of the issue and the limitations of current detection methods.
How the AI Works
Researchers trained three machine learning models using data from nearly 850 women with confirmed IPV cases and a control group of over 5,200 patients. The models analyzed:
- Structured hospital data: Age, medical history, and standard patient information.
- Unstructured medical notes: Doctors’ observations and radiology reports.
- A combination of both data types.
The combined model proved most accurate, correctly identifying risk in 88% of cases. Crucially, the AI could flag potential abuse over three years before patients entered formal intervention programs. By analyzing patterns of physical trauma and comparing them to confirmed abuse cases, the tool offers an early warning system for healthcare providers.
What This Means for Public Health
“This clinical decision support tool could make a significant impact on prediction and prevention of intimate partner violence,” says Qi Duan, program director at the National Institute of Biomedical Imaging and Bioengineering. The AI doesn’t replace clinical judgment; instead, it provides a signal that may encourage healthcare professionals to approach patients with greater sensitivity and offer support.
Future Implementation
Researchers plan to integrate the technology into electronic medical record systems for real-time assessments during routine care. The goal is to shift from reactive intervention to proactive risk recognition, leveraging existing healthcare data to protect vulnerable individuals.
This AI tool represents a major step towards addressing a widespread yet often hidden public health issue. By detecting risk earlier, healthcare systems can potentially save lives and reduce the long-term physical and psychological harm caused by intimate partner violence.
