Researchers created an AI tool leveraging clinical and radiology data to accurately predict patients at risk of intimate partner violence (IPV).
Key Details
- 1NIH-funded study trained AI models on structured clinical data and unstructured text, including radiology reports.
- 2Models evaluated data from nearly 850 IPV-affected female patients and 5,200 matched controls.
- 3Three models were developed: structured data only, unstructured text only, and multimodal (fusion) combining both.
- 4The fusion model achieved 88% accuracy and greater stability, outperforming single-modality models.
- 5Both tabular and fusion models predicted IPV risk an average of three years before intervention center enrollment.
- 6Emergency radiologist Bhati Khurana was senior author; further clinical integration is planned via EMR tools.
Why It Matters
Proactive AI-based IPV risk identification can significantly improve early intervention and patient outcomes, with radiology playing a key role in detection. This approach illustrates how imaging and informatics can address public health challenges beyond traditional diagnostics.
Source
EurekAlert
Related News
•EurekAlert
AI Accelerates Radiopharmaceuticals, Boosts Personalized Dosimetry in Cancer
Machine learning is driving advancements in radiopharmaceutical drug discovery and optimizing patient-specific dosimetry for precision cancer therapy.
•EurekAlert
Physicians Overly Trust Erroneous AI, Ignore Contradictory Evidence
Physicians tend to trust incorrect AI advice, even when evidence contradicts it, suggesting risks in clinical decision-making with AI tools.
•EurekAlert
Concerns Raised Over Unverified Datasets in AI Health Prediction Models
A new study finds widely used AI health prediction models are built on datasets with unverifiable origins, raising safety and validity concerns.