Assessment of medication self-administration using artificial intelligence

Errors in medication self-administration (MSA) lead to poor treatment adherence, increased hospitalizations and higher healthcare costs. These errors are particularly common when delivery involves devices such as inhalers or insulin pens. We present a contactless unobtrusive artificial intelligence (AI) framework that can detect monitor MSA by analyzing the wireless signals patient’s home, without need for physical contact. The system was developed observing conducted volunteers evaluated comparing its prediction with human annotations. Findings from this study demonstrate our approach automatically patients use their (area under curve (AUC) = 0.992) pens (AUC 0.967), assess whether follow appropriate steps using these 0.952). work shows potential of leveraging AI-based solutions improve safety minimal overhead health professionals. Artificial coupled home sensors alert self-medication an manner.