Using AI to identify patients in Smart Healthcare

Objectives: The Taiwan Ministry of Health and Welfare advocates for hospitals to implement patient identification processes, and the Joint Commission on Accreditation of Healthcare Organizations in the United States also mandates positive patient identification. However, investigations have found that 16.5% of hospitals have not implemented these processes, impacting patient safety.Methods: Artificial intelligence (AI) excels at image recognition and can automatically verify patients. From April to June 2024, nine wards in a school-affiliated hospital were selected to establish an environment for physiological data transmission using near field communication (NFC) for patient identification and data transfer.Results: Based on the confusion matrix, the recognition precision is 86.7%, the recall is 72.2%, the model accuracy is 92.6%, and Net Promoter Score (NPS) is 84. Each nurse saved an average of 4.8 minutes in measuring vital signs electronically compared to traditional methods (p = .043).Conclusions: Our findings suggest that AI, combined with NFC technology, can play a significant role in shaping the future of healthcare.