Medical prescriptions often contain handwritten and unstructured information, making accurate interpretation difficult and leading to potential medication errors. Existing healthcare digitisation systems primarily focus on basic text extraction and fail to incorporate intelligent clinical analysis, such as drug interaction detection and medication scheduling. This paper proposes an intelligent AI-powered Clinical Prescription Recognition and Medication Management System using multimodal Optical Character Recognition and transformer-based Natural Language Processing with Role-Based Access Control. The system preprocesses prescription images through enhancement and normalisation techniques, converting them into structured medical representations, including drug names, dosage, frequency, and duration, using multi-entity extraction. Transformer-based models accurately interpret prescription text, while an integrated interaction analysis module identifies potential drug conflicts using knowledge-driven rules. A scheduling engine generates personalised medication plans based on extracted instructions and clinical constraints. The system further provides secure role-based access, enabling patients to receive medication reminders and administrators to monitor prescription history, interaction alerts, and adherence patterns. The proposed framework improves accuracy, safety, and usability compared to traditional OCR-based approaches, enabling intelligent and automated healthcare workflow management.
Prescription Recognition, OCR, Healthcare AI, Multi-Entity Extraction, Drug Interaction Detection, Transformer Models, Medication Scheduling, Clinical NLP.
. AI-Based Clinical Prescription Recognition and Medication Management Using Multimodal OCR and Transformer Models: A Survey. Indian Journal of Modern Research and Reviews. 2026; 4(4):40-45
Download PDF
