Implementation of an On-line Domain-Specific Handwriting Recognition Interface for Electronic Medical Records

• Main Author: Dela Cruz, Viktor Mikhael
• Format: text
• Published: Archīum Ateneo 2020
• Subjects: n/a
• Online Access: https://archium.ateneo.edu/theses-dissertations/401
• Institution: Ateneo De Manila University

The Philippine Republic Act 11223 of 2019 sees electronic health systems as a primary tool in the implementation of the UHC Law. This poses a challenge to physicians who need to include the use of electronic medical records in their practice. This study looked into the usability of an EMR and the feasibility of adding handwriting recognition feature as an alternative mode of data entry. Open-source technologies were uti- lized and embedded in a local EMR for testing. Iterative development phases were practiced to achieve human-centered design. The EMR ob- tained a passing score of 70.76 on the System Usability Scale (SUS). Handwriting recognition was augmented with a medical lexicon and on- tology to expand its basis for recognition. Devices used to test the feature included a laptop with a digital pen and tablet attached to it and a tablet PC with a pen. Accuracy, efficiency, and usability tests yielded positive results with decreased word and character error rates as well as faster encoding speed compared to traditional typewriting. The medical lexi- con was shown to increase recognition accuracy. However, the medical ontology did not yield significant improvements to efficiency. Recommen- dations for future studies include optimizing the lexicon and ontology, creating a lexicon of medical abbreviations, studying the user’s end of handwriting input, ensuring readiness of equipment before testing, and remembering the human in design and development.