Handwriting recognition for medical patient records using semantic information from UMLS ontology | |
| Author | Pluert Lertsukittipongsa |
| Call Number | AIT RSPR no.ICT-09-01 |
| Subject(s) | Semantic networks (Information theory) Telecommunication in medicine |
| Note | Submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Information and Communications Technologies, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Series Statement | Research studies project report ; no. ICT-09-01 |
| Abstract | Electronic medical records (EMR) provide a number of advantages over traditional paper-based records, including ease of storage, ease of transmission, the ability to automatically process the data for functions like drug interaction checking, and the support they provide to evidence-based medicine. But the due to the highly structured nature of current EMR systems, physician acceptance remains low. The ideal EMR would provide electronic storage and processing while providing a more natural way for physicians to record information. This research study attempts to make a step in that direction by developing a system for recognizing handwritten text in EMRs. We explore the possibility to using semantic information contained in the UMLS medical ontology to increase the accuracy of a generic handwriting recognition system. Our system works in two phases. First we use the generic system to see which words can be recognized with high confidence. These are then used as keys to extract related concepts from UMLS. Probabilities of these concepts in the lexicon are increased and a second recognition pass is made. Experimental results show that this use of UMLS improves the recognition accuracy from an average of 62% to an average of 73%. While the baseline accuracy is not high, the results provide evidence that the use of such semantic information can significantly increase accuracy of handwriting recognition |
| Year | 2009 |
| Corresponding Series Added Entry | Asian Institute of Technology. Research studies project report ; no. ICT-09-01 |
| Type | Research Study Project Report (RSPR) |
| School | School of Engineering and Technology (SET) |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Information and Communication Technology (ICT) |
| Chairperson(s) | Haddawy, Peter; |
| Examination Committee(s) | Dailey, Matthew;Siriwan Suebnukarn; |
| Scholarship Donor(s) | Royal Thai Government; |
| Degree | Research Studies Project Report (M. Eng.) - Asian Institute of Technology, 2009 |