Three-Dimensional Visual Patient Based on Electronic Medical Diagnostic Records.
Liehang Shi,Jianyong Sun,Yuanyuan Yang,Tonghui Ling,Mingqing Wang,Yiping Gu,Zhiming Yang,Yanqing Hua,Jianguo Zhang
DOI: https://doi.org/10.1109/jbhi.2017.2702201
IF: 7.7
2018-01-01
IEEE Journal of Biomedical and Health Informatics
Abstract:OBJECTIVE:an innovative concept and method is introduced to use a 3-D anatomical graphic pattern called visual patient (VP) visually to index, represent, and render the medical diagnostic records (MDRs) of a patient, so that a doctor can quickly learn the current and historical medical status of the patient by manipulating VP. The MDRs can be imaging diagnostic reports and DICOM images, laboratory reports and clinical summaries which can have clinical information relating to medical status of human organs or body parts. METHODS:the concept and method included three steps. First, a VP data model called visual index object (VIO) and a VP graphic model called visual anatomic object (VAO) were introduced. Second, a series of processing methods of parsing and extracting key information from MDRs were used to fill the attributes of the VIO model of a patient. Third, a VP system (VPS) was designed to map VIO to VAO, to create a VP instance for each patient. RESULTS:a prototype VPS has been implemented in a simulated hospital PACS/RIS integrated environment. Two evaluation results showed that more than 70% participating radiologists would like to use the VPS in their radiological imaging tasks, and the efficiency of using VPS to review the tested patients' MDRs was 2.24 times higher than that of using PACS/RIS, while the average accuracy by using PACS/RIS was better than that by using VPS; however, this difference was only about 4%. CONCLUSION:the developed VPS can show the medical status of patient organs/sub-organs with 3-D anatomical graphic pattern and will be welcomed by radiologists with better efficiency in reviewing the patients' MDRs and with acceptable accuracy. SIGNIFICANCE:the VP introduces a new way for medical professionals to access and interact with a huge amount of patient records with better efficiency in the big data era.