G L Yang1, Y F Tan, S C Loh, C C T Lim. 1. Biomedical Imaging Laboratory, Agency for Science Technology and Research Singapore, 30 Biopolis Street, #07-01 Matrix, Singapore 138671.
Abstract
INTRODUCTION: Disease registries and databases form an important component of research in clinical medicine, and can be useful to support retrospective studies and prospective clinical trials. However, analysis of radiological imaging databases has not been explored: imaging and clinical data often exist as separate silos of information, even in modern digital-enabled hospitals in Singapore. We describe a computerised method for creating a radiological research database using data from the picture archive and communication system (PACS) and hospital information system (HIS). METHODS: Using a relational database and Java programming language, we created the neuroradiology imaging database (NRID). A web-interface for keyword searches were tested with the clinical data from PACS of a tertiary referral hospital for neurological diseases. Keyword and wildcard searches were conducted for various brain neoplasms and compared to HIS discharge diagnosis. RESULTS: The NRID was deployed successfully and keyword search could be completed in real time. Lists of patients with meningioma, oligodendroglial tumour, neurocytomas, cerebral abscess, and neurocysticercosis could be exported and compared with the HIS discharge diagnosis. Patients with neurological diseases could be obtained by manually combining lists. CONCLUSION: An imaging database can be created using clinical PACS data, which can enable keyword search functions to support brain tumour research. Radiological databases can help support clinical research, but further work needs to be done in order to take full advantage of the potential of digital health information.
INTRODUCTION: Disease registries and databases form an important component of research in clinical medicine, and can be useful to support retrospective studies and prospective clinical trials. However, analysis of radiological imaging databases has not been explored: imaging and clinical data often exist as separate silos of information, even in modern digital-enabled hospitals in Singapore. We describe a computerised method for creating a radiological research database using data from the picture archive and communication system (PACS) and hospital information system (HIS). METHODS: Using a relational database and Java programming language, we created the neuroradiology imaging database (NRID). A web-interface for keyword searches were tested with the clinical data from PACS of a tertiary referral hospital for neurological diseases. Keyword and wildcard searches were conducted for various brain neoplasms and compared to HIS discharge diagnosis. RESULTS: The NRID was deployed successfully and keyword search could be completed in real time. Lists of patients with meningioma, oligodendroglial tumour, neurocytomas, cerebral abscess, and neurocysticercosis could be exported and compared with the HIS discharge diagnosis. Patients with neurological diseases could be obtained by manually combining lists. CONCLUSION: An imaging database can be created using clinical PACS data, which can enable keyword search functions to support brain tumour research. Radiological databases can help support clinical research, but further work needs to be done in order to take full advantage of the potential of digital health information.
Authors: Tiina Rajala; Sami Savio; Jarkko Penttinen; Prasun Dastidar; Mika Kähönen; Hannu Eskola; Risto Miettunen; Väinö Turjanmaa; Ritva Järvenpää Journal: J Digit Imaging Date: 2011-10 Impact factor: 4.056