| Core Technologies |
|
We have an extensive set of research and development plans for improving our engagement with the hospital sector that will bring new technologies to enhance productivity, gain greater and easier access to data to support clinical and translational research, and provide sophisticated language processing. We have commitments to collaborate with the local health community and high profile overseas organisations. The HITRL has three key technologies that it is developing for use in hospitals: 1. High Accuracy intelligent search of clinical notes (ICNS).
1. RPAH:ICU - High Accuracy Search and Information Extraction (ICNS). In an ICU patients are cared for by many different staff and specialists. Each needs to know exactly what processes the others have initiated before they intervene. This usually means reading many pages of clinical notes within a busy schedule leading to missing key interventions and so risking patient safety. A high accuracy search and extraction engine prototype has been installed at the ICU. It will lead to better clinical decision making and increased productivity producing increased patient safety and reduced hospitalisation times. The current software gives basic functionality but needs to be significantly improved in its natural language processing (NLP) functions. 2. RPAH:ED - Trauma Resuscitation - Trauma Information Systems generated by GCIMS The process of resuscitation for trauma cases requires clinical staff to act quickly and accurately. Currently they use paper methods to record the details of cases due to the inability of the resident clinical information to deliver real-time service on the content important to Trauma Care. We are using the Generative Clinical Information Management System (GCIMS) to create a Trauma Information System (TIS) that has a new range of features not available in classical information systems. The new functions are a user alterable interface and data capture process, an underpinning medical terminology to ensure interoperability with other clinical ISs, a completely generalised data analytics language (CliniDAL) that enables staff to ask any ad hoc question in their own medical dialect, and test statistically any hypothesis about data. In effect they can do their research while they do their rounds. The generation of the TIS is complete and it is in beta test mode at the hospital with the data entry of past cases. Future development requires the addition of CliniDAL to the system, and extensive testing of its function prior to commissioning it for service. The TIS is a candidate technology for use in all Trauma units in the country. 3. Clinical Data Analytics Language (CliniDAL) - SWAPS: Anatomical Pathology Data Warehouse - All clinical data is required by law to be retained for up to 20 years hence the process of CISs decaying and becoming moribund creates a serious data rescue task for hospitals. We have been engaged by a number of institutions to perform this rescue task. However subsequently we have advocated that the resulting data warehouse become the basis of the research activities of the clinical unit, and supplied CLINIDAL to support this function. CliniDAL is a language that permits the expression of any ad hoc query that is answerable from the data stores. It allows for restricted natural language input in the physician's specialism dialect. It has an internal map between any input language and the stored data so that it can retrieve the correct answers. It has an NLP function, that uses the high accuracy search tools, so as to enable it to search for answers in the text notes written by the physicians. It has statistical functions so that it can give probabilistic evaluations to hypotheses that are posed to it. CliniDAL is in use at the RPAH:ICU and SWAPS and can be bolted on to any CIS due to its design architecture. It is the query front end for the GCIMS architecture.
Other Collaborating Hospital Departments RPAH:Anatomical Pathology - High accuracy NLP Concord Hospital:ED - High accuracy NLP Westmead Hospital:ED - Clinical workflow and information systems design Westmead Hospital:Cardiology - Rescuing moribund information system Blacktown-MtDruitt Hospital:General Wards - Multidisciplinary Medical Team Information System using GCIMS Prince of Wales: South East Area Laboratory Service(SEALS) - Rescuing moribund information systems Liverpool Hospital:SWAPS - Rescuing moribund information system Childrens Hospital Westmead: Paediatric ICU - CliniDAL
|
