Clinical Language Processing
Language Technology, which is NLP in industrial settings, is one of the critical technologies that HLA has developed. HLA has the ability to extract appropriate content from prose radiology and pathology reports and clinical notes. Reliable information extraction from any type of prose requires special algorithms that understand the statistical behaviour of the language therein. That behaviour has to be learnt from a sample collection of reports. The combination of our linguistic analysis and our statistical algorithms produces tailored information extraction engines for specific needs.
In one Language Technology application, our team of software engineers has designed a new technology for extracting targeted text content from documents and text fields in clinical information systems. These systems can be tailored through trial and error in the workplace under the management and control of the clinical staff. [read more]
Data Analytics is problematic in any clinical setting. However HLA can apply their Clinical Data Analytics technology (CliniDAL) to any clinical information system to support ad hoc questioning in the language of any medical speciality. The CliniDAL package can be installed on any clinical information system or multiple systems and provide combined analysis from the collection of systems. CliniDAL is currently operating in hospital ICU and pathology information systems. [read more]
Language engineering is the use of IT to analyse texts and extract pertinent content. To be able to create working systems of industrial quality there is a large amount of specific software engineering that needs to be present in the infrastructure of any successful product. HLA are international experts in building this infrastructure and have a lead on the rest of the industry in its efficiency at delivering language processing solutions for its clients. It is particularly efficient at producing annotated training sets of linguistic content with its own specially crafted methods for building learning models, enhancing them with human corrections and regenerating the models in an iterative cycle of knowledge discovery and knowledge reuse. [read more]