Book a Demo now +61 2 8915 6242

Cancer Registry Pipeline Services

HLA technology has been developed to analyse unstructured clinical text, using semantics and context to identify a broad range of medical concepts within medical documents. CaRPS is a strategy for automating the flow of documents through cancer registries by Natural Language Processing.

Step 1. Scan paper document
Step 2. OCR to create digital document
Step 3. Validate and correct digital document
Step 4. Case identification as a reportable cancer
Step 5. Abstract pertinent content from the reportable cancer document
Step 6. Codify the abstracted content based on standard coding schemes
Step 7. Inferring cancer stage from extracted content
Step 8. Forward the processing results in an acceptable format

How It Works

Clinical Concepts, Terminologies and Classification: HLA Text Analytics Engine understands clinical text and converts it into clinical concepts defined in terminologies and classification systems such as ICD, LOINC and SNOMED CT.

Automatic Spelling Corrector: HLA’s proprietary spelling corrector is built from manually annotated 90,000+ known clinical misspellings and has an autocorrect accuracy of over 90%.

Grammar Agnostic: HLA Text Analytics Engine deals with a great variety of grammatical expressions, word variations, acronyms, abbreviations and neologisms with comparative ease.

Acronyms and Abbreviations: HLA engines know many thousands of common abbreviations and acronyms, and continues to learn a user’s idiosyncratic usage.

Score and Measures: HLA’s CaRPS has a deep understanding of clinical scores such as the GCS or Apgar score, and can manage a wide range of measurement representations.

Document Structure: HLA understands document structure and searches for content in the appropriate section. For example, a diagnosis in a radiology or pathology report is identified in the Conclusions section.


Faster, more comprehensive coding of patient cases

Search for the widest range of clinical concepts across the organisation

Retrieve historical patient cases that need review and administration for new care strategies

Fast preparation of case histories for meetings, reviews and auditing

Collect information for Policy Development and Validation

Accurate, comprehensive and fast retrieval of patient cohorts

Enhanced tuning of language models for highly targeted retrieval and extraction

Web user access from anywhere, at any time