For the most secured research, Research-oriented Metadata

HAYCORN provide Only de-identified metadata and analysis results generated within the hospital's secure environment, HRS(Healthcare data Research Suite).

Research-oriented Metadata

We designed a universal data model based on international standards. It provides expanded information on multicenter analytics and can be used in many areas of healthcare.

LLM techonology for deep and wide healthcare data coverage

We apply Artificial Intelligence technology, especially Large-scale Language Model(LLM) to extract and classify key information from medical records and refine into usable data. It provides more effective statistical analysis and machine learning for new drug development and clinical research.

NLP examples on pathology records

Named entity recognition (NER) labeling

Performs an entity annotation task to match the important keywords defined for the Clinical Note Record.

Customized NER modeling

Customizable fine-tuning is performed in minimized time period and specializing in each type of unstructured data make more efficient the information extraction process.

The best workspace
to do healthcare data research

De-identified data

HAYCORN utilizes only de-identified data within a secure hospital area.

Data Result approved by Hospital

HAYCORN provides only data result with metadata processed for statistical and scientific research purposes after hospital approval.

Data security and quality

Security

Data privacy is protected by management standards that meet domestic and international compliance standards.

ISO/IEC 27001 (Information Security Management Systems)
ISO 27799 (Health Informatics)
ISO/IEC 27017
ISO/IEC 27018
HIPAA (Health Insurance Portability and Accountability Act)
CSAP SaaS (Cloud Security Assurance Program SaaS)

Quality

Achieve the highest level of data reliability through systematic data quality control, data standardization, and metadata management.

More detail for Data Coverage

By analyzing patient, disease, and treatment data of distribution across populations, you can identify disease occurrence patterns, differences in treatment response, and progression. After service inquiry, check the distribution of the top 50 codes for each major clinical domain and the occurrence of data you are particularly curious about.