Method and system for automating clinical data standards

Inventors

Ramanujam, Ilango

Assignees

Symbiance LLC

Abstract

A method and a clinical data standards (CDS) automated system are provided for automating clinical data standards and generating study data tabulation model (SDTM) artifacts required for a regulatory submission process using a machine learning model and a natural language processing (NLP) engine with minimal user intervention. The CDS automated system extracts metadata from multiple raw datasets automatically using NLP and feeds the extracted metadata into the machine learning model; predicts automatic case report form (CRF) annotations on the extracted metadata and records new learnings onto the CDS automated system; maps one or more raw datasets against a target SDTM variable; generates an SDTM statistical analysis system (SAS) code, an SDTM specification, and one or more SDTM datasets; generates a define package; validates the generated define package and the SDTM artifacts generated throughout the entire cycle; and generates validation reports in real time.

Core Innovation

The invention provides a clinical data standards automated system and a method for automating clinical data standards and generating study data tabulation model (SDTM) artifacts required for a regulatory submission process. The method uses a machine learning model and a natural language processing engine to extract metadata from raw datasets automatically, feed the extracted metadata into the machine learning model, and use the extracted metadata to drive downstream SDTM artifact generation.

The system pre-fills the extracted metadata into a blank case report form (CRF) and predicts automatic CRF annotations on the extracted metadata using the machine learning model. The method records new learnings into the clinical data standards automated system with minimal user intervention, based on annotated metadata information and previous learnings that were recorded from manually acquired annotated metadata at the beginning of a clinical trial process.

The invention maps raw datasets against a target SDTM variable using the extracted metadata and information acquired from previous learnings, and then generates SDTM artifacts for a regulatory submission. It predicts derivation logic from a derivation repository based on mapping combinations, study design, and the extracted metadata, generates an SDTM statistical analysis system (SAS) code that is executable outside the clinical data standards automated system in a SAS application, creates a CDISC define package including a define.xml file and a reviewer’s guide, and validates the generated define package and SDTM artifacts with CDISC or United States Food and Drug Administration (FDA)-published validation rules while generating validation reports in real time.

Claims Coverage

The provided partial content includes three independent claims that cover a method, a system, and a non-transitory computer readable storage medium. Across these independent claims, the main inventive features include automated metadata extraction from raw datasets using natural language processing, CRF pre-filling and automated CRF annotation prediction using a machine learning model with previous learnings, mapping to target SDTM variables, generating SDTM SAS code and SDTM artifacts using derivation logic, generating a CDISC define package with define.xml and a reviewer’s guide, and validating the outputs using CDISC or FDA-published validation rules with real-time validation reports.

Across the independent method claim, system claim, and non-transitory storage medium claim, the shared core coverage automates clinical data standards for regulatory submissions by combining NLP-based metadata extraction, machine learning-based CRF pre-filling and annotation prediction with previous learnings and recorded new learnings, mapping to target SDTM variables, derivation logic prediction to generate executable SDTM SAS code and other SDTM artifacts, creating a CDISC define.xml define package with a reviewer’s guide, and validating the generated artifacts using CDISC or FDA-published validation rules with real-time validation reports.

Stated Advantages

Documented Applications