NAVIGATION
    • Home
    • NLP-AS-A-SERVICE

Automatically Unlock the Potential of Your Oncology Data

HIGH PERFORMANCE. HIGH ACCURACY. HIGH AUTOMATION.

NLP-AS-A-SERVICE

Up to 80% of data needed for cancer-related analysis and reporting is unstructured in free-text formats, making it difficult to access without manual curation…, until now!

why-inspirata-315

Up to 80 percent of cancer-related patient information captured and stored in EHRs and other health IT systems is extremely difficult to access and render actionable, whether it’s for quality analysis, clinical decision support, research or reporting. Often, the most important insights about a patient’s cancer is written in free-text narrative formats (e.g., Word, Excel, PDF, HL7 V2.x, HL7 V3 CDA, XML, RTF, ASCII text); or, it’s contained in faxed/scanned documents (like genomic profiling reports). Most institutions that treat cancer, therefore, must resort to costly, time-consuming and error-prone manual chart review and abstraction.

Inspirata has developed a transformative solution with a proven track record for successfully unlocking the potential of oncology data in healthcare. Using sophisticated Natural Language Processing (NLP), Inspirata’s NLP-as-a-Service has been validated during large-scale pilots* where it successfully completed on-demand processing and abstraction of structured, semi-structured and unstructured clinical documents, such as pathology reports, molecular/genomic test results, diagnostic imaging reports and clinical notes.

Pre-built Oncology-Specific Pipeline Offers Fastest Time-to-Value
Applications315

Other NLP solutions available today require considerable manual oversight and custom pipeline development for each deployment. Conversely, Inspirata’s NLP-as-a-Service has a pre-built oncology-specific pipeline that can extract more than 300 discrete data elements out of the box. These data elements are mapped to consistent biomedical terminology using the Unified Medical Language System (UMLS) ontology. Our NLP also employs proprietary business rules and annotation techniques that reliably identify and extract discrete biomedical concepts with context sensitivity for linguistic variations, negation and spatial associations.

Because our NLP pipeline is prebuilt and simply fine-tuned to achieve the desired data extraction from your documents, Inspirata’s NLP-as-a-Service offers the fastest time-to-value solution available today.

Workflow Delivers Highly Accurate Data Extraction

The NLP-as-a-Service workflow is uncomplicated, yet effective. Documents are imported from the EHR or other health IT systems, and the pre-built NLP pipeline performs data element extraction, transforming the required free text into discrete data elements. When complete, the extracted and transformed document is output as an annotated Clinical Data Architecture (CDA) file for easy document exchange, and the discrete data elements are output in a .csv or JSON file, or a custom format that the hospital can then use to perform data analysis and reporting.

Implementation Is Customized to Your Conventions

A rapid, three-step approach guides configuration and deployment for implementation of Inspirata’s NLP-as-a-Service.

1. Assess: An assessment of clinical document types, content, formats and language is conducted to determine conventions and standards used by your hospital and to account for variations in abbreviations, acronyms and writing styles employed by your clinicians.
2. Configure: Inspirata’s NLP pipeline is configured and fine-tuned by our data analytics experts to achieve the desired data extraction based on the document assessment completed in step 1.
3. Deploy: Deployment for this solution can be on premise or in the cloud. Your documents can be accessed through web services or similar APIs, as well as via direct integration with your clinical systems. Output formats include standard CSV files, custom reports or data ingestion into your downstream applications.

 

*Inspirata’s NLP engine has been validated for document classification, case finding and data extraction using more than 50,000 clinical documents provided by Inspirata customers, academic medical centers and cancer centers.

 

Workflow Solution

Learn more about our end-to-end digital pathology workflow solution.

Learn More
White Paper

Read Dr. Metin Gurcan’s White Paper regarding the lessons learned from radiology.

Read Now
Our Partnerships

Learn how Inspirata delivers on the promise of digital pathology through partnerships.

Learn More