Results for ""
Problem / Objective
The company faced the challenge of converting the historical documents into structured and searchable digital formats that could be used for analysis. Manual conversion was time consuming, costly and lacked accuracy, and the company needed a solution that could be implemented quickly.
Solution / Approach
The company chose to use Data Foundry’s DFDigitalize AI platform that utilized advance Optical Character Recognition(OCR) technology to digitize documents such as PDFs and images, to machine readable structured texts and made the data accessible for reporting and further analysis. The platform achieved an accuracy rate of 90% in converting over 90 thousand historical documents pertaining to clinical trials conducted since 1980s. The digitized documents were secured by a digital time-stamp in order to make them verifiable and establish the integrity and authenticity of the data over an extended period of time. Additionally, the platform used Named Entity Recognition(NER) capabilities for extracting custom color-coded entities. The platform’s NER was able to extract a wide range of information that was relevant to the company's clinical trial design and research efforts. The platform was able to extract insights for identifying the most effective trial designs such as, demographic and inclusion criteria, dosage and administration that were successful in previous trials, and other key factors that could be useful for future trials. The platform also had a human-in-the-loop verification feature where errors and ambiguities were flagged for review and correction by human experts. The ability to extract specific entities from unstructured text using NER was a key factor in the success of the platform in digitizing the company's historical clinical trial documents, and in providing valuable insights for the company's clinical trial research efforts. Through this extraction, the company gained better insight into the data, which was then fed into the platform's analytics module for deeper analysis and actionable insights.
Impact / Implementation
With DF Digitalize AI, the company was able to digitize their historical clinical trial documents and data five times faster than manual methods and with greater accuracy. Utilizing NER capabilities, the company was able to extract crucial information, such as patient demographics, inclusion and exclusion criteria, dosage and administration, and results, which was used to optimize trial design, identify new clinical trial candidates, and potentially improve patient outcomes. The platform allowed validation of OCR and NER output through a user-friendly interface enabling effort savings of 80% compared to manual data extraction and validation. The platform's analytics module provided valuable insights for better trial design, recruitment, site qualification, and identifying new opportunities for clinical trial candidates. Additionally, the solution was quickly implemented in less than 3 months and costed less than alternative solutions available in the market.
