AI-powered technology can access the 90% of medical records (unstructured data) untouched by other systems
Did you know that nearly nine out of ten clinical trials suffer significant delays, slowing down the clinical innovation they aim to support. Trial criteria have exploded in size and complexity over the past decade. They require in-depth review of patients’ medical records, which is currently a slow, manual process, since traditional software cannot analyze the 90% of the record that is “unstructured,” like physician notes, pathology reports, and images.
AI, natural language processing, and machine learning for the win
Applying artificial intelligence (AI) and natural language processing (NLP) to medical records finds more, better-matching patients for clinical trials exponentially faster. AI tools can extract all relevant clinical features from a patient’s records, such as symptoms, diagnoses, treatments, genomics, lifestyle data, and much, much more. From these data points, software can create a multidimensional clinical vector of each patient, which can be matched against complex trial eligibility criteria to find patients, cohorts, and populations “in minutes, rather than months.” This process turns document-based medical records into rich patient graphs that can be easily read and compared by machine.
Advanced AI techniques, like link analysis conducted across patients, discover hidden relationships in the data or detect “fuzzy” medical conditions that are not diagnosed or mentioned by name, but are only hinted at, in an electronic health record. Machine learning allows the system to grow faster and smarter with use, so it can continuously suggest more, better matches.
The software at work: non-small cell lung cancer
Our software integrates all of these capabilities, and the results speak for themselves. A major academic medical center in Los Angeles tested our software to find a biomarker in patient’s records for a trial for non-small cell lung cancer. With the help of the medical center’s IT team, the principal investigator had found 23 eligible patients over the previous 12 months. During the subsequent trial of the Deep 6 software, 58 eligible matches were found in mere minutes, which enabled much faster and more precise patient recruitment.
With the integration of clinical trials into their care options, health organizations benefit from better patient outcomes and satisfaction, a reputation for offering cutting-edge treatments and research, attraction and retention of more talent and patients, and significantly increased revenues.
To see for yourself how Deep 6 AI’s patient recruitment software can match patients to clinical trials in minutes, not months, schedule a demo today.