Lewes, Del. — October 14, 2025 — Leads & Copy — John Snow Labs has announced new capabilities designed to fully automate oncology patient registries, improving accuracy, scalability, and usability for clinical teams. The announcement was made during a keynote session at the Applied AI Summit, taking place online from October 14-16.
Oncology teams rely on high-quality registries to power real-world evidence generation, cohort selection, clinical trial design, outcomes research, and operational planning. However, creating such registries remains one of the most complex challenges in clinical informatics because critical data is rarely explicitly documented and must instead be inferred from thousands of pages of pathology, radiology, genomics, and clinical notes.
Manual abstraction is slow, costly, and increasingly unsustainable. According to the National Program of Cancer Registries, building oncology registries by hand takes an average of 2 hours per cancer case, with complex patients requiring several days of work. One full-time tumor registrar can typically process only 6–10 cases per day, meaning 1,000 patient cases consume over 2,000 hours of labor. The timeliness of cancer reporting is a nationwide concern, with the NPCR aiming for 90% of cases diagnosed within the past 12 months to be reported to central cancer registries, yet only 14% of registries consistently meet this standard. Staffing shortages and growing ePath report volumes exacerbate delays, creating backlogs that can stretch up to 7 months for pathology report processing.
John Snow Labs’ multimodal AI approach addresses these operational and technological challenges directly via multimodal AI and NLP for triage, agentic workflows and auto-consolidation, a registrar-friendly UI with human oversight, and oncology-specific agents. With John Snow Labs, the time to abstract a case drops from 2 hours to 1–2 minutes, representing a 60–100x productivity gain.
“Oncology registrars have long faced the dual challenge of extracting structured data from unstructured sources and consistently applying thousands of pages of guidelines,” said David Talby, CEO, John Snow Labs. “This new capability delivers a practical blueprint to automate patient registries in a way that is accurate, explainable, and production-ready for real-world operations. This is not just about speed – it’s about providing useful and reliable data for public health and research.”
To learn more about automating oncology patient registries, register for the free Applied AI Summit, and tune into our keynote session taking place at 12pm ET on Wednesday, October 15.
Contact:
Gina Devine
Head of Communications
John Snow Labs
gina@johnsnowlabs.com
Source: John Snow Labs
