Cancer informatics
for Cancer Centers

Nature Digital Medicine PRISM: Patient Records Interpretation for Semantic clinical trial Matching system using large language models

American Cancer Society and Color Health to Provide Free At-Home Colorectal Cancer Screening in Underserved Rural Communities

Leading Progress Against Cancer

By Drs Karen Knudsen & Othman Laraki.

Conversation with The Cancer Letter: NCI’s new chief data scientist Warren Kibbe tells us about efforts to get “AI-ready” - July 12, 2024

"Proportion and number of cancer cases and deaths attributable to potentially modifiable risk factors in the United States, 2019" American Cancer Society

The UCSF Division of Clinical Informatics and Digital Transformation (DoC-IT) and UCSF Health have received a $5 million gift from Ken and Kathy Hao to develop a cutting-edge, real-time, continuous, and automated artificial intelligence (AI) monitoring platform for clinical care. The Impact Monitoring Platform for AI in Clinical Care (IMPACC) aims to bridge the gap between the rapid evolution of AI technologies used by clinicians and the essential need for robust, ongoing assessment of their efficacy, safety, and equity. Julia Adler-Milstein, PhD , chief of the UCSF Division of Clinical Informatics and Digital Transformation (DoC-IT), and Sara Murray, MD, MAS , chief health AI officer at UCSF Health, will lead the pioneering collaboration. “This philanthropic gift is transformative in many ways,” said Adler-Milstein. “It comes at a critical juncture as the healthcare industry more broadly integrates AI into clinical practice. Through IMPACC and this collaborative effort, we are poised to improve patient care at UCSF while advancing the science of how to assess AI tools in real-world use.” Currently, the healthcare field lacks established protocols for ongoing AI monitoring, leading to risks of adverse outcomes for patients and healthcare providers that go undetected. While assessments are conducted to determine the suitability of new AI technologies for safe integration into clinical environments before deployment, once they are deployed, health systems need a way to promptly identify any issues in their real-world performance. IMPACC will fill this urgent need by shifting from planned, periodic, manual monitoring of a focused set of measures to real-time, continuous, automated, and longitudinal monitoring across a broad measure set with specified criteria for escalation to human review and intervention.  Full Article: https://docit.ucsf.edu/news/ucsf-and-ucsf-health-receive-pivotal-donation-support-first-continuous-ai-monitoring-platform

I’m most proud of how my colleagues and I evolved CBIIT’s focus and activities to enable NCI’s scientific mission. My approach had always been to collaborate with colleagues from various NCI divisions, offices, and centers (DOCs). In doing this, we understood each other’s specific needs and ensured that our work together aligned with those needs. This wasn’t always CBIIT’s approach though, and while the change started before I became director, creating the necessary collaborations across the institute and developing the projects to support our scientific partners took consistent and intentional focus. CBIIT has been working closely with the other NCI DOCs for years, supporting individual projects and concerted efforts in major programs (such as the NCI Cancer Research Data Commons [CRDC]), data sharing, and NCI Intramural Research Program data management. In hindsight, I might have tried to initiate this outreach and move these relationships forward more broadly and even more quickly. I also would have ensured that we invested more time and some dedicated funding in innovation. To Article: https://datascience.cancer.gov/news-events/blog/dr-tony-kerlavage-reflects-his-time-nci-cbiit

The aims of our case-control study were (1) to develop an automated 3-dimensional (3D) Convolutional Neural Network (CNN) for detection of pancreatic ductal adenocarcinoma (PDA) on diagnostic computed tomography scans (CTs), (2) evaluate its generalizability on multi-institutional public data sets, (3) its utility as a potential screening tool using a simulated cohort with high pretest probability, and (4) its ability to detect visually occult preinvasive cancer on prediagnostic CTs.