Coronavirus COVID-19 Updates: uc.edu/publichealth
Search By:
How will the explosion of AI impact cancer research and care? Last month, the University of Cincinnati Cancer Center held the tenth annual Charge Against Cancer Symposium on “Leveraging Artificial Intelligence and Data Sciences to Conquer Cancer” to tackle this topic. Local and national speakers gave lectures on leveraging data in healthcare, advancing precision medicine using artificial intelligence, using and leveraging data in professional sports and how it applies to medicine, AI-enabled imaging, developing data networks, and the growing use and influence of Chat GPT.
On Friday, August 25th, the University of Cincinnati Cancer Center held the tenth annual Charge Against Cancer Retreat. With this year’s theme being Leveraging Artificial Intelligence and Data Sciences to Conquer Cancer, the executive committee was charged with developing an event to help place the University of Cincinnati Cancer Center at the forefront of implementing AI to ensure this powerful technological advance is used to its full capacity.
The Charge Against Cancer was developed by Cancer Center co-director William Barrett, MD, nearly a decade ago. “We started the Charge Against Cancer event nine years ago to, on an annual basis, focus on important and often novel aspects of cancer prevention, detection, treatment and/or recovery. These events have brought people together from varied backgrounds and disciplines and have stimulated original ideas and collaborations. These events have each been provocative, inspiring and team building,” said Barrett. Charge Against Cancer Executive Committee:
Thomas J. Herzog, MD, who co-hosted the event alongside Dr. Barrett, recognized artificial intelligence's impact on cancer research and care.
“Artificial intelligence has far-reaching implications for cancer research and clinical care,” Herzog said. “Research has become incredibly costly and inefficient, especially with large clinical trials. AI may allow us to make these trials more efficient, conserving resources while decreasing inefficiencies. Furthermore, clinical care could change substantially not only in diagnosing disease but also in evaluating options for care through imaging and assessment of biomarkers. Building AI networks could facilitate the development of larger databases that could be valuable in generating new hypotheses for novel treatment options.”
Three external speakers and four speakers from the University of Cincinnati were featured during this year’s Charge Against Cancer Retreat.
Keynote Speakers:
Upon the conclusion of presentations, the attendees were divided into seven working
groups to develop a project idea to employ artificial intelligence and/or data science to positively impact one or more of the four areas: cancer prevention, detection, treatment, and/or recovery. Once each team had pitched their ideas, the winning team was determined through a voting process and awarded $75,000 to transform their idea into a reality.
Brett M. Harnett, MS-IS, principal investigator of the winning team’s project, is not only very grateful for the opportunity but also ecstatic to begin actualizing the idea.
“Our idea – “Get PAIRED (Pancreatic AI Repository for Early Detection)” – is to use the relatively new data transport standard called Fast Health Interoperability Resources (FHIR) to automate the extraction of clinical data from two to three sample community hospitals who often refer cancer patients to UC Health,” Harnett explained. “The aim is to create an automatic, centralized regional registry that will provide high-quality and high-quantity data to enable machine-learning methods for the detection of early predictors of pancreatic cancer.”
The “Get PAIRED” data model will be hosted on a UC Health/BMI server infrastructure and designed to leverage an established clinical data standard not yet implemented at UC Health supporting ingestion and transport standards such as HL7 FHIR APIs.
“The idea is to implement artificial intelligence in a way that enhances human insight and allows early intervention to be considered,” he said. “For instance, data from patients who visit one of the identified community hospitals and match the criteria derived from the validated training models of the algorithm developed from verified pancreatic cancer diagnosis at UC Health will be sent to the “Get PAIRED” pipeline for AI assessment and reporting. The system will include basic descriptive analytics, such as aggregating data into visualizations, without additional statistical application. To meet the project's aims, a dashboard will deliver the AI-driven outcomes/predictions where human insight can be enhanced, and early intervention can be considered.”
“Get PAIRED (Pancreatic AI Repository for Early Detection)” Contributors:
231 Albert Sabin Way, Suite 2005Cincinnati, OH 45267Phone: 513-558-2177Fax: 513-558-2666