2 Clarke Drive
Suite 100
Cranbury, NJ 08512
© 2024 MJH Life Sciences™ and OncLive - Clinical Oncology News, Cancer Expert Insights. All rights reserved.
Amma Asare, MD, PhD on the use of AI for predicting PFS in ovarian cancer
Amma Asare, MD, PhD, gynecologic oncology Fellow, The University of Texas MD Anderson Cancer Center, discusses the investigation of an artificial intelligence (AI) modeling tool intended to predict progression-free survival (PFS) for patients with ovarian cancer.
Asare and colleagues sought to employ laparoscopic imaging and machine-learning techniques to identify patterns associated with PFS outcomes for patients with ovarian cancer. The retrospective study included 99 patients with ovarian cancer who had a PFS that was classified as long, as well as 16 patients with PFS classified as short.
Findings presented at the 2024 SGO Annual Meeting on Women’s Cancer showed that the model predicted 73 of these patients would experience long PFS, whereas 42 would have short PFS. This translated to a sensitivity of 88% and a specificity of 72%.
Asare underscores the novelty and potential of this approach, acknowledging the unfamiliarity physicians may still have with AI and its potential uses within practice. Collaborative efforts are paramount in navigating this terrain, she adds, noting that computational scientists and machine-learning experts played pivotal roles in guiding the integration of clinical data with advanced analytical tools.
The project's core objective was the automation of prognostic assessments, specifically in determining the likelihood of long or short PFS in patients with ovarian cancer. By leveraging machine-learning algorithms, investigators sought to streamline this process, potentially enhancing clinical decision-making and better individualizing patient treatment decisions. Notably, this endeavor marks a departure from conventional methodologies, toward data-driven approaches in oncological practice, Asare mentions.
The utilization of laparoscopic assessment videos as primary data sources underscores the interdisciplinary nature of the project, bridging clinical and computational domains, she continues. Through the amalgamation of diverse expertise, the team aimed to optimize the integration of clinical information with machine-learning models, fostering a symbiotic relationship between medical practitioners and technology.
Asare also emphasizes the essential role of interdisciplinary collaboration in navigating the complexities of AI integration in health care. Moreover, the initiative underscores the potential of machine-learning technologies to augment clinical decision-making processes, she concludes.