Baltimore, MD — With the rapid advancement of artificial intelligence (AI), predictive medicine is becoming a pivotal part of healthcare, particularly in cancer treatment. Predictive medicine leverages algorithms and large datasets to help physicians better understand tumor growth patterns and patient-specific drug responses, enabling more precise and personalized cancer therapies.
What Are the Benefits and Limitations of AI According to Dr. Elana Fertig and Dr. Daniel Bergman?
Researchers at the University of Maryland School of Medicine (UMSOM) caution, however, that AI should not be used in isolation. In a recent commentary published in Nature Biotechnology on April 14, 2025, Dr. Elana Fertig, Director of the Institute for Genome Sciences (IGS), and Dr. Daniel Bergman, an IGS scientist, emphasize that AI’s predictive power is strongest when combined with traditional mathematical modeling.
“AI models primarily learn from existing datasets to predict outcomes, while mathematical models incorporate biological knowledge to address specific questions,” said Dr. Fertig. “When data is limited—as is often the case with emerging cancer treatments like immunotherapy—AI can produce biased or less reliable results. Mathematical modeling provides mechanistic insights that enhance understanding.”
Mathematical modeling enables the simulation of virtual cancer and healthy cells interacting within a tumor under different treatment scenarios, offering a specificity that AI currently cannot replicate, explained Dr. Bergman.
How Can Ethical Data Sharing Improve AI and Cancer Research? — Perspectives by Dr. Elana Fertig and Team
In addition to advocating for the integration of AI and mathematical modeling, the researchers highlight the importance of using diverse population datasets and promoting open data sharing to improve accuracy and reproducibility in cancer research.
Reproducibility is a known challenge in scientific research, with over 70% of researchers reporting failed attempts to replicate others’ experiments. In a related commentary published on April 15, 2025, in Cell Reports Medicine, Dr. Fertig and colleagues discuss ethical data sharing frameworks that protect patient privacy while enabling transparency and scientific integrity.
“Ethical and responsible data sharing not only advances AI research but also informs public health policy, maximizing benefits for patients worldwide,” said Dr. Dmitrijs Lvovs, lead author of the Cell Reports Medicine commentary.
To ethically share genomic and clinical data, researchers must secure informed patient consent, ensure data quality, harmonize data collection standards, and employ secure, open-source analysis platforms.
The University of Maryland’s research underscores a balanced approach to AI in medicine, advocating for collaboration between cutting-edge AI techniques and established scientific methodologies to accelerate discovery and improve patient outcomes in cancer care.
By