TITLE: AI Cancer Research Projects to Launch at Ohio State in January

AI Cancer Research Projects Set to Begin at James Cancer Hospital

Six artificial intelligence-based research projects aimed at predicting cancer risks and detecting tumors are scheduled to commence development in January at the James Cancer Hospital and Solove Research Institute.

The initiatives, led by radiation oncologist Dr. Simeng Zhu, will leverage AI algorithms to identify tumors and forecast recurrence risks across various cancer types. The projects will utilize the Ascend supercomputer cluster at the Ohio Supercomputer Center to process complex data sets.

"AI machine learning is becoming very popular, and I believe it represents the future," Zhu said in an interview. He emphasized the growing role of AI in medical advancements, particularly in oncology.

Each project will involve approximately two student researchers, drawn from diverse academic backgrounds including computer science, engineering and pre-medicine. Zhu selected participants based on their enthusiasm for learning rather than prior expertise.

"The main criterion is their willingness to learn, which determines the pace of knowledge acquisition," Zhu explained. This approach fosters rapid skill development and interdisciplinary collaboration.

Project Structure and Methodology

The research will proceed in three distinct phases. The initial stage focuses on data curation, gathering and organizing medical datasets essential for training AI models. The second phase involves model development, where students will program the algorithms to analyze patterns in cancer data.

The final phase entails algorithm validation, testing the models against real-world clinical scenarios to ensure accuracy and reliability. This structured process aims to produce robust tools for clinical use.

Hari Garish, a second-year computer science and engineering student, is participating in one of the projects. He highlighted the value of gaining domain-specific knowledge in healthcare.

"To improve as a machine learning practitioner, it's crucial to build conceptual understanding of the industry," Garish said. He noted the opportunity to learn from Zhu about integrating AI into medicine, a skill applicable across sectors.

Pre-medical students are also integral to the teams, providing clinical insights that enhance the algorithms' relevance. "Involving them is vital because they contribute the medical perspective, which is essential for developing effective medical AI," Zhu said.

Garish added that AI implementation spans industries, but applying it to healthcare offers unique challenges and rewards. "You can enter any field and find AI data analysis needs," he observed.

Challenges and Risks in Medical AI

While promising, integrating AI into oncology carries significant risks due to the high stakes involved. Zhu compared non-medical AI errors, such as misclassifying a cat as a dog, to clinical misjudgments that could affect patient outcomes.

"In clinical prediction, the standards are much higher," he cautioned. The projects will prioritize rigorous validation to mitigate errors and ensure ethical application.

The interdisciplinary nature of the teams addresses these challenges by combining technical prowess with medical knowledge. Students will gain hands-on experience in a cutting-edge field, potentially influencing future careers in AI and healthcare.

As AI adoption accelerates in medicine, these projects at the James Cancer Hospital represent a proactive step toward personalized cancer care. Development begins in January 2026, with potential for broader implementation pending successful validation.