Pushing the Boundaries of Cancer Care

Personalized medicine is the future of healthcare.

That is the perspective Mohamed Abazeed brings to his Northwestern Medicine radiation oncology lab, where he and a team of investigators blend medicine with biology and technology to create new ways to fight cancer.

"Personalized medicine, to us, represents a transformative approach to healthcare, one where treatment is tailored specifically to the individual characteristics of each patient," said Abazeed, a medical doctor who specializes in the management of patients with lung cancer. "Rather than adopting a one-size-fits-all approach, personalized medicine considers a patient’s unique genetic makeup, imaging features, and disease characteristics to devise the most effective radiation treatment strategy."

To do that, Abazeed uses genomics and artificial intelligence (AI) — with a focus on deep learning algorithms — to develop individualized treatment plans that optimize efficacy and minimize side effects. The goal is to predict how a patient will respond to therapy and then adjust the approach as needed.

"Our lab has pioneered the application of generative AI in radiation oncology, including computer vision and large language models (LLMs) to assist in clinical decision-making," he said. "By integrating these advanced technologies, we aim to push the boundaries of what’s possible in cancer treatment and improve outcomes for patients." 

For the past two years, those boundaries have been pushed thanks to students in Northwestern Engineering's Master of Science in Artificial Intelligence (MSAI) program. MSAI students positively contributed to the lab's critical projects, enhancing the learning experience while advancing the work, said Peng (Troy) Teo, a radiation oncology instructor in Northwestern's Feinberg School of Medicine and an investigator in Abazeed's lab.

For one project, MSAI students used advanced deep learning techniques to develop AI models for tumor segmentation. The focus was to accurately delineate tumor boundaries in medical imaging to help with precise radiation therapy planning.

Teo said the students significantly refined the model's accuracy, which led to more reliable and efficient tumor targeting.

The second project explored image-based prediction of pneumonitis, a serious side effect of radiation therapy. The students helped create deep learning models that analyze medical images to predict the likelihood of pneumonitis in patients who undergo treatment. The goal was to identify at-risk patients early so that adjustments could be made to their treatment plans and minimize the risk of further complications. 

"The collaboration with MSAI students helped advance our understanding and refine the predictive models," Teo said. "Their work contributed to one of the most accurate models of this toxicity we've developed so far."

Abazeed appreciates the enthusiasm and creative energy he sees from MSAI students. Since working with the program, he's been impressed by students' willingness to collaborate and learn.

The experience has been mutually beneficial, he said. 

"Working with our lab provides MSAI students with a unique opportunity to apply their computational skills in a highly specialized and impactful domain," Abazeed said. "This experience allows them to see firsthand how AI can be integrated into real-world healthcare settings, addressing complex challenges such as tumor segmentation and cancer treatment optimization."

The students also get to see how their work can ultimately impact patients.

"Personalized medicine embodies the promise of precision, where every treatment decision is informed by a deep understanding of the individual patient," Abazeed said. "This not only improves clinical outcomes but also empowers patients by involving them more directly in their care, fostering a more personalized and compassionate approach to medicine."