Data scientist Dr. Aline Talhouk, PhD’13 works within a clinical research team, where she has learned to understand the language of medicine—a cutting-edge advantage that is helping her develop a model that can save many women with endometrial cancer from unnecessarily aggressive surgery and treatments for a faster return to good health.

“A clinician’s skillset is in healing the body, so we wouldn’t expect them to know what’s possible in terms of informatics, and computer scientists may not be aware of relevant medical problems,” says Dr. Talhouk, assistant professor in UBC’s department of obstetrics and gynecology.

“The opportunity to be invested in both fields means I’m able to bridge this gap.” —Dr. Aline Talhouk

She uses a type of artificial intelligence—machine learning, or computer algorithms that can learn from hundreds of thousands of molecular markers from many patients—to classify endometrial cancer into four subtypes with different survival rates.

Endometrial cancer, the most common gynecological cancer in the developed world, is often treatable by simple surgery (hysterectomy). Less than 15 per cent of patients have the aggressive, deadly form of the disease. These patients need aggressive surgery by a gynecologic oncologist and often have to endure exhaustion and nausea brought on by chemotherapy and radiation afterward. The goal is for doctors to identify as early as possible the many patients with less aggressive forms who will be cured by surgery alone and can be treated in their community.

“To know we prevented someone from going through aggressive surgery and the toxicities and side-effects of cancer treatment without changing their outcome, I never thought I’d have a dream like that. But it’s so meaningful to me today,” says Dr. Talhouk, a member of the Gynecologic Cancer Initiative, a provincial collaboration focused on accelerating reproductive cancer research into clinical care.

Dr. Talhouk developed and validated the model over the past five years—a quick turn-around by today’s standards in medical research—thanks to her collaboration with gynecologic oncologist Dr. Jessica McAlpine, associate professor and co-head of UBC’s division of gynecologic oncology and Dr. Chew Wei MBBS [HK] FRCOG [ENG] Memorial Chair in Gynecologic Oncology.

Dr. McAlpine is currently conducting a British Columbia and international clinical trial to test how the model can change medical care. Also, the World Health Organization recently adopted this made-in-B.C. classification of endometrial cancer subtypes.

B.C. has one of the richest health data ecosystems in the world. Dr. Talhouk works to improve how gynecology data is collected, stored, accessed and used across the province to support the development of better risk prediction models for reproductive cancers.

“There’s so much progress going on in the artificial intelligence and machine learning fields that we can apply in so many settings to improve the outcomes of our patients. The future of medicine is with data sciences.”