Bringing AI to doctor’s offices: Spotlight on John-Jose Nunez

Imagine completing a full-time medical residency in psychiatry while also earning your computer science masters degree part-time over four years? That’s exactly what UBC computer science alumnus Dr. John-Jose Nunez managed to do successfully.

Today, Dr. Nunez is a psychiatrist and Clinical Research Fellow with BC Cancer and the UBC Mood Disorders Centre, where he is combining his passions. He conducts research using Artificial Intelligence (AI) to predict a cancer patient’s need for psychosocial services like psychiatry and counselling, and in another research study: he uses it to predict the survival of cancer patients after their initial oncology consultation.

Both studies use Natural Language Processing (NLP) – a branch of AI that gives computers the ability to understand text and spoken words similar to how human beings can. And according to the findings, this approach goes a long way to aiding patients in their journey with cancer.

The research Dr. Nunez conducted to predict a patient’s need for psychosocial services uses NLP to analyze an oncologist’s clinical dictations in order to find clues that a patient may be experiencing depression or anxiety. Why is it so important to identify depression or anxiety within cancer patients? Because on average, cancer patients with significant depression and anxiety do not survive as long as cancer patients who do not have those symptoms. That’s because they have more difficulty following through with recommended cancer treatments and/or tolerating the side effects of treatment.

His other research on predicting cancer survival rates resulted in findings recently published here:  Predicting the Survival of Patients with Cancer from their Initial Oncology Consultation Document Using Natural Language Processing. This was a collaborative and cross-discipline study authored by Dr. John-Jose Nunez (UBC Psychiatry, BC Cancer and UBC Computer Science), Dr. Bonnie Leung (BC Cancer), Dr. Cheryl Ho (BC Cancer), Dr. Alan T. Bates  (BC Cancer & UBC Psychiatry), and Dr. Raymond T. Ng (UBC Computer Science).

In the study, the researchers used AI to predict a 6-month, 36-month and 60-month cancer survival rate with a greater than 80 per cent accuracy (close to 85 per cent). That’s comparable, and in often cases better, to previous models used to predict cancer survival. Most importantly is that this model may be able to predict survival without focusing on one particular cancer type (which previous models have been limited to). Nunez equates the 85 per cent survival rate prediction accuracy with being an A minus student: “Not perfect, but very good,” he said.

Dr. Nunez explained the intersection of his areas of research focus, “There is a gigantic gap between computers and health. Doctors need the most recent evidence and data, but they are still human and are obviously limited by how much they can fit in their brains. Plus, everything advances so fast. This is where AI and data can help. Not to replace doctors, but to be a tool working alongside the doctor to combine data with knowledge, while being current.”

Can computer science cure cancer?

Dr. Nunez believes computer science has the potential to help cure some cancers and to improve the lives of people with cancer. “We have already started to cure some cancers. By aiding cancer care with tools like AI, we can improve the survival rate, enhance medical protocols and improve the quality of life for patients in hospice care.” 

Nunez could not have known just how relevant this subject would be, as his own father was diagnosed with cancer. He is now in remission.

In terms of pursuing his MSc with UBC computer science, Nunez says he has one regret: “I should have transferred to computer science as an undergrad after I realized how much I liked it and how well I did,” he reflected. Nunez recounts having received a mark of 100% in “Introduction to Computer Science 110” with Gregor Kiczales. “I loved every minute of my Masters. I had fantastic mentors and having the ability to pursue it part time was awesome.” 

He recalls some of his favourite classes and professors, “I loved my Natural Language Processing class with Giuseppe Carenini, and Tamara Munzner’s Data Vis course was so helpful and great. Of course, Dr. Raymond Ng was such an asset to my work and an amazing mentor. He provided me with such an incredible opportunity to explore the intersection of computer science and medicine.”

When asked what he would advise students considering grad school, Dr. Nunez says, “Pursue grad school because you have a passion and you have identified a problem that you want to solve. Also, computer science is a great degree before you go to med school! The application of AI within health isn’t going away. It infiltrates every aspect of health research and can be such an important tool to understand and utilize in medicine. If you have interest in computer science, go for it, because you can then use all those CS skills to address all the issues that computer science can help with.”

Dr. Nunez also wishes to acknowledge those who have funded his research: the UBC Institute of Mental Health (IMH) Marshall Fellows Program and the BC Cancer Foundation.

For someone like Nunez, the collision of his disciplines can create significant headway for mankind when it comes to pervasive diseases like cancer. It’s these types of interdisciplinary applications that computer science thrives on.