Pioneering AI models to better diagnose lung disease in women

Kalysta Makimoto, a TMU doctoral candidate in Biomedical Physics, is among this year’s 10 Ontario Women’s Health Scholars. Her groundbreaking research is aimed at improving diagnosis and health outcomes for women with Chronic Obstructive Pulmonary Disease (COPD). 

The progressive lung disease impacts men and women almost equally, affecting 9.4% of men aged 35 and older and 9.1% of women in the same age range. However, women are severely under-diagnosed.

Makimoto’s interest in COPD is personal. While in TMU’s undergraduate medical physics program, her grandmother was diagnosed with the disease, which causes the airways in the lungs to become swollen and blocked making it difficult to breathe. 

“At that time, I wasn’t sure if I wanted to pursue medical school or graduate school. One of the projects in my fourth-year thesis course was looking into COPD. I chose it to learn more about the disease. That’s when I fell in love with research.” 

COPD  is the sixth leading cause  (external link) of death in Canada and the second-leading cause of hospitalization. According to the Canadian Lung Association (external link) , more than 2 million Canadians are living with the treatable but not curable disease and it’s estimated another million are living with COPD without knowing it because it is under-diagnosed – this is particularly true for women. 

“COPD has historically been thought of as being a male-dominant disease, largely because smoking is the main risk factor and smoking prevalence has traditionally been higher among men than women,” says Makimoto.  

There are several other key differences between the sexes when it comes to COPD that haven’t been taken into account when it comes to how the disease is diagnosed – until Makimoto’s research. For example, women tend to have smaller lung sizes and airways than men. Men tend to have more emphysema than women but women tend to have more severe symptoms that require hospitalization, and are at a higher risk of dying from respiratory failure. 

Makimoto is developing machine learning prediction models for the whole population as well as for men only and women only that can predict COPD status based on computed tomography (CT) imaging with the ultimate goal of improving detection accuracy, prediction of who is at risk of developing COPD and treatment. “Ideally, we’d like to identify sex-specific features that might be more predictive of the disease since we know the disease presents differently in males and females,” she says. 

“Being able to identify exactly what’s changing in the lungs could help doctors identify optimal treatments for specific patients that could help improve disease management, quality of life, and outcomes. Since females are being under-diagnosed, the main goal is to diagnose them at the same level as men.” 

Makitmoto’s research also provides a new, more accessible way to diagnose COPD. 

The current standard for diagnosing COPD is with spirometry, which measures lung function. However, this is only collected when someone presents with symptoms and/or a smoking history. Integrating CT images with machine learning creates an opportunity to identify individuals with COPD or who are at risk of progressing to COPD that haven’t been diagnosed. 

“Unfortunately, my grandmother passed away from COPD in 2022,” says Makimoto. “On a personal level, the Ontario Women’s Health Scholars award will allow me to work on a project that can help improve care for women. From a research perspective, it confirms that what I’m doing is important. ”

Makimoto has received several post-doctoral fellowship offers and is finalizing her decision. “The differences between males and females in health research is not widely investigated for machine learning in COPD. There are so many unknowns that need to be answered. I have another project stemming out of this one that will also be focused on trying to address the gap between males and females in terms of disease progression.”

Makamoto’s advice to the next generation of women looking to enter STEM fields: You can do it. “As young as three or four, I remember loving math. In high school, I discovered my love for biology. Medical physics allows me to apply math theories to biology. It’s the right fit for me. More women are entering these fields all the time. The research lab I work in is led by a woman and is equally split between men and women. If you’re interested in STEM, do it wholeheartedly; try your best.”