John Shepherd, PhD in front of the UH Cancer Center building

14 April 2018

Expert medical imager joins UH Cancer Center

John Shepherd, PhD, joins the University of Hawaiʻi Cancer Center as a senior researcher in Epidemiology from the University of California, San Francisco (UCSF) School of Medicine. He was at UCSF in the Radiology Department for 19 years and is a National Institutes of Health Building Interdisciplinary Research Careers in Women’s Health Fellow, a 2013 Fulbright Scholar, and a frequent consultant to the International Atomic Energy Agency (IAEA). He is most known for his expertise in quantitative X-ray imaging using machine learning for bone densitometry, body composition and mammography applications.

What are some of your research goals?
Ultimately, I want to reduce the burden of cancer on individuals. I approach this by studying risk factors for cancer using medical imaging.

Some examples of my work include using 3-dimensional optical scans of the body to measure body shape and how it relates to obesity and cancer risk. We do this by comparing body shape to fat, muscle and bone distributions that currently can only be measured using less accessible x-ray imaging scans. Another example is the use of screening mammograms to estimate not just if the person has cancer, but how likely she is to develop breast cancer in the future. The most exciting tool we use in these investigations is a type of artificial intelligence called “deep learning.” Using deep learning, we are extracting so much more information from medical images than we could even five years ago.

How will your research be beneficial to the people of Hawaiʻi?
Women in Hawaiʻi have not been included in any of the U.S. breast cancer risk models because there has not been a large scale effort to collect screening mammograms. I am working with my colleagues at the UH Cancer Center to create the first Hawaiʻi and Pacific Islands Mammography Registry. Our goal is to teach deep learning algorithms how to read mammograms to not only detect cancer but to identify the women who will develop cancer in the next five years. This requires literally several hundred thousand mammograms to train the algorithms, but we end up with very accurate models to predict who is at high risk of breast cancer that are specific to our unique ethnically-diverse population of women in Hawaiʻi. Once identified, women at high risk can then work with their doctors to reduce their breast cancer risk and hopefully avoid getting breast cancer.

I also study obesity using X-ray and 3-D optical images, the kind that are used for fitting clothes and video games. Obesity is a strong risk factor for cancers in general including breast cancer and our 3-D optical methods are a very accessible and accurate way to assess a person’s fat and muscle status.

What does it mean to you to join the UH Cancer Center?
There is lots of terrific work going on in the UH Cancer Center and I am honored to join the effort. I have been working with the Center on various projects for more than 10 years now. First, it was to develop a way to measure breast density in young girls with Gertraud Maskarinec, MD, PhD and Rachel Novotny, PhD. Later, I worked with Loic Le Marchand, MD, PhD and Unhee Lim, PhD on the Multiethnic Cohort Study to quantify body composition and cancer risk using dual-energy X-ray and MRI images. Many of my current projects were inspired by these collaborations. I hope to go much deeper into my research now that I am faculty at the Center.

Are you excited to move to Hawaiʻi?
I am a surfer and that should say it all! Cowabunga! But before coming, I was only really familiar with the Waikīkī area. I am super excited about getting to know the people, culture, and food of all of Hawaiʻi. Since arriving, I have been enjoying swimming the length of Ala Moana Park, biking to work from Mānoa, and eating lots and lots of poke!

My wife Jessica and my three daughters, Hewson, Maggie and Sarah are just very honored and happy to be part of the UH Cancer Center ʻohana.