Skip to content

Shepherd Research Lab

    The Makawalu Study: Breast Cancer Screening in the Pacific using Portable Ultrasound

    Overview

    Advanced stage breast cancer (stages III and IV) rates in the Pacific are much higher than in the USA mainland, especially where mammography services do not exist or have low accessibility. Hawaii and Guam both have high insurance coverage and mammography access similar to the United States mainland, but have much higher advanced stage rates, 60% and 15% for Guam and Hawaii, respectively, compared to 10% for women in Northern California. Further, the advanced staging in Hawaii dramatically varies by race/ethnicity with a high of 17% (Hawaiian) and a low of 10% (Japanese). The reason for these disparities across ethnicities in Hawaii and the larger Pacific is not well understood but clearly multifactorial.

    Objective/Aims

    Our overall hypothesis is that introducing portable handheld ultrasound (US) systems coupled with an artificial intelligence (AI) detection algorithm and operated by a trained healthcare worker will reduce advanced stage cancer rates in areas where mammography is not available.

    For this to be true, the US screening approach would have to have similar sensitivity and specificity for breast cancer detection as screening mammography read by a radiologist. This approach would be very low cost and could also be extended to be used for point of care breast biopsies.

    Our overall objective is to reduce the mortality of breast cancer in the Pacific by reducing the advanced breast cancer stage rate, resulting in a reduced burden of cancer. We believe the primary cause of very high advanced stage breast cancer is the lack of mammographic screening. If successful, the implementation of our paradigm could reduce overall breast cancer mortality rates by identifying breast cancers earlier, when they are more responsive to treatment.

    Our specific aim is to identify cofactors (breast density, ethnicity, BMI, etc.) that impact the sensitivity and specificity for malignant breast cancer lesion detection on clinical ultrasound images with AI support for the following readers: 1) Radiologists, 2) Non-radiologist Doctor, and 3) Graduate Research Assistants (surrogate healthcare worker).

    Research Team

    Jami Fukui

    Principal Investigator

    UH Cancer Center

    Dustin Valdez profile photo 1

    Dustin Valdez

    Graduate Student Assistant

    UH Mānoa/UHCC

    John Shepherd, PhD

    Co-Principal Investigator/Mentor

    UH Cancer Center

    Thomas Wolfgruber profile photo

    Thomas Wolfgruber

    IT Support

    UH Cancer Center

    Funding Source(s)

    UH Cancer Center Seed Grant
    This grant is funded by the UH Cancer Center’s P30 support grant.
    01/01/2021 – 12/31/2021