Point-of-care ultrasound and deep learning for breast cancer diagnostics in low-resource settings

Topic: Point-of-care ultrasound and deep learning for breast cancer diagnostics in low-resource settings

When: 11 September at 12.00-13.15

Where: Online

Speakers: 

• Ida Arvidsson, Computer Vision and Machine Learning, Mathematical Sciences, Lund University
• Jennie Karlsson, Computer Vision and Machine Learning, Mathematical Sciences, Lund University

Spoken language: English

Abstract

Despite breast cancer being the most common cancer globally, women in low- and middle-income countries have limited access to health care leading to late-stage diagnosis and poor survival rates. To address this shortness, a possible solution would be an accessible breast diagnostic tool consisting of a point-of-care ultrasound (POCUS) device paired with a smartphone-based algorithm, to examine suspicious tissue in the ultrasound images. In an ongoing project we are developing such an algorithm based on deep learning. Some of the challenges we have tried to tackle include making the algorithm generalise well to ultrasound probes from different vendors, detect out-of-distribution images such as ultrasound images of too poor quality, as well as integrate uncertainty assessment in the algorithm’s predictions. 

Bios

Ida Arvidsson is a postdoctoral researcher at the Division of Computer Vision and Machine Learning at the Centre for Mathematical Sciences, Lund University. She received her Ph.D. degree in Mathematics in 2021, titled "Applications of Deep Learning in Medical Image Analysis - Grading of Prostate Cancer and Detection of Coronary Artery Disease". Her main research interest is machine learning with a focus on deep learning and medical applications.

Jennie Karlsson is a doctoral student at the Division of Computer Vision and Machine Learning at the Centre for Mathematical Sciences, Lund University. Her research is within the field of deep learning for medical applications, focused on detecting breast cancer in ultrasound imaging.