Trustworthy Weakly Supervised Breast Cancer Detection in Ultrasound Imaging

Abstract: Breast cancer has continued to be a significant health issue in recent years, being the most frequently diagnosed cancer among women in the United States. Early-stage breast cancer lacks apparent symptoms in the early stage of breast cancer; thus, many patients miss the best chance to cure it. Breast Ultrasound (BUS) imaging has emerged as a crucial diagnostic tool. Yet, the inherent challenges in BUS imaging, such as low contrast and noise, impede accurate breast cancer diagnosis by doctors. Computer-aided-diagnosis (CAD) systems are proposed to help radiologists interpret BUS images, make a more precise diagnosis, and reduce their workload. Breast cancer detection plays a crucial role in the CAD system. Training a fully supervised model for detecting breast cancer necessitates many manual annotations. Several weakly supervised frameworks for object detection have been developed to minimize the requirement for extensive annotations. During the SpF 2023 award period, the research team developed two such frameworks for natural images and breast cancer detection. However, trustworthiness is the most critical aspect of smart health applications. Constructing a trustworthy deep-learning model requires extensive datasets. Additionally, previous methods do not offer a qualitative assessment of the models’ trustworthiness. To overcome these hurdles, the researchers propose a new, trustworthy, weakly supervised framework for breast cancer detection named BUSwiNet. This framework utilizes only image-level labels (cancer/non cancer). It is capable of identifying the bounding boxes of breast tumors in breast ultrasound (BUS) images, as well as determining whether the tumor is cancerous. Moreover, the research aims to develop a method using Bayesian Neural Networks to evaluate the trustworthiness of breast cancer detection models quantitatively.

Objective: The project is focused on two primary objectives: 1) to develop an innovative weakly supervised framework for breast cancer detection named BUSwiNet. 2) to employ Bayesian Theory to evaluate the trustworthiness of BUSwiNet and the models developed during the SpF 2023 period to identify the most reliable model. Furthermore, in response to the growing interest in AI and smart health, the research team intends to establish a summer learning environment for three undergraduate students. It is designed to spark their interest in AI and healthcare and prepare them for future careers.

Significance and expected outcomes: The importance of this project lies in several key areas: Firstly, it is expected to not only increase the accuracy of breast cancer diagnosis but also has the potential to be adapted for other types of cancers, such as lung and brain cancers. Secondly, the method for estimating trustworthiness could provide a strategy applicable across various contexts. Thirdly, this project offers an excellent opportunity to mentor undergraduate students in my initial year as a faculty member at Kean University, engaging in research collaboration with students and enhancing faculty-student relationships. The expected research outputs include 1) our proposed BUSwiNet to outperform existing methods on three public BUS datasets and 2) a publicly available software tool for breast cancer detection and trustworthiness evaluation, 3) three peer-reviewed papers: two conference papers addressing objective one and a journal article focusing on objective 2.

Location: Kean University

Academic Term: Summer 2024

People

• Principal Investigator: Dr. Meng Xu
• Co-Principal Investigator: Dr. Kuan Huang
• Undergraduate Student Researchers:
  • Maryam Ahmed
  • Joanna Loja
  • Armando Mendez

Publications

• To be updated …

Acknowledgements

This project is funded by the Students Partnering with Faculty (SpF) program at Kean University.