Integrating Physic-Informed Neural Network, Bayesian and Convolutional Neural Networks for early breast cancer detection using thermography

Grant funding 2023-2025

The purpose of the project is to develop an intelligent computer-based breast cancer diagnosis system with the ultimate aim of helping to meet WHO’s BSE objective for minimization of the fatality rate of breast cancer. The system includes the integration of BN and CNN together with PINN using thermographic data collected from patients. Combining trained BNs and trained CNNs allows to predict tumors with interpretability and explainability, and PINN links forward and inverse thermal modeling to extract patient-specific tissue parameters, determines tumor sizes and locations for precision diagnosis.

According to the World Health Organization 2.3 million women were diagnosed with breast cancer, and 685 thousand women died from this disease in 2020. Moreover 7.8 million women were diagnosed with breast cancer from 2015 to 2020. This makes breast cancer the most predominant type of cancer among others. At early stages breast cancer is highly treatable, therefore it is vital to diagnose the disease at the earliest possible stages. This project aims to develop an early warning tool for the development of breast cancer based on the detection of growing tumors within the breast. This tool focuses on the initial “precautionary measure” phase which would be suitable for mass screening in remote areas whereby expensive equipment for diagnosis may not be readily available. Those identified as high-risk of breast cancer development can then be sent for more detailed examinations.

Scientific results will be presented in the form of scientific publications, applications for inventions of the Republic of Kazakhstan and European countries. At least 3 (three) articles and (or) reviews in peer-reviewed scientific publications in the scientific direction of the project, indexed in the Science Citation Index Expanded and included in the 1 (first), 2 (second) and (or) 3 (third) quartile according to impact factor in the Web of Science database and (or) having a CiteScore percentile in the Scopus database of at least 50 (fifty), during the whole project implementation