Breast cancer detection using machine learning with thermograms in an edge computing scenario

Abstract

The second cause of death in the world is cancer. Although breast cancer is the more common cause of death among women, the chance of survival can be increased by detecting cancer in the early stages. For this aim, there are different tests such as Magnetic Resonance Imaging (MRI), mammogram, ultrasound, thermogram and among these tests, the mammogram is the one which is used more frequently. Regarding the advantages and the results, which are achieved by thermogram, it can be a good alternative or complement for the mammogram if we can improve the weaknesses of the thermogram. For this reason, in this research, we work on a thermogram to see the possibility of having a good performance. On the other hand, we train another model by sending personal patients' information to see the effect of these data to improve the performance of breast cancer detection. In the end, we plan to separate the process between edge and core host to do the process faster, safer, and cost-effective.