Heat transfer capacity in millimeter size breast cancer cells analysis through thermal imaging and FDNCNN for primary stage identification. (February 2023)

• Record Type: Journal Article
• Authors: Geetha, P.; UmaMaheswari, S.
• Is Part Of: Biomedical signal processing and control. Volume 80:Part 2(2023)
• Publication Date: 2023-02
• Subjects: Breast phantom -- Thermal imaging -- Heating of phantom -- Breast cancer cell size-based detection -- Tissue heating; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.104361
• Languages: English
• ISSNs: 1746-8094

Abstract

Abstract: In this paper, cancer cells in human breast of below 256 cell (2 cm) counts are detected through proposed Feature DnCNN (FDnCNN) enhancement algorithm. Cancer cell count below 256 is termed as T1b and T1c stages. FDnCNN is applied to the breast thermal images obtained after external heating of breast phantom through warm water. Induced heat due to warm water immersion of breast phantom helps for detection of the small size cancer cell through heat transfer capacity between normal and cancer cells. However, traditional method of breast cancer detection through thermal imaging never detects the cancer cell size of below 2 cm (256-cell) in the images. Thermal imaging breast cancer detects cancer cell of above 256 cells count in breast and never below cancer cell of count 256 cells. The proposed method of acquiring the thermal image after induced heating on the breast phantom, leads to higher interpretation of cancer cells below 256 cell count because of increase in thermogenesis of tissue. The increase in thermogenesis in cancer tissues improves visibility of cancer cell below 256 cell count. Moreover, induced heat based phantom thermal images need enhancement algorithm for better visualization of cancer cell count of below 256. The proposed FDnCNN enhancement algorithm performs better than traditional algorithms and provides an accuracy of cancer cell measurement of about 95%, when compared to ground truth measurement of cancer cells.