Breast cancer is one of the crucially prevailing cancers among women. Early detection and diagnosis of breast cancer is very important for patient survival. Currently, mammography is the most common method used for detection of breast cancer. However, mammography has certain limitations. It is not recommended to be used on women under the age of 45 years, cannot be used on lactating women and is inconclusive on women with high density of breast. Mammography in correlation with sono-mammography is widely used currently for screening in hospitals .

Infrared imaging of the breast or thermography is an imaging technique which detects suspected malignancy based on thermal changes on the chest. Due to increased blood circulation and metabolic activity of the tumour, the temperature of the tissue is higher around the regions where a tumor is present. A high resolution thermal camera can be used to capture this thermal information. This thermal image representing the temperature variations observed on the breast is called a breast thermogram. A thermal image provides a visual representation of the temperature distribution and can be used for clinical interpretation of abnormal breast conditions as an adjunct modality. Particularly, radiologists who are also certified thermologists are able to interpret these breast thermograms to determine its malignancy or benign condition.  However, manual interpretation of these themograms is very cumbersome and error prone.  Thermalytix is a new method of interpreting thermal images using CAD anovel Artificial Intelligence and Machine Learning Algorithms and has shown promising results in early experiments.

Our proposed research is to improve breast cancer screening through low cost, mobile, non-invasive and non-contact thermal screening without causing any discomfort to patients and develop accurate interpretation algorithms using machine learning, image processing and computer vision techniques to give automated screening capabilities.

In this study, sensitivity of Thermalytix© (an automated thermographic screening algorithms developed by NIRAMAI) will be compared to the sensitivity of standard screening modalities. The entry criteria for the study will be subjects who show possible symptoms of suspected breast cancer, such as pain, lump and so on.

The primary objectives of the study are to assess a non-inferiority of sensitivity of Thermalytix© methodology over the sensitivity of standard screening procedures by not more than 10% and study comparative diagnostic performance of Thermalytix© with standard screening modalities. The endpoint measurements would be sensitivity, specificity, PPV (Positive Predictive Values) and NPV (Negative Predictive Values) in the detection of breast cancer in women.

Secondary Objectives would be to study the influence of effects of patient characteristics on diagnostic accuracy. Patient characteristics would include, Age, Lesion type, Pathologic diagnosis, Menopausal and hormonal status, Breast density and Family history. Also to study how Thermalytix© can complement the standard modalities in breast cancer screening in future

This protocol for human research study will be conducted according to standards of effective Good Clinical Practice (International Conference on Harmonization (ICH) Guidelines), if applicable the Indian regulations and applicable Ethical Guidelines for Biomedical Research involving Human Participants issued by the Indian Council of Medical Research. The trial protocol and subject documentation will be reviewed and approved by the Institutional Ethical Review Boards of the participating trial sites.