(https://www.fda.gov/medical-devices/safety-communications/fda-warns-thermography-should-not-be-used-place-mammography-detect-diagnose-or-screen-breast-cancer).
Off late some customers are requesting clarity on the above safety communications warning issued by US FDA about usage of thermography for screening, detection or diagnosis. There are some important considerations showing why the Meditherma solution is not the same as Thermography, and those differences are detailed below:

1. Meditherma solution uses Infralytix© which is different from Thermography. Though the sensing method used by Infralytix includes thermal images (measurement of temperature variations on the skin), Meditherma Infralytix Solution uses a unique AI-enabled software that gives objective assessment of the thermal analysis, which is missing in thermography. Additionally, demography data about the patient as well as a novel vascularity analysis is performed. So, we should treat Infralytix as a new solution for detecting breast abnormalities.
2. Thermography clinics provide a report based on visual analysis of thermal images. Thermal images are very complex with about 400,000 temperature points shown as a color pixels, where each pixel can be any of the 2000 colors. Analyzing and interpreting these thermal images with just visual inspection is a huge cognitive overload for even an experienced thermographer. So, the manual diagnosis used in Thermography is very error prone. Hence we agree with FDA guidance of using visual analysis of thermal images (thermography) as an adjunct to existing methodologies.
3. Thermography clinics are also NOT able to provide any report with one sitting of the patient. They usually ask the patient to come back in 3 months and do a visual inspection of 2 different sets of images taken 3 months apart. This results in potentially missing a case that was evident in the first sitting, and delaying the identification of malignancy for the patient. Meditherma solution is able to analyse the thermal images for every screening instance and provide a quantitative report.
4. Meditherma solution is based on machine learning models – that have the ability to analyse all the 400,000 color points both in imaging domain as well as temperature domain. A difference of 0.02 deg C is perceived by Meditherma software and hence the analysis is very sensitive, unlike thermography where two different shades of a color are hard to distinguish.
5. Meditherma machine learning models are calibrated with historical data labelled with current standard of care. The way Meditherma models were trained was based on labels learnt from mammography + sono-mammography + biopsy. It was not trained using reports of thermographers, hence Meditherma results are more accurate and better than current standard of care.
6. In order to make an objective comparison between Thermography and Infralytix, a clinical trial on 303 patients was conducted in a blinded retrospective analysis. This was peer-reviewed and published in International Conference on Advances in Breast Cancer Treatments, Kyoto, Japan, April 26-27, 2018. The results of this comparison showed a remarkable improvement in performance of Infralytix over Thermography. An increase of 70% was seen in Positive Predictive Value over Thermography was seen, with very high levels of sensitivity and specificity.

Infralytix is available only in India at this time and usage in India does not fall under the purview of the above  mentioned US FDA communication.  We are under the process of getting an FDA clearance for our AI-based solution as well.