Thermography and breast cancer an analysis of a blind reading.

Thermography, a pictorial display of the infrared radiation of the breast, has been used for approximately 25 years as an aid in the evaluation of breast problems. Its inception dates to Lawson’s observations in 1956-1958 that a breast cancer may be warmer than surrounding tissue.’-’ He also showed that the venous blood draining the cancer may be warmer than its arterial supply. Many investigatorsCI6 have utilized thermography in conjunction with clinical examination and/or mammography for the evaluation of symptomatic patients and have reported varying true positive rates with its use. It has been and continues to be widely used in Europe, both as a diagnostic modality and in the follow-up and prognostic evaluation of breast cancer patient^.""^ Although the attractions of the techniques are widely recognized (noninvasive, inexpensive, with capability of rapid large volume screening), the modality has not been widely accepted in the medical community. This probably is due to the perceived difficulty in sensitivity and specificity levels of the technique and the lack of specific, discrete, objective criteria for its interpretive use. Many anecdotal reports and discussions vaunting the value of thermography as a diagnostic aid are in the medical literature. However, little supporting statistical evaluation is available. ReveszI9 has discussed this and recommended using relative-operating characteristic (ROC) curves as a means of measuring the detectability of thermography. He also has shown that