A Wavelet-Based method for multifractal analysis of medical signals: Application to dynamic infrared thermograms of breast cancer

Abstract

We use the wavelet transform modulus maxima (WTMM) method to perform multifractal analysis of the temporal fluctuations of breast skin temperature recorded using infrared (IR) thermography. When investigating thermograms collected from a panel of patients with breast cancer and some female volunteers with healthy breasts, we show that the multifractal complexity of temperature fluctuations observed on intact breast is lost in mammary glands with malignant tumors. These results highlight dynamics IR imaging as a very valuable non-invasive technique for preliminary screening in asymptomatic women to identify those with risk of breast cancer. Besides potential clinical impact, they also shed a new light on physiological changes that may precede anatomical alterations in breast cancer development.