Hand pressure estimation by image entropy for a real-time breast self-examination multimedia system

The degree of hand pressure applied on a breast determines the possibilities of lump detection during breast self-examination. In this paper, a method for estimating hand pressure is presented by means of analysing image entropy for an image sequence obtained through a Web camera. This is achieved by, firstly, calculating the difference between the current and initial images followed by the entropy of the difference. The estimated pressure is proven to be accurate through a human-like silicon breast experiment. The dynamic model of the breast is set up, close to Hook's rule, by means of verification of the proposed pressure estimation. The algorithm is simple and the calculation is limited in a local region of the images, which is beneficial for real-time realization.