SKIN TUMOR DETECTION THROUGH IMAGE PROCESSING

Abstract

In this paper a new method for skin tumor detection is developed using image processing. Considered a variation
formulation, the energy of which combines a diffuse interface phase-field model (regularization term) and a loglikelihood computed using nonparametric estimates (data attachment term). Adopted CNN with the exact solutions
which have the advantage to avoid space discretization and numerical instabilities. The resulting algorithm is simple
and easy to implement in multi-dimensions. Concerning applications, focused on skin tumor segmentation. The
clinical dataset used for the experiments is composed of 15 images with the ground truth given by a dermatologist.
Comparisons with the reference methods, the proposed method is more robust to the choice of the volume
initialization. Moreover, the flexibility introduced by the diffuse interface, the sensitivity increases by 12% if the
initialization is inside the lesion, and the Dice index increases by 59%, if the initialization covers the entire lesion.
The results show that this new method is well designed to tackle the problem of underestimation of tumor volumes.