Detection of Masses in Mammograms Using Cellular Neural Networks, Hidden Markov Models and Ripley's K Function
Wener Borges de Sampaio, Edgar Moraes Diniz, Aristófanes Corrêa Silva, Anselmo Cardoso de Paiva · 2009
Breast cancer shows high frequency and its psychological effects affect the female's perception of sexuality and their personal image. The mammographic images processing has contributed to the detection and diagnosis of breast nodules, and contributes as an important tool, reducing the diagnosis uncertainty. This work presents a computational methodology that helps the expert in the task of mass detection based on mammographic images. To achieve this, hidden Markov model and Ripley's K function were used to detect masses, segmented by cellular neural networks. In the tests methodology demonstrate a sensitivity of 94.62%, with 92.57% of specificity, 93.60% of accuracy rate and an average of 0.53 false positives per image.