Classification of Cervical Cancer Cells Using HMLP Network With Confidence Percentage and Confidence Level Analysis

• Autores: N.A. Mat-Isa, Mohd Yusoff Mashor, Mohamed Othman
• Localización: International journal of the computer, the internet and management, ISSN 0858-7027, Vol. 11, Nº. 1 (ENE-ABR), 2003, págs. 17-29
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • In most previous studies, the analysis on the ability of neural networks to be used as a good cervical cancer diagnosis technique is only based on accuracy, sensitivity, specificity, false negative and false positive. In the current study, we go one step further by introducing analysis of diagnosis confidence percentage and diagnosis confidence level to analyse the ability of neural network to produce a good diagnosis performance. The current study used hybrid multilayered perceptron (HMLP) network to diagnose cervical cancer in the early stage by classifying cervical cells into normal, LSIL and HSIL cell. The proposed diagnosis confidence percentage and diagnosis confidence level analysis have been proved to give clearer picture on the strength or confidence level of each diagnosis, which is done by HMLP network.