Automated identification of cancerous smears using various competitive intelligent techniques

Abstract

In this study the performance of various intelligent methodologies is compared in the task of pap-smear diagnosis. The selected intelligent methodologies are briefly described and explained, and then, the acquired results are presented and discussed for their comprehensibility and usefulness to medical staff, either for fault diagnosis tasks, or for the construction of automated computer-assisted classification of smears. The intelligent methodologies used for the construction of pap-smear classifiers, are different clustering approaches, feature selection, neuro-fuzzy systems, inductive machine learning, genetic programming, and second order neural networks. Acquired results reveal the power of most intelligent techniques to obtain high quality solutions in this difficult problem of medical diagnosis. Some of the methods obtain almost perfect diagnostic accuracy in test data, but the outcome lacks comprehensibility. On the other hand, results scoring high in terms of comprehensibility are acquired from some methods, but with the drawback of achieving lower diagnostic accuracy. The experimental data used in this study were collected at a previous stage, for the purpose of combining intelligent diagnostic methodologies with other existing computer imaging technologies towards the construction of an automated smear cell classification device.

Particle swarm optimization for pap-smear diagnosis

“…The secondorder neural networks approach obtained the highest diagnostic accuracy approaching 99% on new (test) data selecting nine features as the most important for the classification task. In Dounias et al (2006) all the previous results are summarized and discussed extensively. Except of the intelligent approaches, a number of metaheuristic algorithms have been proposed for the solution of the problem (Marinakis & Dounias, in press;Marinakis & Dounias, 2006a;Marinakis & Dounias, 2006b).…”

“…Each cell is described by 20 features extracted from pictures of single human cells. Several intelligent methodologies have been previously applied on the specific domain, with variable classification performance, see Ampazis, Dounias, and Jantzen (2004), Byriel (1999), Dounias et al (2006), Jantzen, Norup, Dounias, and Bjerregaard (2005), Marinakis and Dounias (in press), Marinakis and Dounias (2006a), Marinakis and Dounias (2006b), Martin (2003), Norup (2005), Tsakonas et al (2004).…”

Intelligent and nature inspired optimization methods in medicine: the Pap smear cell classification problem

“…The collaborative medical doctors of our research group, Georgia Panagi and Eva Panourgias, are greatly acknowledged for providing their helpful information and expertise (Dounias et al , 2006) regarding the influence of intelligent techniques in daily medical practice.…”

“…All the previous results are summarized and discussed extensively in Dounias et al (2006). Table 8 shows the errors (the root mean squared error, the error for group N and for group P, F N % and F P %, and the overall error OE%) for the proposed algorithms in the two-class problem for both data sets and for both metrics.…”

“…Each cell is described by 20 features extracted from pictures of single human cells. Several intelligent methodologies have been previously applied in this specific domain, with variable classification performance (see Byriel, 1999; Martin, 2003; Ampazis et al , 2004; Tsakonas et al , 2004; Jantzen et al , 2005; Norup, 2005; Dounias et al , 2006).…”