Multiclass Classification of Cervical Cancer Tissues by Hidden Markov Model

TL;DR

This paper presents a hidden Markov model approach for multiclass classification of cervical cancer tissues using time series data from microscopy images, achieving promising accuracy.

Contribution

The study introduces a novel application of hidden Markov models for multiclass tissue classification based on optical image analysis.

Findings

Higher accuracy in multiclass classification

Effective use of time series from refractive index fluctuations

Validated on differential interference contrast images

Abstract

In this paper, we report a hidden Markov model based multiclass classification of cervical cancer tissues. This model has been validated directly over time series generated by the medium refractive index fluctuations extracted from differential interference contrast images of healthy and different stages of cancer tissues. The method shows promising results for multiclass classification with higher accuracy.