Segmentation of human intestinal parasites from microscopy images using localized mean-separation based active contour model

• Published in: Biomedical engineering letters Vol. 3; no. 3; pp. 179 - 189
• Main Authors: Rema, M., Nair, Madhu S.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01-09-2013; 대한의용생체공학회
• Subjects: Biological and Medical Physics; Biomedical Engineering and Bioengineering; Biomedicine; Biophysics; Engineering; Medical and Radiation Physics; Original Article; 의공학; Human Intestinal Parasites; Active Contour Model; Segmentation; Thresholding; Bright field Microscopy
• ISSN: 2093-9868; 2093-985X
• DOI: 10.1007/s13534-013-0101-3

Purpose
Identification of human intestinal parasites from microscopy images of fecal sample is an important and time consuming process in the diagnosis of intestinal parasitosis. Automatic image processing can be applied to segment and identify the parasite but the presence of fecal impurities makes this process a challenging one. This paper presents a framework for segmentation of bright field microscopy images of fecal sample that contain both parasites and impurities.
Methods
The proposed framework uses thresholding, morphological opening and Active Contour Model (ACM). Contour is initialized using thresholding and morphological opening and the contour is evolved using Localized Mean-Separation based Active Contour Model (LMS-ACM).
Results
This framework is simple, fast and yields good results even when the parasites are overlapped with impurities. The accuracy of the method is tested by comparing the results with manually segmented images.
Conclusions
As accurate segmentation of the objects is the first and important step in identification process, this work is a promising approach towards the automatic diagnosis of human intestinal parasitosis.