Please use this identifier to cite or link to this item:
Type: Artigo
Title: Medical-laboratory algorithm WTH-MO for segmentation of digital images of blood cells: a new methodology for making hemograms
Author: Monteiro, Ana M.
Iano, Yuzo
França, Reinaldo
Arthur, Rangel
Abstract: Inter-disciplinarity has been responsible for solving several problems in the medical areas, being that many of which are derived from the use of engineering techniques. One problem that affects a large part of the world's population is the difficulty of access to health, often caused by a lack of financial resources. One of the most requested medical tests due to its ability to detect a wide range of diseases is the hemogram. However, this methodology is often inaccessible due to its cost, which for disadvantaged populations can be seen as high. Based on this, the present study aims to develop an algorithm for the segmentation of digital images of blood smears, which is capable of detecting and counting erythrocytes, leukocytes and platelets simultaneously, with high accuracy and low cost. The algorithm was developed in the Matlab software environment, based on the Watershed Transform, Transformed Hough methodologies and operations methodologies. The hybrid algorithm was named WHT-MO. The experiments were conducted using 10 digital blood smear images. The results showed high accuracy (90% and 100%) in the detection of the three cell types. Therefore, the WHT-MO algorithm demonstrates that the hemogram can be performed using a simple computer, dispensing with the need to acquire high-cost hematological equipment, which directly impacts the final price of the examination for patients. In addition, high accuracy demonstrates that it is possible to perform the hemogram at a reduced cost without loss of quality and reliability
Subject: Algoritmos de computador
Country: Reino Unido
Editor: UK Simulation Society
Rights: Aberto
Identifier DOI: 10.5013/ijssst.a.20.s1.19
Date Issue: 2019
Appears in Collections:FEEC - Artigos e Outros Documentos
FT - Artigos e Outros Documentos

Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.