Cargando…

Low cost automated whole smear microscopy screening system for detection of acid fast bacilli

BACKGROUND: In countries with high tuberculosis (TB) burden, there is urgent need for rapid, large-scale screening to detect smear-positive patients. We developed a computer-aided whole smear screening system that focuses in real-time, captures images and provides diagnostic grading, for both bright...

Descripción completa

Detalles Bibliográficos
Autores principales: Law, Yan Nei, Jian, Hanbin, Lo, Norman W. S., Ip, Margaret, Chan, Mia Mei Yuk, Kam, Kai Man, Wu, Xiaohua
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Public Library of Science 2018
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5777646/
https://www.ncbi.nlm.nih.gov/pubmed/29357378
http://dx.doi.org/10.1371/journal.pone.0190988
_version_ 1783294219456086016
author Law, Yan Nei
Jian, Hanbin
Lo, Norman W. S.
Ip, Margaret
Chan, Mia Mei Yuk
Kam, Kai Man
Wu, Xiaohua
author_facet Law, Yan Nei
Jian, Hanbin
Lo, Norman W. S.
Ip, Margaret
Chan, Mia Mei Yuk
Kam, Kai Man
Wu, Xiaohua
author_sort Law, Yan Nei
collection PubMed
description BACKGROUND: In countries with high tuberculosis (TB) burden, there is urgent need for rapid, large-scale screening to detect smear-positive patients. We developed a computer-aided whole smear screening system that focuses in real-time, captures images and provides diagnostic grading, for both bright-field and fluorescence microscopy for detection of acid-fast-bacilli (AFB) from respiratory specimens. OBJECTIVES: To evaluate the performance of dual-mode screening system in AFB diagnostic algorithms on concentrated smears with auramine O (AO) staining, as well as direct smears with AO and Ziehl-Neelsen (ZN) staining, using mycobacterial culture results as gold standard. METHODS: Adult patient sputum samples requesting for M. tuberculosis cultures were divided into three batches for staining: direct AO-stained, direct ZN-stained and concentrated smears AO-stained. All slides were graded by an experienced microscopist, in parallel with the automated whole smear screening system. Sensitivity and specificity of a TB diagnostic algorithm in using the screening system alone, and in combination with a microscopist, were evaluated. RESULTS: Of 488 direct AO-stained smears, 228 were culture positive. These yielded a sensitivity of 81.6% and specificity of 74.2%. Of 334 direct smears with ZN staining, 142 were culture positive, which gave a sensitivity of 70.4% and specificity of 76.6%. Of 505 concentrated smears with AO staining, 250 were culture positive, giving a sensitivity of 86.4% and specificity of 71.0%. To further improve performance, machine grading was confirmed by manual smear grading when the number of AFBs detected fell within an uncertainty range. These combined results gave significant improvement in specificity (AO-direct:85.4%; ZN-direct:85.4%; AO-concentrated:92.5%) and slight improvement in sensitivity while requiring only limited manual workload. CONCLUSION: Our system achieved high sensitivity without substantially compromising specificity when compared to culture results. Significant improvement in specificity was obtained when uncertain results were confirmed by manual smear grading. This approach had potential to substantially reduce workload of microscopists in high burden countries.
format Online
Article
Text
id pubmed-5777646
institution National Center for Biotechnology Information
language English
publishDate 2018
publisher Public Library of Science
record_format MEDLINE/PubMed
spelling pubmed-57776462018-02-05 Low cost automated whole smear microscopy screening system for detection of acid fast bacilli Law, Yan Nei Jian, Hanbin Lo, Norman W. S. Ip, Margaret Chan, Mia Mei Yuk Kam, Kai Man Wu, Xiaohua PLoS One Research Article BACKGROUND: In countries with high tuberculosis (TB) burden, there is urgent need for rapid, large-scale screening to detect smear-positive patients. We developed a computer-aided whole smear screening system that focuses in real-time, captures images and provides diagnostic grading, for both bright-field and fluorescence microscopy for detection of acid-fast-bacilli (AFB) from respiratory specimens. OBJECTIVES: To evaluate the performance of dual-mode screening system in AFB diagnostic algorithms on concentrated smears with auramine O (AO) staining, as well as direct smears with AO and Ziehl-Neelsen (ZN) staining, using mycobacterial culture results as gold standard. METHODS: Adult patient sputum samples requesting for M. tuberculosis cultures were divided into three batches for staining: direct AO-stained, direct ZN-stained and concentrated smears AO-stained. All slides were graded by an experienced microscopist, in parallel with the automated whole smear screening system. Sensitivity and specificity of a TB diagnostic algorithm in using the screening system alone, and in combination with a microscopist, were evaluated. RESULTS: Of 488 direct AO-stained smears, 228 were culture positive. These yielded a sensitivity of 81.6% and specificity of 74.2%. Of 334 direct smears with ZN staining, 142 were culture positive, which gave a sensitivity of 70.4% and specificity of 76.6%. Of 505 concentrated smears with AO staining, 250 were culture positive, giving a sensitivity of 86.4% and specificity of 71.0%. To further improve performance, machine grading was confirmed by manual smear grading when the number of AFBs detected fell within an uncertainty range. These combined results gave significant improvement in specificity (AO-direct:85.4%; ZN-direct:85.4%; AO-concentrated:92.5%) and slight improvement in sensitivity while requiring only limited manual workload. CONCLUSION: Our system achieved high sensitivity without substantially compromising specificity when compared to culture results. Significant improvement in specificity was obtained when uncertain results were confirmed by manual smear grading. This approach had potential to substantially reduce workload of microscopists in high burden countries. Public Library of Science 2018-01-22 /pmc/articles/PMC5777646/ /pubmed/29357378 http://dx.doi.org/10.1371/journal.pone.0190988 Text en © 2018 Law et al http://creativecommons.org/licenses/by/4.0/ This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/) , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
spellingShingle Research Article
Law, Yan Nei
Jian, Hanbin
Lo, Norman W. S.
Ip, Margaret
Chan, Mia Mei Yuk
Kam, Kai Man
Wu, Xiaohua
Low cost automated whole smear microscopy screening system for detection of acid fast bacilli
title Low cost automated whole smear microscopy screening system for detection of acid fast bacilli
title_full Low cost automated whole smear microscopy screening system for detection of acid fast bacilli
title_fullStr Low cost automated whole smear microscopy screening system for detection of acid fast bacilli
title_full_unstemmed Low cost automated whole smear microscopy screening system for detection of acid fast bacilli
title_short Low cost automated whole smear microscopy screening system for detection of acid fast bacilli
title_sort low cost automated whole smear microscopy screening system for detection of acid fast bacilli
topic Research Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5777646/
https://www.ncbi.nlm.nih.gov/pubmed/29357378
http://dx.doi.org/10.1371/journal.pone.0190988
work_keys_str_mv AT lawyannei lowcostautomatedwholesmearmicroscopyscreeningsystemfordetectionofacidfastbacilli
AT jianhanbin lowcostautomatedwholesmearmicroscopyscreeningsystemfordetectionofacidfastbacilli
AT lonormanws lowcostautomatedwholesmearmicroscopyscreeningsystemfordetectionofacidfastbacilli
AT ipmargaret lowcostautomatedwholesmearmicroscopyscreeningsystemfordetectionofacidfastbacilli
AT chanmiameiyuk lowcostautomatedwholesmearmicroscopyscreeningsystemfordetectionofacidfastbacilli
AT kamkaiman lowcostautomatedwholesmearmicroscopyscreeningsystemfordetectionofacidfastbacilli
AT wuxiaohua lowcostautomatedwholesmearmicroscopyscreeningsystemfordetectionofacidfastbacilli