Bayesian Bacterial Detection Using Irregularly Sampled Optical Endomicroscopy Images

Ahmed Karam Mohammed Abdelkarim, Yoann Altmann, Ahsan Akram, Paul McCool, Antonios Perperidis, Kevin Dhaliwal, Stephen McLaughlin

Research output: Contribution to journalArticle

Abstract

Pneumonia is a major cause of morbidity and mortality of patients in intensive care. Rapid determination of the presence and gram status of the pathogenic bacteria in the distal lung may enable a more tailored treatment regime. Optical Endomicroscopy (OEM) is an emerging medical imaging platform with preclinical and clinical utility. Pulmonary OEM via multi-core fibre bundles has the potential to provide in vivo, in situ, fluorescent molecular signatures of the causes of infection and inflammation. This paper presents a Bayesian approach for bacterial detection in OEM images. The model considered assumes that the observed pixel fluorescence is a linear combination of the actual intensity value associated with tissues or background, corrupted by additive Gaussian noise and potentially by an additional sparse outlier term modelling anomalies (bacteria). The bacteria detection problem is formulated in a Bayesian framework and prior distributions are assigned to the unknown model parameters. A Markov chain Monte Carlo algorithm based on a partially collapsed Gibbs sampler is used to sample the posterior distribution of the unknown parameters. The proposed algorithm is first validated by simulations conducted using synthetic datasets for which good performance is obtained. Analysis is then conducted using two ex vivo lung datasets in which fluorescently labelled bacteria are present in the distal lung. A good correlation between bacteria counts identified by a trained clinician and those of the proposed method, which detects most of the manually annotated regions, is observed.
Original languageEnglish
Pages (from-to)18-31
Number of pages14
JournalMedical Image Analysis
Volume57
Early online date24 Jun 2019
DOIs
Publication statusPublished - Oct 2019

Fingerprint

bacterium
pneumonia
morbidity
Markov chain
outlier
sampler
pixel
fluorescence
detection
anomaly
mortality
modeling
simulation
parameter
distribution

Keywords

  • Bacteria detection
  • Bayesian estimation
  • Irregular spatial sampling
  • Optical endomicroscopy
  • Outlier detection

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging
  • Computer Vision and Pattern Recognition
  • Health Informatics
  • Computer Graphics and Computer-Aided Design

Cite this

Karam Mohammed Abdelkarim, Ahmed ; Altmann, Yoann ; Akram, Ahsan ; McCool, Paul ; Perperidis, Antonios ; Dhaliwal, Kevin ; McLaughlin, Stephen. / Bayesian Bacterial Detection Using Irregularly Sampled Optical Endomicroscopy Images. In: Medical Image Analysis. 2019 ; Vol. 57. pp. 18-31.
@article{3b1b59430f4047b2950533e36b741e85,
title = "Bayesian Bacterial Detection Using Irregularly Sampled Optical Endomicroscopy Images",
abstract = "Pneumonia is a major cause of morbidity and mortality of patients in intensive care. Rapid determination of the presence and gram status of the pathogenic bacteria in the distal lung may enable a more tailored treatment regime. Optical Endomicroscopy (OEM) is an emerging medical imaging platform with preclinical and clinical utility. Pulmonary OEM via multi-core fibre bundles has the potential to provide in vivo, in situ, fluorescent molecular signatures of the causes of infection and inflammation. This paper presents a Bayesian approach for bacterial detection in OEM images. The model considered assumes that the observed pixel fluorescence is a linear combination of the actual intensity value associated with tissues or background, corrupted by additive Gaussian noise and potentially by an additional sparse outlier term modelling anomalies (bacteria). The bacteria detection problem is formulated in a Bayesian framework and prior distributions are assigned to the unknown model parameters. A Markov chain Monte Carlo algorithm based on a partially collapsed Gibbs sampler is used to sample the posterior distribution of the unknown parameters. The proposed algorithm is first validated by simulations conducted using synthetic datasets for which good performance is obtained. Analysis is then conducted using two ex vivo lung datasets in which fluorescently labelled bacteria are present in the distal lung. A good correlation between bacteria counts identified by a trained clinician and those of the proposed method, which detects most of the manually annotated regions, is observed.",
keywords = "Bacteria detection, Bayesian estimation, Irregular spatial sampling, Optical endomicroscopy, Outlier detection",
author = "{Karam Mohammed Abdelkarim}, Ahmed and Yoann Altmann and Ahsan Akram and Paul McCool and Antonios Perperidis and Kevin Dhaliwal and Stephen McLaughlin",
year = "2019",
month = "10",
doi = "10.1016/j.media.2019.06.009",
language = "English",
volume = "57",
pages = "18--31",
journal = "Medical Image Analysis",
issn = "1361-8415",
publisher = "Elsevier",

}

Bayesian Bacterial Detection Using Irregularly Sampled Optical Endomicroscopy Images. / Karam Mohammed Abdelkarim, Ahmed; Altmann, Yoann; Akram, Ahsan; McCool, Paul; Perperidis, Antonios; Dhaliwal, Kevin; McLaughlin, Stephen.

In: Medical Image Analysis, Vol. 57, 10.2019, p. 18-31.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Bayesian Bacterial Detection Using Irregularly Sampled Optical Endomicroscopy Images

AU - Karam Mohammed Abdelkarim, Ahmed

AU - Altmann, Yoann

AU - Akram, Ahsan

AU - McCool, Paul

AU - Perperidis, Antonios

AU - Dhaliwal, Kevin

AU - McLaughlin, Stephen

PY - 2019/10

Y1 - 2019/10

N2 - Pneumonia is a major cause of morbidity and mortality of patients in intensive care. Rapid determination of the presence and gram status of the pathogenic bacteria in the distal lung may enable a more tailored treatment regime. Optical Endomicroscopy (OEM) is an emerging medical imaging platform with preclinical and clinical utility. Pulmonary OEM via multi-core fibre bundles has the potential to provide in vivo, in situ, fluorescent molecular signatures of the causes of infection and inflammation. This paper presents a Bayesian approach for bacterial detection in OEM images. The model considered assumes that the observed pixel fluorescence is a linear combination of the actual intensity value associated with tissues or background, corrupted by additive Gaussian noise and potentially by an additional sparse outlier term modelling anomalies (bacteria). The bacteria detection problem is formulated in a Bayesian framework and prior distributions are assigned to the unknown model parameters. A Markov chain Monte Carlo algorithm based on a partially collapsed Gibbs sampler is used to sample the posterior distribution of the unknown parameters. The proposed algorithm is first validated by simulations conducted using synthetic datasets for which good performance is obtained. Analysis is then conducted using two ex vivo lung datasets in which fluorescently labelled bacteria are present in the distal lung. A good correlation between bacteria counts identified by a trained clinician and those of the proposed method, which detects most of the manually annotated regions, is observed.

AB - Pneumonia is a major cause of morbidity and mortality of patients in intensive care. Rapid determination of the presence and gram status of the pathogenic bacteria in the distal lung may enable a more tailored treatment regime. Optical Endomicroscopy (OEM) is an emerging medical imaging platform with preclinical and clinical utility. Pulmonary OEM via multi-core fibre bundles has the potential to provide in vivo, in situ, fluorescent molecular signatures of the causes of infection and inflammation. This paper presents a Bayesian approach for bacterial detection in OEM images. The model considered assumes that the observed pixel fluorescence is a linear combination of the actual intensity value associated with tissues or background, corrupted by additive Gaussian noise and potentially by an additional sparse outlier term modelling anomalies (bacteria). The bacteria detection problem is formulated in a Bayesian framework and prior distributions are assigned to the unknown model parameters. A Markov chain Monte Carlo algorithm based on a partially collapsed Gibbs sampler is used to sample the posterior distribution of the unknown parameters. The proposed algorithm is first validated by simulations conducted using synthetic datasets for which good performance is obtained. Analysis is then conducted using two ex vivo lung datasets in which fluorescently labelled bacteria are present in the distal lung. A good correlation between bacteria counts identified by a trained clinician and those of the proposed method, which detects most of the manually annotated regions, is observed.

KW - Bacteria detection

KW - Bayesian estimation

KW - Irregular spatial sampling

KW - Optical endomicroscopy

KW - Outlier detection

UR - http://www.scopus.com/inward/record.url?scp=85067949911&partnerID=8YFLogxK

U2 - 10.1016/j.media.2019.06.009

DO - 10.1016/j.media.2019.06.009

M3 - Article

VL - 57

SP - 18

EP - 31

JO - Medical Image Analysis

JF - Medical Image Analysis

SN - 1361-8415

ER -