Stroke Risk Stratification and its Validation using Ultrasonic Echolucent Carotid Wall Plaque Morphology

A Machine Learning Paradigm

Tadashi Araki, Pankaj K. Jain, Harman S. Suri, Narendra D. Londhe, Nobutaka Ikeda, Ayman El-Baz, Vimal K. Shrivastava, Luca Saba, Andrew Nicolaides, Shoaib Shafique, John R. Laird, Ajay Gupta, Jasjit S. Suri

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

Stroke risk stratification based on grayscale morphology of the ultrasound carotid wall has recently been shown to have a promise in classification of high risk versus low risk plaque or symptomatic versus asymptomatic plaques. In previous studies, this stratification has been mainly based on analysis of the far wall of the carotid artery. Due to the multifocal nature of atherosclerotic disease, the plaque growth is not restricted to the far wall alone. This paper presents a new approach for stroke risk assessment by integrating assessment of both the near and far walls of the carotid artery using grayscale morphology of the plaque. Further, this paper presents a scientific validation system for stroke risk assessment. Both these innovations have never been presented before. The methodology consists of an automated segmentation system of the near wall and far wall regions in grayscale carotid B-mode ultrasound scans. Sixteen grayscale texture features are computed, and fed into the machine learning system. The training system utilizes the lumen diameter to create ground truth labels for the stratification of stroke risk. The cross-validation procedure is adapted in order to obtain the machine learning testing classification accuracy through the use of three sets of partition protocols: (5, 10, and Jack Knife). The mean classification accuracy over all the sets of partition protocols for the automated system in the far and near walls is 95.08% and 93.47%, respectively. The corresponding accuracies for the manual system are 94.06% and 92.02%, respectively. The precision of merit of the automated machine learning system when compared against manual risk assessment system are 98.05% and 97.53% for the far and near walls, respectively. The ROC of the risk assessment system for the far and near walls is close to 1.0 demonstrating high accuracy.

Original languageEnglish (US)
Pages (from-to)77-96
Number of pages20
JournalComputers in Biology and Medicine
Volume80
DOIs
StatePublished - Jan 1 2017

Fingerprint

Ultrasonics
Learning systems
Risk assessment
Stroke
Carotid Arteries
Atherosclerotic Plaques
Jacks
Labels
Textures
Innovation
Machine Learning
Growth
Testing

Keywords

  • Carotid wall
  • Far
  • Machine learning
  • Near
  • Precision of merit
  • ROC
  • Segmentation
  • Stroke
  • Ultrasound

ASJC Scopus subject areas

  • Computer Science Applications
  • Health Informatics

Cite this

Stroke Risk Stratification and its Validation using Ultrasonic Echolucent Carotid Wall Plaque Morphology : A Machine Learning Paradigm. / Araki, Tadashi; Jain, Pankaj K.; Suri, Harman S.; Londhe, Narendra D.; Ikeda, Nobutaka; El-Baz, Ayman; Shrivastava, Vimal K.; Saba, Luca; Nicolaides, Andrew; Shafique, Shoaib; Laird, John R.; Gupta, Ajay; Suri, Jasjit S.

In: Computers in Biology and Medicine, Vol. 80, 01.01.2017, p. 77-96.

Research output: Contribution to journalArticle

Araki, T, Jain, PK, Suri, HS, Londhe, ND, Ikeda, N, El-Baz, A, Shrivastava, VK, Saba, L, Nicolaides, A, Shafique, S, Laird, JR, Gupta, A & Suri, JS 2017, 'Stroke Risk Stratification and its Validation using Ultrasonic Echolucent Carotid Wall Plaque Morphology: A Machine Learning Paradigm', Computers in Biology and Medicine, vol. 80, pp. 77-96. https://doi.org/10.1016/j.compbiomed.2016.11.011
Araki, Tadashi ; Jain, Pankaj K. ; Suri, Harman S. ; Londhe, Narendra D. ; Ikeda, Nobutaka ; El-Baz, Ayman ; Shrivastava, Vimal K. ; Saba, Luca ; Nicolaides, Andrew ; Shafique, Shoaib ; Laird, John R. ; Gupta, Ajay ; Suri, Jasjit S. / Stroke Risk Stratification and its Validation using Ultrasonic Echolucent Carotid Wall Plaque Morphology : A Machine Learning Paradigm. In: Computers in Biology and Medicine. 2017 ; Vol. 80. pp. 77-96.
@article{5bf9126903964a22b0a2a2eace12d7c8,
title = "Stroke Risk Stratification and its Validation using Ultrasonic Echolucent Carotid Wall Plaque Morphology: A Machine Learning Paradigm",
abstract = "Stroke risk stratification based on grayscale morphology of the ultrasound carotid wall has recently been shown to have a promise in classification of high risk versus low risk plaque or symptomatic versus asymptomatic plaques. In previous studies, this stratification has been mainly based on analysis of the far wall of the carotid artery. Due to the multifocal nature of atherosclerotic disease, the plaque growth is not restricted to the far wall alone. This paper presents a new approach for stroke risk assessment by integrating assessment of both the near and far walls of the carotid artery using grayscale morphology of the plaque. Further, this paper presents a scientific validation system for stroke risk assessment. Both these innovations have never been presented before. The methodology consists of an automated segmentation system of the near wall and far wall regions in grayscale carotid B-mode ultrasound scans. Sixteen grayscale texture features are computed, and fed into the machine learning system. The training system utilizes the lumen diameter to create ground truth labels for the stratification of stroke risk. The cross-validation procedure is adapted in order to obtain the machine learning testing classification accuracy through the use of three sets of partition protocols: (5, 10, and Jack Knife). The mean classification accuracy over all the sets of partition protocols for the automated system in the far and near walls is 95.08{\%} and 93.47{\%}, respectively. The corresponding accuracies for the manual system are 94.06{\%} and 92.02{\%}, respectively. The precision of merit of the automated machine learning system when compared against manual risk assessment system are 98.05{\%} and 97.53{\%} for the far and near walls, respectively. The ROC of the risk assessment system for the far and near walls is close to 1.0 demonstrating high accuracy.",
keywords = "Carotid wall, Far, Machine learning, Near, Precision of merit, ROC, Segmentation, Stroke, Ultrasound",
author = "Tadashi Araki and Jain, {Pankaj K.} and Suri, {Harman S.} and Londhe, {Narendra D.} and Nobutaka Ikeda and Ayman El-Baz and Shrivastava, {Vimal K.} and Luca Saba and Andrew Nicolaides and Shoaib Shafique and Laird, {John R.} and Ajay Gupta and Suri, {Jasjit S.}",
year = "2017",
month = "1",
day = "1",
doi = "10.1016/j.compbiomed.2016.11.011",
language = "English (US)",
volume = "80",
pages = "77--96",
journal = "Computers in Biology and Medicine",
issn = "0010-4825",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Stroke Risk Stratification and its Validation using Ultrasonic Echolucent Carotid Wall Plaque Morphology

T2 - A Machine Learning Paradigm

AU - Araki, Tadashi

AU - Jain, Pankaj K.

AU - Suri, Harman S.

AU - Londhe, Narendra D.

AU - Ikeda, Nobutaka

AU - El-Baz, Ayman

AU - Shrivastava, Vimal K.

AU - Saba, Luca

AU - Nicolaides, Andrew

AU - Shafique, Shoaib

AU - Laird, John R.

AU - Gupta, Ajay

AU - Suri, Jasjit S.

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Stroke risk stratification based on grayscale morphology of the ultrasound carotid wall has recently been shown to have a promise in classification of high risk versus low risk plaque or symptomatic versus asymptomatic plaques. In previous studies, this stratification has been mainly based on analysis of the far wall of the carotid artery. Due to the multifocal nature of atherosclerotic disease, the plaque growth is not restricted to the far wall alone. This paper presents a new approach for stroke risk assessment by integrating assessment of both the near and far walls of the carotid artery using grayscale morphology of the plaque. Further, this paper presents a scientific validation system for stroke risk assessment. Both these innovations have never been presented before. The methodology consists of an automated segmentation system of the near wall and far wall regions in grayscale carotid B-mode ultrasound scans. Sixteen grayscale texture features are computed, and fed into the machine learning system. The training system utilizes the lumen diameter to create ground truth labels for the stratification of stroke risk. The cross-validation procedure is adapted in order to obtain the machine learning testing classification accuracy through the use of three sets of partition protocols: (5, 10, and Jack Knife). The mean classification accuracy over all the sets of partition protocols for the automated system in the far and near walls is 95.08% and 93.47%, respectively. The corresponding accuracies for the manual system are 94.06% and 92.02%, respectively. The precision of merit of the automated machine learning system when compared against manual risk assessment system are 98.05% and 97.53% for the far and near walls, respectively. The ROC of the risk assessment system for the far and near walls is close to 1.0 demonstrating high accuracy.

AB - Stroke risk stratification based on grayscale morphology of the ultrasound carotid wall has recently been shown to have a promise in classification of high risk versus low risk plaque or symptomatic versus asymptomatic plaques. In previous studies, this stratification has been mainly based on analysis of the far wall of the carotid artery. Due to the multifocal nature of atherosclerotic disease, the plaque growth is not restricted to the far wall alone. This paper presents a new approach for stroke risk assessment by integrating assessment of both the near and far walls of the carotid artery using grayscale morphology of the plaque. Further, this paper presents a scientific validation system for stroke risk assessment. Both these innovations have never been presented before. The methodology consists of an automated segmentation system of the near wall and far wall regions in grayscale carotid B-mode ultrasound scans. Sixteen grayscale texture features are computed, and fed into the machine learning system. The training system utilizes the lumen diameter to create ground truth labels for the stratification of stroke risk. The cross-validation procedure is adapted in order to obtain the machine learning testing classification accuracy through the use of three sets of partition protocols: (5, 10, and Jack Knife). The mean classification accuracy over all the sets of partition protocols for the automated system in the far and near walls is 95.08% and 93.47%, respectively. The corresponding accuracies for the manual system are 94.06% and 92.02%, respectively. The precision of merit of the automated machine learning system when compared against manual risk assessment system are 98.05% and 97.53% for the far and near walls, respectively. The ROC of the risk assessment system for the far and near walls is close to 1.0 demonstrating high accuracy.

KW - Carotid wall

KW - Far

KW - Machine learning

KW - Near

KW - Precision of merit

KW - ROC

KW - Segmentation

KW - Stroke

KW - Ultrasound

UR - http://www.scopus.com/inward/record.url?scp=85000741772&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85000741772&partnerID=8YFLogxK

U2 - 10.1016/j.compbiomed.2016.11.011

DO - 10.1016/j.compbiomed.2016.11.011

M3 - Article

VL - 80

SP - 77

EP - 96

JO - Computers in Biology and Medicine

JF - Computers in Biology and Medicine

SN - 0010-4825

ER -