A novel online Variance Based Instance Selection (VBIS) method for efficient atypicality detection in chest radiographs

Mohammad Alzubaidi, Vineeth Balasubramanian, Ameet Patel, Sethuraman Panchanathan, John A. Black

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Chest radiographs are complex, heterogeneous medical images that depict many different types of tissues, and many different types of abnormalities. A radiologist develops a sense of what visual textures are typical for each anatomic region within chest radiographs by viewing a large set of "normal" radiographs over a period of years. As a result, an expert radiologist is able to readily detect atypical features. In our previous research, we modeled this type of learning by (1) collecting a large set of "normal" chest radiographs, (2) extracting local textural and contour features from anatomical regions within these radiographs, in the form of high-dimensional feature vectors, (3) using a distance-based transductive machine learning method to learn what it typical for each anatomical region, and (4) computing atypicality scores for the anatomical regions in test radiographs. That research demonstrated that the transductive One-Nearest-Neighbor (1NN) method was effective for identifying atypical regions in chest radiographs. However, the large set of training instances (and the need to compute a distance to each of these instances in a high dimensional space) made the transductive method computationally expensive. This paper discusses a novel online Variance Based Instance Selection (VBIS) method for use with the Nearest Neighbor classifier, that (1) substantially reduced the computational cost of the transductive 1NN method, while maintaining a high level of effectiveness in identifying regions of chest radiographs with atypical content, and (2) allowed the incremental incorporation of training data from new informative chest radiographs as they are encountered in day-to-day clinical work.

Original languageEnglish (US)
Title of host publicationMedical Imaging 2012: Image Processing
Volume8314
DOIs
StatePublished - 2012
Externally publishedYes
EventMedical Imaging 2012: Image Processing - San Diego, CA, United States
Duration: Feb 6 2012Feb 9 2012

Other

OtherMedical Imaging 2012: Image Processing
CountryUnited States
CitySan Diego, CA
Period2/6/122/9/12

Fingerprint

chest
Thorax
Learning systems
Classifiers
Textures
Tissue
education
Costs
machine learning
abnormalities
classifiers
Research
learning
textures
Learning
costs
Costs and Cost Analysis

Keywords

  • Anomaly detection
  • Atypicality detection
  • Chest x-rays
  • Computer aided diagnosis
  • Instance selection
  • Machine learning
  • Nearest neighbor
  • Radiology training

ASJC Scopus subject areas

  • Atomic and Molecular Physics, and Optics
  • Electronic, Optical and Magnetic Materials
  • Biomaterials
  • Radiology Nuclear Medicine and imaging

Cite this

Alzubaidi, M., Balasubramanian, V., Patel, A., Panchanathan, S., & Black, J. A. (2012). A novel online Variance Based Instance Selection (VBIS) method for efficient atypicality detection in chest radiographs. In Medical Imaging 2012: Image Processing (Vol. 8314). [83144Z] https://doi.org/10.1117/12.911154

A novel online Variance Based Instance Selection (VBIS) method for efficient atypicality detection in chest radiographs. / Alzubaidi, Mohammad; Balasubramanian, Vineeth; Patel, Ameet; Panchanathan, Sethuraman; Black, John A.

Medical Imaging 2012: Image Processing. Vol. 8314 2012. 83144Z.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Alzubaidi, M, Balasubramanian, V, Patel, A, Panchanathan, S & Black, JA 2012, A novel online Variance Based Instance Selection (VBIS) method for efficient atypicality detection in chest radiographs. in Medical Imaging 2012: Image Processing. vol. 8314, 83144Z, Medical Imaging 2012: Image Processing, San Diego, CA, United States, 2/6/12. https://doi.org/10.1117/12.911154
Alzubaidi M, Balasubramanian V, Patel A, Panchanathan S, Black JA. A novel online Variance Based Instance Selection (VBIS) method for efficient atypicality detection in chest radiographs. In Medical Imaging 2012: Image Processing. Vol. 8314. 2012. 83144Z https://doi.org/10.1117/12.911154
Alzubaidi, Mohammad ; Balasubramanian, Vineeth ; Patel, Ameet ; Panchanathan, Sethuraman ; Black, John A. / A novel online Variance Based Instance Selection (VBIS) method for efficient atypicality detection in chest radiographs. Medical Imaging 2012: Image Processing. Vol. 8314 2012.
@inproceedings{16dadea3a2cb4471b136be2e1f771a73,
title = "A novel online Variance Based Instance Selection (VBIS) method for efficient atypicality detection in chest radiographs",
abstract = "Chest radiographs are complex, heterogeneous medical images that depict many different types of tissues, and many different types of abnormalities. A radiologist develops a sense of what visual textures are typical for each anatomic region within chest radiographs by viewing a large set of {"}normal{"} radiographs over a period of years. As a result, an expert radiologist is able to readily detect atypical features. In our previous research, we modeled this type of learning by (1) collecting a large set of {"}normal{"} chest radiographs, (2) extracting local textural and contour features from anatomical regions within these radiographs, in the form of high-dimensional feature vectors, (3) using a distance-based transductive machine learning method to learn what it typical for each anatomical region, and (4) computing atypicality scores for the anatomical regions in test radiographs. That research demonstrated that the transductive One-Nearest-Neighbor (1NN) method was effective for identifying atypical regions in chest radiographs. However, the large set of training instances (and the need to compute a distance to each of these instances in a high dimensional space) made the transductive method computationally expensive. This paper discusses a novel online Variance Based Instance Selection (VBIS) method for use with the Nearest Neighbor classifier, that (1) substantially reduced the computational cost of the transductive 1NN method, while maintaining a high level of effectiveness in identifying regions of chest radiographs with atypical content, and (2) allowed the incremental incorporation of training data from new informative chest radiographs as they are encountered in day-to-day clinical work.",
keywords = "Anomaly detection, Atypicality detection, Chest x-rays, Computer aided diagnosis, Instance selection, Machine learning, Nearest neighbor, Radiology training",
author = "Mohammad Alzubaidi and Vineeth Balasubramanian and Ameet Patel and Sethuraman Panchanathan and Black, {John A.}",
year = "2012",
doi = "10.1117/12.911154",
language = "English (US)",
isbn = "9780819489630",
volume = "8314",
booktitle = "Medical Imaging 2012: Image Processing",

}

TY - GEN

T1 - A novel online Variance Based Instance Selection (VBIS) method for efficient atypicality detection in chest radiographs

AU - Alzubaidi, Mohammad

AU - Balasubramanian, Vineeth

AU - Patel, Ameet

AU - Panchanathan, Sethuraman

AU - Black, John A.

PY - 2012

Y1 - 2012

N2 - Chest radiographs are complex, heterogeneous medical images that depict many different types of tissues, and many different types of abnormalities. A radiologist develops a sense of what visual textures are typical for each anatomic region within chest radiographs by viewing a large set of "normal" radiographs over a period of years. As a result, an expert radiologist is able to readily detect atypical features. In our previous research, we modeled this type of learning by (1) collecting a large set of "normal" chest radiographs, (2) extracting local textural and contour features from anatomical regions within these radiographs, in the form of high-dimensional feature vectors, (3) using a distance-based transductive machine learning method to learn what it typical for each anatomical region, and (4) computing atypicality scores for the anatomical regions in test radiographs. That research demonstrated that the transductive One-Nearest-Neighbor (1NN) method was effective for identifying atypical regions in chest radiographs. However, the large set of training instances (and the need to compute a distance to each of these instances in a high dimensional space) made the transductive method computationally expensive. This paper discusses a novel online Variance Based Instance Selection (VBIS) method for use with the Nearest Neighbor classifier, that (1) substantially reduced the computational cost of the transductive 1NN method, while maintaining a high level of effectiveness in identifying regions of chest radiographs with atypical content, and (2) allowed the incremental incorporation of training data from new informative chest radiographs as they are encountered in day-to-day clinical work.

AB - Chest radiographs are complex, heterogeneous medical images that depict many different types of tissues, and many different types of abnormalities. A radiologist develops a sense of what visual textures are typical for each anatomic region within chest radiographs by viewing a large set of "normal" radiographs over a period of years. As a result, an expert radiologist is able to readily detect atypical features. In our previous research, we modeled this type of learning by (1) collecting a large set of "normal" chest radiographs, (2) extracting local textural and contour features from anatomical regions within these radiographs, in the form of high-dimensional feature vectors, (3) using a distance-based transductive machine learning method to learn what it typical for each anatomical region, and (4) computing atypicality scores for the anatomical regions in test radiographs. That research demonstrated that the transductive One-Nearest-Neighbor (1NN) method was effective for identifying atypical regions in chest radiographs. However, the large set of training instances (and the need to compute a distance to each of these instances in a high dimensional space) made the transductive method computationally expensive. This paper discusses a novel online Variance Based Instance Selection (VBIS) method for use with the Nearest Neighbor classifier, that (1) substantially reduced the computational cost of the transductive 1NN method, while maintaining a high level of effectiveness in identifying regions of chest radiographs with atypical content, and (2) allowed the incremental incorporation of training data from new informative chest radiographs as they are encountered in day-to-day clinical work.

KW - Anomaly detection

KW - Atypicality detection

KW - Chest x-rays

KW - Computer aided diagnosis

KW - Instance selection

KW - Machine learning

KW - Nearest neighbor

KW - Radiology training

UR - http://www.scopus.com/inward/record.url?scp=84860754372&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84860754372&partnerID=8YFLogxK

U2 - 10.1117/12.911154

DO - 10.1117/12.911154

M3 - Conference contribution

AN - SCOPUS:84860754372

SN - 9780819489630

VL - 8314

BT - Medical Imaging 2012: Image Processing

ER -