Simultaneous sparsity model for histopathological image representation and classification

Umamahesh Srinivas, Hojjat Seyed Mousavi, Vishal Monga, Arthur Hattel, Bhushan M. Jayarao

Research output: Contribution to journalArticle

48 Citations (Scopus)

Abstract

The multi-channel nature of digital histopathological images presents an opportunity to exploit the correlated color channel information for better image modeling. Inspired by recent work in sparsity for single channel image classification, we propose a new simultaneous sparsity model for multi-channel histopathological image representation and classification (SHIRC). Essentially, we represent a histopathological image as a sparse linear combination of training examples under suitable channel-wise constraints. Classification is performed by solving a newly formulated simultaneous sparsity-based optimization problem. A practical challenge is the correspondence of image objects (cellular and nuclear structures) at different spatial locations in the image. We propose a robust locally adaptive variant of SHIRC (LA-SHIRC) to tackle this issue. Experiments on two challenging real-world image data sets: 1) mammalian tissue images acquired by pathologists of the animal diagnostics lab (ADL) at Pennsylvania State University, and 2) human intraductal breast lesions, reveal the merits of our proposal over state-of-the-art alternatives. Further, we demonstrate that LA-SHIRC exhibits a more graceful decay in classification accuracy against the number of training images which is highly desirable in practice where generous training per class is often not available.

Original languageEnglish (US)
Article number6739999
Pages (from-to)1163-1179
Number of pages17
JournalIEEE Transactions on Medical Imaging
Volume33
Issue number5
DOIs
StatePublished - Jan 1 2014

Fingerprint

Image classification
Animals
Cellular Structures
Tissue
Color
Breast
Experiments
Pathologists
Datasets

All Science Journal Classification (ASJC) codes

  • Software
  • Radiological and Ultrasound Technology
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

@article{4468855d846447ff83d7567ae613fac3,
title = "Simultaneous sparsity model for histopathological image representation and classification",
abstract = "The multi-channel nature of digital histopathological images presents an opportunity to exploit the correlated color channel information for better image modeling. Inspired by recent work in sparsity for single channel image classification, we propose a new simultaneous sparsity model for multi-channel histopathological image representation and classification (SHIRC). Essentially, we represent a histopathological image as a sparse linear combination of training examples under suitable channel-wise constraints. Classification is performed by solving a newly formulated simultaneous sparsity-based optimization problem. A practical challenge is the correspondence of image objects (cellular and nuclear structures) at different spatial locations in the image. We propose a robust locally adaptive variant of SHIRC (LA-SHIRC) to tackle this issue. Experiments on two challenging real-world image data sets: 1) mammalian tissue images acquired by pathologists of the animal diagnostics lab (ADL) at Pennsylvania State University, and 2) human intraductal breast lesions, reveal the merits of our proposal over state-of-the-art alternatives. Further, we demonstrate that LA-SHIRC exhibits a more graceful decay in classification accuracy against the number of training images which is highly desirable in practice where generous training per class is often not available.",
author = "Umamahesh Srinivas and Mousavi, {Hojjat Seyed} and Vishal Monga and Arthur Hattel and Jayarao, {Bhushan M.}",
year = "2014",
month = "1",
day = "1",
doi = "10.1109/TMI.2014.2306173",
language = "English (US)",
volume = "33",
pages = "1163--1179",
journal = "IEEE Transactions on Medical Imaging",
issn = "0278-0062",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "5",

}

Simultaneous sparsity model for histopathological image representation and classification. / Srinivas, Umamahesh; Mousavi, Hojjat Seyed; Monga, Vishal; Hattel, Arthur; Jayarao, Bhushan M.

In: IEEE Transactions on Medical Imaging, Vol. 33, No. 5, 6739999, 01.01.2014, p. 1163-1179.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Simultaneous sparsity model for histopathological image representation and classification

AU - Srinivas, Umamahesh

AU - Mousavi, Hojjat Seyed

AU - Monga, Vishal

AU - Hattel, Arthur

AU - Jayarao, Bhushan M.

PY - 2014/1/1

Y1 - 2014/1/1

N2 - The multi-channel nature of digital histopathological images presents an opportunity to exploit the correlated color channel information for better image modeling. Inspired by recent work in sparsity for single channel image classification, we propose a new simultaneous sparsity model for multi-channel histopathological image representation and classification (SHIRC). Essentially, we represent a histopathological image as a sparse linear combination of training examples under suitable channel-wise constraints. Classification is performed by solving a newly formulated simultaneous sparsity-based optimization problem. A practical challenge is the correspondence of image objects (cellular and nuclear structures) at different spatial locations in the image. We propose a robust locally adaptive variant of SHIRC (LA-SHIRC) to tackle this issue. Experiments on two challenging real-world image data sets: 1) mammalian tissue images acquired by pathologists of the animal diagnostics lab (ADL) at Pennsylvania State University, and 2) human intraductal breast lesions, reveal the merits of our proposal over state-of-the-art alternatives. Further, we demonstrate that LA-SHIRC exhibits a more graceful decay in classification accuracy against the number of training images which is highly desirable in practice where generous training per class is often not available.

AB - The multi-channel nature of digital histopathological images presents an opportunity to exploit the correlated color channel information for better image modeling. Inspired by recent work in sparsity for single channel image classification, we propose a new simultaneous sparsity model for multi-channel histopathological image representation and classification (SHIRC). Essentially, we represent a histopathological image as a sparse linear combination of training examples under suitable channel-wise constraints. Classification is performed by solving a newly formulated simultaneous sparsity-based optimization problem. A practical challenge is the correspondence of image objects (cellular and nuclear structures) at different spatial locations in the image. We propose a robust locally adaptive variant of SHIRC (LA-SHIRC) to tackle this issue. Experiments on two challenging real-world image data sets: 1) mammalian tissue images acquired by pathologists of the animal diagnostics lab (ADL) at Pennsylvania State University, and 2) human intraductal breast lesions, reveal the merits of our proposal over state-of-the-art alternatives. Further, we demonstrate that LA-SHIRC exhibits a more graceful decay in classification accuracy against the number of training images which is highly desirable in practice where generous training per class is often not available.

UR - http://www.scopus.com/inward/record.url?scp=84899714454&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84899714454&partnerID=8YFLogxK

U2 - 10.1109/TMI.2014.2306173

DO - 10.1109/TMI.2014.2306173

M3 - Article

C2 - 24770920

AN - SCOPUS:84899714454

VL - 33

SP - 1163

EP - 1179

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

SN - 0278-0062

IS - 5

M1 - 6739999

ER -