A multiresolution enhancement to generic classifiers of subcellular protein location images

Thomas Merryman, Keridon Williams, Gowri Srinivasa, Amina Chebira, Jelena Kovacevic

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

Abstract

We propose an algorithm for the classification of fluorescence microscopy images depicting the spatial distribution of proteins within the cell. The problem is at the forefront of the current trend in biology towards understanding the role and function of all proteins. The importance of protein subcellular location was pointed out by Murphy, whose group produced the first automated system for classification of images depicting these locations, based on diverse feature sets and combinations of classifiers. With the addition of the simplest multiresolution features, the same group obtained the highest reported accuracy of 91.5% for the denoised 2D HeLa data set. Here, we aim to improve upon that system by adding the true power of multiresolution - adaptivity. In the process, we build a system able to work with any feature sets and any classifiers, which we denote as a Generic Classification System (GCS). Our system consists of multiresolution (MR) decomposition in the front, followedby feature computation and classification in each subband, yielding local decisions. This is followed by the crucial step of combining all those local decisions into a global one, while at the same time ensuring that the resulting system does no worse than a no-decomposition one. On a nondenoised data set and a much smaller number of features (a combination of texture and Zernicke moment features) and a neural network classifier, we obtain a high accuracy of 89.8%, effectively proving that the space-frequency localized information in the subbands adds to the discriminative power of the system.

Original languageEnglish (US)
Title of host publication2006 3rd IEEE International Symposium on Biomedical Imaging
Subtitle of host publicationFrom Nano to Macro - Proceedings
Pages570-573
Number of pages4
Volume2006
StatePublished - Nov 17 2006
Event2006 3rd IEEE International Symposium on Biomedical Imaging: From Nano to Macro - Arlington, VA, United States
Duration: Apr 6 2006Apr 9 2006

Other

Other2006 3rd IEEE International Symposium on Biomedical Imaging: From Nano to Macro
CountryUnited States
CityArlington, VA
Period4/6/064/9/06

Fingerprint

Classifiers
Proteins
Decomposition
Fluorescence microscopy
Spatial distribution
Textures
Neural networks

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Merryman, T., Williams, K., Srinivasa, G., Chebira, A., & Kovacevic, J. (2006). A multiresolution enhancement to generic classifiers of subcellular protein location images. In 2006 3rd IEEE International Symposium on Biomedical Imaging: From Nano to Macro - Proceedings (Vol. 2006, pp. 570-573). [1624980]

A multiresolution enhancement to generic classifiers of subcellular protein location images. / Merryman, Thomas; Williams, Keridon; Srinivasa, Gowri; Chebira, Amina; Kovacevic, Jelena.

2006 3rd IEEE International Symposium on Biomedical Imaging: From Nano to Macro - Proceedings. Vol. 2006 2006. p. 570-573 1624980.

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

Merryman, T, Williams, K, Srinivasa, G, Chebira, A & Kovacevic, J 2006, A multiresolution enhancement to generic classifiers of subcellular protein location images. in 2006 3rd IEEE International Symposium on Biomedical Imaging: From Nano to Macro - Proceedings. vol. 2006, 1624980, pp. 570-573, 2006 3rd IEEE International Symposium on Biomedical Imaging: From Nano to Macro, Arlington, VA, United States, 4/6/06.
Merryman T, Williams K, Srinivasa G, Chebira A, Kovacevic J. A multiresolution enhancement to generic classifiers of subcellular protein location images. In 2006 3rd IEEE International Symposium on Biomedical Imaging: From Nano to Macro - Proceedings. Vol. 2006. 2006. p. 570-573. 1624980
Merryman, Thomas ; Williams, Keridon ; Srinivasa, Gowri ; Chebira, Amina ; Kovacevic, Jelena. / A multiresolution enhancement to generic classifiers of subcellular protein location images. 2006 3rd IEEE International Symposium on Biomedical Imaging: From Nano to Macro - Proceedings. Vol. 2006 2006. pp. 570-573
@inproceedings{4de6cb2f3114406698da3eff6a059f8b,
title = "A multiresolution enhancement to generic classifiers of subcellular protein location images",
abstract = "We propose an algorithm for the classification of fluorescence microscopy images depicting the spatial distribution of proteins within the cell. The problem is at the forefront of the current trend in biology towards understanding the role and function of all proteins. The importance of protein subcellular location was pointed out by Murphy, whose group produced the first automated system for classification of images depicting these locations, based on diverse feature sets and combinations of classifiers. With the addition of the simplest multiresolution features, the same group obtained the highest reported accuracy of 91.5{\%} for the denoised 2D HeLa data set. Here, we aim to improve upon that system by adding the true power of multiresolution - adaptivity. In the process, we build a system able to work with any feature sets and any classifiers, which we denote as a Generic Classification System (GCS). Our system consists of multiresolution (MR) decomposition in the front, followedby feature computation and classification in each subband, yielding local decisions. This is followed by the crucial step of combining all those local decisions into a global one, while at the same time ensuring that the resulting system does no worse than a no-decomposition one. On a nondenoised data set and a much smaller number of features (a combination of texture and Zernicke moment features) and a neural network classifier, we obtain a high accuracy of 89.8{\%}, effectively proving that the space-frequency localized information in the subbands adds to the discriminative power of the system.",
author = "Thomas Merryman and Keridon Williams and Gowri Srinivasa and Amina Chebira and Jelena Kovacevic",
year = "2006",
month = "11",
day = "17",
language = "English (US)",
isbn = "0780395778",
volume = "2006",
pages = "570--573",
booktitle = "2006 3rd IEEE International Symposium on Biomedical Imaging",

}

TY - GEN

T1 - A multiresolution enhancement to generic classifiers of subcellular protein location images

AU - Merryman, Thomas

AU - Williams, Keridon

AU - Srinivasa, Gowri

AU - Chebira, Amina

AU - Kovacevic, Jelena

PY - 2006/11/17

Y1 - 2006/11/17

N2 - We propose an algorithm for the classification of fluorescence microscopy images depicting the spatial distribution of proteins within the cell. The problem is at the forefront of the current trend in biology towards understanding the role and function of all proteins. The importance of protein subcellular location was pointed out by Murphy, whose group produced the first automated system for classification of images depicting these locations, based on diverse feature sets and combinations of classifiers. With the addition of the simplest multiresolution features, the same group obtained the highest reported accuracy of 91.5% for the denoised 2D HeLa data set. Here, we aim to improve upon that system by adding the true power of multiresolution - adaptivity. In the process, we build a system able to work with any feature sets and any classifiers, which we denote as a Generic Classification System (GCS). Our system consists of multiresolution (MR) decomposition in the front, followedby feature computation and classification in each subband, yielding local decisions. This is followed by the crucial step of combining all those local decisions into a global one, while at the same time ensuring that the resulting system does no worse than a no-decomposition one. On a nondenoised data set and a much smaller number of features (a combination of texture and Zernicke moment features) and a neural network classifier, we obtain a high accuracy of 89.8%, effectively proving that the space-frequency localized information in the subbands adds to the discriminative power of the system.

AB - We propose an algorithm for the classification of fluorescence microscopy images depicting the spatial distribution of proteins within the cell. The problem is at the forefront of the current trend in biology towards understanding the role and function of all proteins. The importance of protein subcellular location was pointed out by Murphy, whose group produced the first automated system for classification of images depicting these locations, based on diverse feature sets and combinations of classifiers. With the addition of the simplest multiresolution features, the same group obtained the highest reported accuracy of 91.5% for the denoised 2D HeLa data set. Here, we aim to improve upon that system by adding the true power of multiresolution - adaptivity. In the process, we build a system able to work with any feature sets and any classifiers, which we denote as a Generic Classification System (GCS). Our system consists of multiresolution (MR) decomposition in the front, followedby feature computation and classification in each subband, yielding local decisions. This is followed by the crucial step of combining all those local decisions into a global one, while at the same time ensuring that the resulting system does no worse than a no-decomposition one. On a nondenoised data set and a much smaller number of features (a combination of texture and Zernicke moment features) and a neural network classifier, we obtain a high accuracy of 89.8%, effectively proving that the space-frequency localized information in the subbands adds to the discriminative power of the system.

UR - http://www.scopus.com/inward/record.url?scp=33750932166&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33750932166&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0780395778

SN - 9780780395770

VL - 2006

SP - 570

EP - 573

BT - 2006 3rd IEEE International Symposium on Biomedical Imaging

ER -