COMPUTER AIDED IMAGE SEGMENTATION AND CLASSIFICATION FOR VIABLE AND NON-VIABLE TUMOR IDENTIFICATION IN OSTEOSARCOMA

Harish Babu Arunachalam, Rashika Mishra, Bogdan Armaselu, Ovidiu Daescu, Maria Martinez, Patrick J Leavey, Dinesh Rakheja, Kevin B Cederberg, Anita L Sengupta, Molly Ni'suilleabhain

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Osteosarcoma is one of the most common types of bone cancer in children. To gauge the extent of cancer treatment response in the patient after surgical resection, the H&E stained image slides are manually evaluated by pathologists to estimate the percentage of necrosis, a time consuming process prone to observer bias and inaccuracy. Digital image analysis is a potential method to automate this process, thus saving time and providing a more accurate evaluation. The slides are scanned in Aperio Scanscope, converted to digital Whole Slide Images (WSIs) and stored in SVS format. These are high resolution images, of the order of 109 pixels, allowing up to 40X magnification factor. This paper proposes an image segmentation and analysis technique for segmenting tumor and non-tumor regions in histopathological WSIs of osteosarcoma datasets. Our approach is a combination of pixel-based and object-based methods which utilize tumor properties such as nuclei cluster, density, and circularity to classify tumor regions as viable and non-viable. A K-Means clustering technique is used for tumor isolation using color normalization, followed by multi-threshold Otsu segmentation technique to further classify tumor region as viable and non-viable. Then a Flood-fill algorithm is applied to cluster similar pixels into cellular objects and compute cluster data for further analysis of regions under study. To the best of our knowledge this is the first comprehensive solution that is able to produce such a classification for Osteosarcoma cancer. The results are very conclusive in identifying viable and non-viable tumor regions. In our experiments, the accuracy of the discussed approach is 100% in viable tumor and coagulative necrosis identification while it is around 90% for fibrosis and acellular/hypocellular tumor osteoid, for all the sampled datasets used. We expect the developed software to lead to a significant increase in accuracy and decrease in inter-observer variability in assessment of necrosis by the pathologists and a reduction in the time spent by the pathologists in such assessments.

Original languageEnglish (US)
Pages (from-to)195-206
Number of pages12
JournalPacific Symposium on Biocomputing. Pacific Symposium on Biocomputing
Volume22
StatePublished - 2016

Fingerprint

Osteosarcoma
Neoplasms
Necrosis
Observer Variation
Bone Neoplasms
Cluster Analysis
Fibrosis
Software
Color

ASJC Scopus subject areas

  • Medicine(all)

Cite this

COMPUTER AIDED IMAGE SEGMENTATION AND CLASSIFICATION FOR VIABLE AND NON-VIABLE TUMOR IDENTIFICATION IN OSTEOSARCOMA. / Arunachalam, Harish Babu; Mishra, Rashika; Armaselu, Bogdan; Daescu, Ovidiu; Martinez, Maria; Leavey, Patrick J; Rakheja, Dinesh; Cederberg, Kevin B; Sengupta, Anita L; Ni'suilleabhain, Molly.

In: Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing, Vol. 22, 2016, p. 195-206.

Research output: Contribution to journalArticle

@article{7f67d14677164b1d8030c451cf63ee5b,
title = "COMPUTER AIDED IMAGE SEGMENTATION AND CLASSIFICATION FOR VIABLE AND NON-VIABLE TUMOR IDENTIFICATION IN OSTEOSARCOMA",
abstract = "Osteosarcoma is one of the most common types of bone cancer in children. To gauge the extent of cancer treatment response in the patient after surgical resection, the H&E stained image slides are manually evaluated by pathologists to estimate the percentage of necrosis, a time consuming process prone to observer bias and inaccuracy. Digital image analysis is a potential method to automate this process, thus saving time and providing a more accurate evaluation. The slides are scanned in Aperio Scanscope, converted to digital Whole Slide Images (WSIs) and stored in SVS format. These are high resolution images, of the order of 109 pixels, allowing up to 40X magnification factor. This paper proposes an image segmentation and analysis technique for segmenting tumor and non-tumor regions in histopathological WSIs of osteosarcoma datasets. Our approach is a combination of pixel-based and object-based methods which utilize tumor properties such as nuclei cluster, density, and circularity to classify tumor regions as viable and non-viable. A K-Means clustering technique is used for tumor isolation using color normalization, followed by multi-threshold Otsu segmentation technique to further classify tumor region as viable and non-viable. Then a Flood-fill algorithm is applied to cluster similar pixels into cellular objects and compute cluster data for further analysis of regions under study. To the best of our knowledge this is the first comprehensive solution that is able to produce such a classification for Osteosarcoma cancer. The results are very conclusive in identifying viable and non-viable tumor regions. In our experiments, the accuracy of the discussed approach is 100{\%} in viable tumor and coagulative necrosis identification while it is around 90{\%} for fibrosis and acellular/hypocellular tumor osteoid, for all the sampled datasets used. We expect the developed software to lead to a significant increase in accuracy and decrease in inter-observer variability in assessment of necrosis by the pathologists and a reduction in the time spent by the pathologists in such assessments.",
author = "Arunachalam, {Harish Babu} and Rashika Mishra and Bogdan Armaselu and Ovidiu Daescu and Maria Martinez and Leavey, {Patrick J} and Dinesh Rakheja and Cederberg, {Kevin B} and Sengupta, {Anita L} and Molly Ni'suilleabhain",
year = "2016",
language = "English (US)",
volume = "22",
pages = "195--206",
journal = "Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing",
issn = "2335-6936",

}

TY - JOUR

T1 - COMPUTER AIDED IMAGE SEGMENTATION AND CLASSIFICATION FOR VIABLE AND NON-VIABLE TUMOR IDENTIFICATION IN OSTEOSARCOMA

AU - Arunachalam, Harish Babu

AU - Mishra, Rashika

AU - Armaselu, Bogdan

AU - Daescu, Ovidiu

AU - Martinez, Maria

AU - Leavey, Patrick J

AU - Rakheja, Dinesh

AU - Cederberg, Kevin B

AU - Sengupta, Anita L

AU - Ni'suilleabhain, Molly

PY - 2016

Y1 - 2016

N2 - Osteosarcoma is one of the most common types of bone cancer in children. To gauge the extent of cancer treatment response in the patient after surgical resection, the H&E stained image slides are manually evaluated by pathologists to estimate the percentage of necrosis, a time consuming process prone to observer bias and inaccuracy. Digital image analysis is a potential method to automate this process, thus saving time and providing a more accurate evaluation. The slides are scanned in Aperio Scanscope, converted to digital Whole Slide Images (WSIs) and stored in SVS format. These are high resolution images, of the order of 109 pixels, allowing up to 40X magnification factor. This paper proposes an image segmentation and analysis technique for segmenting tumor and non-tumor regions in histopathological WSIs of osteosarcoma datasets. Our approach is a combination of pixel-based and object-based methods which utilize tumor properties such as nuclei cluster, density, and circularity to classify tumor regions as viable and non-viable. A K-Means clustering technique is used for tumor isolation using color normalization, followed by multi-threshold Otsu segmentation technique to further classify tumor region as viable and non-viable. Then a Flood-fill algorithm is applied to cluster similar pixels into cellular objects and compute cluster data for further analysis of regions under study. To the best of our knowledge this is the first comprehensive solution that is able to produce such a classification for Osteosarcoma cancer. The results are very conclusive in identifying viable and non-viable tumor regions. In our experiments, the accuracy of the discussed approach is 100% in viable tumor and coagulative necrosis identification while it is around 90% for fibrosis and acellular/hypocellular tumor osteoid, for all the sampled datasets used. We expect the developed software to lead to a significant increase in accuracy and decrease in inter-observer variability in assessment of necrosis by the pathologists and a reduction in the time spent by the pathologists in such assessments.

AB - Osteosarcoma is one of the most common types of bone cancer in children. To gauge the extent of cancer treatment response in the patient after surgical resection, the H&E stained image slides are manually evaluated by pathologists to estimate the percentage of necrosis, a time consuming process prone to observer bias and inaccuracy. Digital image analysis is a potential method to automate this process, thus saving time and providing a more accurate evaluation. The slides are scanned in Aperio Scanscope, converted to digital Whole Slide Images (WSIs) and stored in SVS format. These are high resolution images, of the order of 109 pixels, allowing up to 40X magnification factor. This paper proposes an image segmentation and analysis technique for segmenting tumor and non-tumor regions in histopathological WSIs of osteosarcoma datasets. Our approach is a combination of pixel-based and object-based methods which utilize tumor properties such as nuclei cluster, density, and circularity to classify tumor regions as viable and non-viable. A K-Means clustering technique is used for tumor isolation using color normalization, followed by multi-threshold Otsu segmentation technique to further classify tumor region as viable and non-viable. Then a Flood-fill algorithm is applied to cluster similar pixels into cellular objects and compute cluster data for further analysis of regions under study. To the best of our knowledge this is the first comprehensive solution that is able to produce such a classification for Osteosarcoma cancer. The results are very conclusive in identifying viable and non-viable tumor regions. In our experiments, the accuracy of the discussed approach is 100% in viable tumor and coagulative necrosis identification while it is around 90% for fibrosis and acellular/hypocellular tumor osteoid, for all the sampled datasets used. We expect the developed software to lead to a significant increase in accuracy and decrease in inter-observer variability in assessment of necrosis by the pathologists and a reduction in the time spent by the pathologists in such assessments.

UR - http://www.scopus.com/inward/record.url?scp=85020745978&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020745978&partnerID=8YFLogxK

M3 - Article

VL - 22

SP - 195

EP - 206

JO - Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing

JF - Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing

SN - 2335-6936

ER -