Gland segmentation in colon histology images using hand-crafted features and convolutional neural networks

Wenqi Li, Siyamalan Manivannan, Shazia Akbar, Jianguo Zhang, Emanuele Trucco, Stephen J. McKenna

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

  • 15 Citations

Abstract

We investigate glandular structure segmentation in colon histology images as a window-based classification problem. We compare and combine methods based on fine-tuned convolutional neural networks (CNN) and hand-crafted features with support vector machines (HC-SVM). On 85 images of H&E-stained tissue, we find that fine-tuned CNN outperforms HC-SVM in gland segmentation measured by pixel-wise Jaccard and Dice indices. For HC-SVM we further observe that training a second-level window classifier on the posterior probabilities – as an output refinement – can substantially improve the segmentation performance. The final performance of HC-SVM with refinement is comparable to that of CNN. Furthermore, we show that by combining and refining the posterior probability out-puts of CNN and HC-SVM together, a further performance boost is obtained.
LanguageEnglish
Title of host publication2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI)
Subtitle of host publicationFrom Nano to Macro, ISBI 2016 - Proceedings
PublisherIEEE
Pages1405-1408
Number of pages4
VolumeJune-2016
ISBN (Electronic)9781479923496
ISBN (Print)9781479923502
DOIs
Publication statusPublished - 16 Jun 2016
Event2016 IEEE 13th International Symposium on Biomedical Imaging - Clarion Congress Hotel, Prague, Czech Republic
Duration: 13 Apr 201616 Apr 2016
http://biomedicalimaging.org/2016/ (Link to Conference website)

Conference

Conference2016 IEEE 13th International Symposium on Biomedical Imaging
Abbreviated titleISBI 2016
CountryCzech Republic
CityPrague
Period13/04/1616/04/16
Internet address

Fingerprint

Histology
Support vector machines
Neural networks
Refining
Classifiers
Pixels
Tissue

Keywords

  • Glands
  • Image segmentation
  • Support vector machines
  • Feature extraction
  • Dictionaries
  • Training
  • Neural networks

Cite this

Li, W., Manivannan, S., Akbar, S., Zhang, J., Trucco, E., & McKenna, S. J. (2016). Gland segmentation in colon histology images using hand-crafted features and convolutional neural networks. In 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI): From Nano to Macro, ISBI 2016 - Proceedings (Vol. June-2016, pp. 1405-1408). IEEE. https://doi.org/10.1109/ISBI.2016.7493530
Li, Wenqi ; Manivannan, Siyamalan ; Akbar, Shazia ; Zhang, Jianguo ; Trucco, Emanuele ; McKenna, Stephen J. / Gland segmentation in colon histology images using hand-crafted features and convolutional neural networks. 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI): From Nano to Macro, ISBI 2016 - Proceedings. Vol. June-2016 IEEE, 2016. pp. 1405-1408
@inproceedings{e940cd620d854ad5859c82fd944956aa,
title = "Gland segmentation in colon histology images using hand-crafted features and convolutional neural networks",
abstract = "We investigate glandular structure segmentation in colon histology images as a window-based classification problem. We compare and combine methods based on fine-tuned convolutional neural networks (CNN) and hand-crafted features with support vector machines (HC-SVM). On 85 images of H&E-stained tissue, we find that fine-tuned CNN outperforms HC-SVM in gland segmentation measured by pixel-wise Jaccard and Dice indices. For HC-SVM we further observe that training a second-level window classifier on the posterior probabilities – as an output refinement – can substantially improve the segmentation performance. The final performance of HC-SVM with refinement is comparable to that of CNN. Furthermore, we show that by combining and refining the posterior probability out-puts of CNN and HC-SVM together, a further performance boost is obtained.",
keywords = "Glands, Image segmentation, Support vector machines, Feature extraction, Dictionaries, Training, Neural networks",
author = "Wenqi Li and Siyamalan Manivannan and Shazia Akbar and Jianguo Zhang and Emanuele Trucco and McKenna, {Stephen J.}",
note = "S. Manivannan is supported by EPSRC grant EP/M005976/1",
year = "2016",
month = "6",
day = "16",
doi = "10.1109/ISBI.2016.7493530",
language = "English",
isbn = "9781479923502",
volume = "June-2016",
pages = "1405--1408",
booktitle = "2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI)",
publisher = "IEEE",

}

Li, W, Manivannan, S, Akbar, S, Zhang, J, Trucco, E & McKenna, SJ 2016, Gland segmentation in colon histology images using hand-crafted features and convolutional neural networks. in 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI): From Nano to Macro, ISBI 2016 - Proceedings. vol. June-2016, IEEE, pp. 1405-1408, 2016 IEEE 13th International Symposium on Biomedical Imaging, Prague, Czech Republic, 13/04/16. https://doi.org/10.1109/ISBI.2016.7493530

Gland segmentation in colon histology images using hand-crafted features and convolutional neural networks. / Li, Wenqi; Manivannan, Siyamalan; Akbar, Shazia; Zhang, Jianguo; Trucco, Emanuele; McKenna, Stephen J.

2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI): From Nano to Macro, ISBI 2016 - Proceedings. Vol. June-2016 IEEE, 2016. p. 1405-1408.

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

TY - GEN

T1 - Gland segmentation in colon histology images using hand-crafted features and convolutional neural networks

AU - Li, Wenqi

AU - Manivannan, Siyamalan

AU - Akbar, Shazia

AU - Zhang, Jianguo

AU - Trucco, Emanuele

AU - McKenna, Stephen J.

N1 - S. Manivannan is supported by EPSRC grant EP/M005976/1

PY - 2016/6/16

Y1 - 2016/6/16

N2 - We investigate glandular structure segmentation in colon histology images as a window-based classification problem. We compare and combine methods based on fine-tuned convolutional neural networks (CNN) and hand-crafted features with support vector machines (HC-SVM). On 85 images of H&E-stained tissue, we find that fine-tuned CNN outperforms HC-SVM in gland segmentation measured by pixel-wise Jaccard and Dice indices. For HC-SVM we further observe that training a second-level window classifier on the posterior probabilities – as an output refinement – can substantially improve the segmentation performance. The final performance of HC-SVM with refinement is comparable to that of CNN. Furthermore, we show that by combining and refining the posterior probability out-puts of CNN and HC-SVM together, a further performance boost is obtained.

AB - We investigate glandular structure segmentation in colon histology images as a window-based classification problem. We compare and combine methods based on fine-tuned convolutional neural networks (CNN) and hand-crafted features with support vector machines (HC-SVM). On 85 images of H&E-stained tissue, we find that fine-tuned CNN outperforms HC-SVM in gland segmentation measured by pixel-wise Jaccard and Dice indices. For HC-SVM we further observe that training a second-level window classifier on the posterior probabilities – as an output refinement – can substantially improve the segmentation performance. The final performance of HC-SVM with refinement is comparable to that of CNN. Furthermore, we show that by combining and refining the posterior probability out-puts of CNN and HC-SVM together, a further performance boost is obtained.

KW - Glands

KW - Image segmentation

KW - Support vector machines

KW - Feature extraction

KW - Dictionaries

KW - Training

KW - Neural networks

UR - http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=7493530&searchWithin%3Dgland+segmentation+in+colon%26filter%3DAND%28p_IS_Number%3A7493185%29

U2 - 10.1109/ISBI.2016.7493530

DO - 10.1109/ISBI.2016.7493530

M3 - Conference contribution

SN - 9781479923502

VL - June-2016

SP - 1405

EP - 1408

BT - 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI)

PB - IEEE

ER -

Li W, Manivannan S, Akbar S, Zhang J, Trucco E, McKenna SJ. Gland segmentation in colon histology images using hand-crafted features and convolutional neural networks. In 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI): From Nano to Macro, ISBI 2016 - Proceedings. Vol. June-2016. IEEE. 2016. p. 1405-1408 https://doi.org/10.1109/ISBI.2016.7493530