Towards Grading Gleason Score using Generically Trained Deep convolutional Neural Networks
Källén, Hanna LU ; Molin, Jesper ; Heyden, Anders LU
; Lundström, Claes
and Åström, Karl
LU
(2016)
2016 IEEE International Symposium on Biomedical Imaging
p.1163-1167
- Abstract
- We developed an automatic algorithm with the purpose to assist pathologists to report Gleason score on malignant prostatic adenocarcinoma specimen. In order to detect and classify the cancerous tissue, a deep convolutional neural network that had been pre-trained on a large set of photographic images was used. A specific aim was to support intuitive interaction with the result, to let pathologists adjust and correct the output. Therefore, we have designed an algorithm that makes a spatial classification of the whole slide into the same growth patterns as pathologists do. The 22-layer network was cut at an earlier layer and the output from that layer was used to train both a random forest classifier and a support vector machines... (More)
- We developed an automatic algorithm with the purpose to assist pathologists to report Gleason score on malignant prostatic adenocarcinoma specimen. In order to detect and classify the cancerous tissue, a deep convolutional neural network that had been pre-trained on a large set of photographic images was used. A specific aim was to support intuitive interaction with the result, to let pathologists adjust and correct the output. Therefore, we have designed an algorithm that makes a spatial classification of the whole slide into the same growth patterns as pathologists do. The 22-layer network was cut at an earlier layer and the output from that layer was used to train both a random forest classifier and a support vector machines classifier. At a specific layer a small patch of the image was used to calculate a feature vector and an image is represented by a number of those vectors. We have classified both the individual patches and the entire images. The classification results were compared for different scales of the images and feature vectors from two different layers from the network. Testing was made on a dataset consisting of 213 images, all containing a single class, benign tissue or Gleason score 3-5. Using 10-fold cross validation the accuracy per patch was 81 %. For whole images, the accuracy was increased to 89 %. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/839c04b4-164c-4145-8620-9a6f2e5139c4
- author
- Källén, Hanna
LU
; Molin, Jesper
; Heyden, Anders
LU
; Lundström, Claes
and Åström, Karl
LU
- organization
- publishing date
- 2016
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- keywords
- Prostate Cancer, Gleason Score, Deep Learning, Convolutional Neural Networks
- host publication
- 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI)
- pages
- 5 pages
- publisher
- IEEE - Institute of Electrical and Electronics Engineers Inc.
- conference name
- 2016 IEEE International Symposium on Biomedical Imaging
- conference location
- Prague, Czech Republic
- conference dates
- 2016-04-13 - 2016-04-16
- external identifiers
-
- scopus:84978387816
- wos:000386377400275
- ISBN
- 978-1-4799-2349-6
- DOI
- 10.1109/ISBI.2016.7493473
- language
- English
- LU publication?
- yes
- id
- 839c04b4-164c-4145-8620-9a6f2e5139c4
- date added to LUP
- 2016-06-01 11:51:33
- date last changed
- 2023-12-20 18:34:44
@inproceedings{839c04b4-164c-4145-8620-9a6f2e5139c4,
abstract = {{We developed an automatic algorithm with the purpose to assist pathologists to report Gleason score on malignant prostatic adenocarcinoma specimen. In order to detect and classify the cancerous tissue, a deep convolutional neural network that had been pre-trained on a large set of photographic images was used. A specific aim was to support intuitive interaction with the result, to let pathologists adjust and correct the output. Therefore, we have designed an algorithm that makes a spatial classification of the whole slide into the same growth patterns as pathologists do. The 22-layer network was cut at an earlier layer and the output from that layer was used to train both a random forest classifier and a support vector machines classifier. At a specific layer a small patch of the image was used to calculate a feature vector and an image is represented by a number of those vectors. We have classified both the individual patches and the entire images. The classification results were compared for different scales of the images and feature vectors from two different layers from the network. Testing was made on a dataset consisting of 213 images, all containing a single class, benign tissue or Gleason score 3-5. Using 10-fold cross validation the accuracy per patch was 81 %. For whole images, the accuracy was increased to 89 %.}},
author = {{Källén, Hanna and Molin, Jesper and Heyden, Anders and Lundström, Claes and Åström, Karl}},
booktitle = {{2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI)}},
isbn = {{978-1-4799-2349-6}},
keywords = {{Prostate Cancer; Gleason Score; Deep Learning; Convolutional Neural Networks}},
language = {{eng}},
pages = {{1163--1167}},
publisher = {{IEEE - Institute of Electrical and Electronics Engineers Inc.}},
title = {{Towards Grading Gleason Score using Generically Trained Deep convolutional Neural Networks}},
url = {{https://lup.lub.lu.se/search/files/8106067/ISBI2016_prostata.pdf}},
doi = {{10.1109/ISBI.2016.7493473}},
year = {{2016}},
}