Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

Insights into the behaviour of multi-task deep neural networks for medical image segmentation

Abstract: 

Glandular morphology is used by pathologists to assess the malignancy of different adenocarcinomas. This process involves conducting gland segmentation task. The common approach in specialised domains, such as medical imaging, is to design complex architectures in a multi-task learning setup. Generally, these approaches rely on substantial postprocessing efforts. Moreover, a predominant notion is that general purpose models are not suitable for gland instance segmentation. We analyse the behaviour of two architectures: SA-FCN and Mask R-CNN. We compare the impact of post-processing on the final predictive results and the performance of generic and specific models for the gland segmentation problem. Our results highlight the dependency of post-processing on tailored models as well as comparable results when using a generic model. Thus, in the interest of time, it is worth considering to use and improve generic models as opposed to design complex architectures when tackling new domains.

up
1 user has voted: Lukasz Bienias

Paper Details

Authors:
Juanjo R. Guillamon, Line H. Nielsen
Submitted On:
1 November 2019 - 11:34am
Short Link:
Type:
Presentation Slides
Event:
Presenter's Name:
Lukasz Tomasz Bienias
Document Year:
2019
Cite

Document Files

Insigths_into_the_behaviour_of_multi_task_deep_neural_networks_for_medical_image_segmentation.pdf

(81)

Subscribe

[1] Juanjo R. Guillamon, Line H. Nielsen, "Insights into the behaviour of multi-task deep neural networks for medical image segmentation", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4902. Accessed: Sep. 20, 2020.
@article{4902-19,
url = {http://sigport.org/4902},
author = {Juanjo R. Guillamon; Line H. Nielsen },
publisher = {IEEE SigPort},
title = {Insights into the behaviour of multi-task deep neural networks for medical image segmentation},
year = {2019} }
TY - EJOUR
T1 - Insights into the behaviour of multi-task deep neural networks for medical image segmentation
AU - Juanjo R. Guillamon; Line H. Nielsen
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4902
ER -
Juanjo R. Guillamon, Line H. Nielsen. (2019). Insights into the behaviour of multi-task deep neural networks for medical image segmentation. IEEE SigPort. http://sigport.org/4902
Juanjo R. Guillamon, Line H. Nielsen, 2019. Insights into the behaviour of multi-task deep neural networks for medical image segmentation. Available at: http://sigport.org/4902.
Juanjo R. Guillamon, Line H. Nielsen. (2019). "Insights into the behaviour of multi-task deep neural networks for medical image segmentation." Web.
1. Juanjo R. Guillamon, Line H. Nielsen. Insights into the behaviour of multi-task deep neural networks for medical image segmentation [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4902