A Lightweight Deep Learning Approach for Liver Segmentation
Liver segmentation is a prerequisite for various hepatic interventions and is a time-consuming manual task performed by radiology experts. Recently, various computationally expensive deep learning architectures tackled this aspect without considering the resource limitations of a real-life clinical...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-12-01
|
Series: | Mathematics |
Subjects: | |
Online Access: | https://www.mdpi.com/2227-7390/11/1/95 |
_version_ | 1797440506042515456 |
---|---|
author | Smaranda Bogoi Andreea Udrea |
author_facet | Smaranda Bogoi Andreea Udrea |
author_sort | Smaranda Bogoi |
collection | DOAJ |
description | Liver segmentation is a prerequisite for various hepatic interventions and is a time-consuming manual task performed by radiology experts. Recently, various computationally expensive deep learning architectures tackled this aspect without considering the resource limitations of a real-life clinical setup. In this paper, we investigated the capabilities of a lightweight model, UNeXt, in comparison with the U-Net model. Moreover, we conduct a broad analysis at the micro and macro levels of these architectures by using two training loss functions: soft dice loss and unified focal loss, and by substituting the commonly used ReLU activation function, with the novel Funnel activation function. An automatic post-processing step that increases the overall performance of the models is also proposed. Model training and evaluation were performed on a public database—LiTS. The results show that the UNeXt model (Funnel activation, soft dice loss, post-processing step) achieved a 0.9902 dice similarity coefficient on the whole CT volumes in the test set, with 15× fewer parameters in nearly 4× less inference time, compared to its counterpart, U-Net. Thus, lightweight models can become the new standard in medical segmentation, and when implemented thoroughly can alleviate the computational burden while preserving the capabilities of a parameter-heavy architecture. |
first_indexed | 2024-03-09T12:09:09Z |
format | Article |
id | doaj.art-e3ae3f6b83f84f29bdc4ea5d90ab54f9 |
institution | Directory Open Access Journal |
issn | 2227-7390 |
language | English |
last_indexed | 2024-03-09T12:09:09Z |
publishDate | 2022-12-01 |
publisher | MDPI AG |
record_format | Article |
series | Mathematics |
spelling | doaj.art-e3ae3f6b83f84f29bdc4ea5d90ab54f92023-11-30T22:54:58ZengMDPI AGMathematics2227-73902022-12-011119510.3390/math11010095A Lightweight Deep Learning Approach for Liver SegmentationSmaranda Bogoi0Andreea Udrea1Department of Automatic Control and Systems Engineering, Faculty of Computer Science and Automatic Control, University “Politehnica” of Bucharest, 060042 Bucharest, RomaniaDepartment of Automatic Control and Systems Engineering, Faculty of Computer Science and Automatic Control, University “Politehnica” of Bucharest, 060042 Bucharest, RomaniaLiver segmentation is a prerequisite for various hepatic interventions and is a time-consuming manual task performed by radiology experts. Recently, various computationally expensive deep learning architectures tackled this aspect without considering the resource limitations of a real-life clinical setup. In this paper, we investigated the capabilities of a lightweight model, UNeXt, in comparison with the U-Net model. Moreover, we conduct a broad analysis at the micro and macro levels of these architectures by using two training loss functions: soft dice loss and unified focal loss, and by substituting the commonly used ReLU activation function, with the novel Funnel activation function. An automatic post-processing step that increases the overall performance of the models is also proposed. Model training and evaluation were performed on a public database—LiTS. The results show that the UNeXt model (Funnel activation, soft dice loss, post-processing step) achieved a 0.9902 dice similarity coefficient on the whole CT volumes in the test set, with 15× fewer parameters in nearly 4× less inference time, compared to its counterpart, U-Net. Thus, lightweight models can become the new standard in medical segmentation, and when implemented thoroughly can alleviate the computational burden while preserving the capabilities of a parameter-heavy architecture.https://www.mdpi.com/2227-7390/11/1/95CT liver segmentationlightweight neural networkLiTSautomatic post-processing |
spellingShingle | Smaranda Bogoi Andreea Udrea A Lightweight Deep Learning Approach for Liver Segmentation Mathematics CT liver segmentation lightweight neural network LiTS automatic post-processing |
title | A Lightweight Deep Learning Approach for Liver Segmentation |
title_full | A Lightweight Deep Learning Approach for Liver Segmentation |
title_fullStr | A Lightweight Deep Learning Approach for Liver Segmentation |
title_full_unstemmed | A Lightweight Deep Learning Approach for Liver Segmentation |
title_short | A Lightweight Deep Learning Approach for Liver Segmentation |
title_sort | lightweight deep learning approach for liver segmentation |
topic | CT liver segmentation lightweight neural network LiTS automatic post-processing |
url | https://www.mdpi.com/2227-7390/11/1/95 |
work_keys_str_mv | AT smarandabogoi alightweightdeeplearningapproachforliversegmentation AT andreeaudrea alightweightdeeplearningapproachforliversegmentation AT smarandabogoi lightweightdeeplearningapproachforliversegmentation AT andreeaudrea lightweightdeeplearningapproachforliversegmentation |