The University of Southampton
University of Southampton Institutional Repository

Validation of a deep learning model for automatic segmentation of skeletal muscle and adipose tissue on L3 abdominal CT images

Validation of a deep learning model for automatic segmentation of skeletal muscle and adipose tissue on L3 abdominal CT images
Validation of a deep learning model for automatic segmentation of skeletal muscle and adipose tissue on L3 abdominal CT images
Background: body composition assessment using abdominal computed tomography (CT) images is increasingly applied in clinical and translational research. Manual segmentation of body compartments on L3 CT images is time-consuming and requires significant expertise. Robust high-throughput automated segmentation is key to assess large patient cohorts and ultimately, to support implementation into routine clinical practice. By training a deep learning neural network (DLNN) with several large trial cohorts and performing external validation on a large independent cohort, we aim to demonstrate the robust performance of our automatic body composition segmentation tool for future use in patients.

Methods: L3 CT images and expert-drawn segmentations of skeletal muscle, visceral adipose tissue, and subcutaneous adipose tissue of patients undergoing abdominal surgery were pooled (n = 3,187) to train a DLNN. The trained DLNN was then externally validated in a cohort with L3 CT images of patients with abdominal cancer (n = 2,535). Geometric agreement between automatic and manual segmentations was evaluated by computing two-dimensional Dice Similarity (DS). Agreement between manual and automatic annotations were quantitatively evaluated in the test set using Lin’s Concordance Correlation Coefficient (CCC) and Bland-Altman’s Limits of Agreement (LoA).

Results: the DLNN showed rapid improvement within the first 10,000 training steps and stopped improving after 38,000 steps. There was a strong concordance between automatic and manual segmentations with median DS for skeletal muscle, visceral adipose tissue, and subcutaneous adipose tissue of 0.97 (interquartile range, IQR: 0.95-0.98), 0.98 (IQR: 0.95-0.98), and 0.95 (IQR: 0.92-0.97), respectively. Concordance correlations were excellent: skeletal muscle 0.964 (0.959-0.968), visceral adipose tissue 0.998 (0.998-0.998), and subcutaneous adipose tissue 0.992 (0.991-0.993). Bland-Altman metrics (relative to approximate median values in parentheses) indicated only small and clinically insignificant systematic offsets : 0.23 HU (0.5%), 1.26 cm2.m-2 (2.8%), -1.02 cm2.m-2 (1.7%), and 3.24 cm2.m-2 (4.6%) for skeletal muscle average radiodensity, skeletal muscle index, visceral adipose tissue index, and subcutaneous adipose tissue index, respectively. Assuming the decision thresholds by Martin et al. for sarcopenia and low muscle radiation attenuation, results for sensitivity (0.99 and 0.98 respectively), specificity (0.87 and 0.98 respectively), and overall accuracy (0.93) were all excellent.

Conclusion: we developed and validated a deep learning model for automated analysis of body composition of patients with cancer. Due to the design of the DLNN, it can be easily implemented in various clinical infrastructures and used by other research groups to assess cancer patient cohorts or develop new models in other fields.
medRxiv
van Dijk, David P.J.
6e67c1e2-ca01-454d-bbee-fc0f7ff3f56f
Volmer, Leroy F.
725c649d-1c95-4597-93e3-57c78a283fe6
Brecheisen, Ralph
5005a1a8-c7fd-471f-87dd-dfe51cf86291
Dolan, Ross D.
f657b693-22df-40e4-b28a-7c5c104c91fa
Bryce, Adam S.
19270b25-d4b5-41e7-9d01-bf60ccd0e743
Chang, David K.
3787d5fc-a2c5-4e41-8a59-892f651ab108
McMillan, Donald C.
75ef9613-a319-4c91-a476-8081409684d2
Stoot, Jan H.M.B.
9c14b0f7-f667-411e-9553-5ef618b630ac
West, Malcolm A.
98b67e58-9875-4133-b236-8a10a0a12c04
Rensen, Sander S.
5d809235-a5ee-48d9-804e-89f711ee7e07
Dekker, Andre
e8837cf1-73c6-434b-a573-810622eb701a
Wee, Leonard
2c6395df-0366-4d7d-902d-b84ab4ec9acb
Olde Damink, Steven W.M.
504fcca4-2739-494f-a203-70f0d756d3f7
Tweed, T.T.T.
e3054e11-d59a-40ec-9a1a-3075ec1fc4c4
Tummers, S.
b67042b3-f007-406d-8a19-b20e4f119bcd
van der Kroft, G.
be91999f-0d58-4620-9411-85a0f8bde1e3
Ligthart, M.A.P.
11082169-9e97-4fd8-b3eb-3213292b2a37
Aberle, M.R.
7176966d-917a-44db-abf7-bbabe95cb60f
Tim, L.
fa339743-fdfa-46bb-a6f9-e38185bc8a05
Bongers, B.C.
b08a9261-102a-42b3-b0c2-153e0439b894
Ubachs, J.
e678867f-c0cd-43b9-912f-b0722d991939
Kruitwagen, R.F.P.M.
dcb34269-3014-42cd-a669-cbf7a781274f
Pugh, S.
83010563-0865-446c-ba71-95e2a45d9562
Primrose, J.N.
d85f3b28-24c6-475f-955b-ec457a3f9185
Bridgewater, J.A.
22a97d4c-b9df-4f88-b413-d3193d2d14b7
Pucher, P.H.
88d1340c-f1df-448d-a816-84fdefca48a1
Curtis, N.J.
c66321fd-90f1-48ca-9019-377712a9e318
Dreyer, S.B.
612da26a-171d-4d66-9652-44e5eeaedda7
Body Composition Collaborative
van Dijk, David P.J.
6e67c1e2-ca01-454d-bbee-fc0f7ff3f56f
Volmer, Leroy F.
725c649d-1c95-4597-93e3-57c78a283fe6
Brecheisen, Ralph
5005a1a8-c7fd-471f-87dd-dfe51cf86291
Dolan, Ross D.
f657b693-22df-40e4-b28a-7c5c104c91fa
Bryce, Adam S.
19270b25-d4b5-41e7-9d01-bf60ccd0e743
Chang, David K.
3787d5fc-a2c5-4e41-8a59-892f651ab108
McMillan, Donald C.
75ef9613-a319-4c91-a476-8081409684d2
Stoot, Jan H.M.B.
9c14b0f7-f667-411e-9553-5ef618b630ac
West, Malcolm A.
98b67e58-9875-4133-b236-8a10a0a12c04
Rensen, Sander S.
5d809235-a5ee-48d9-804e-89f711ee7e07
Dekker, Andre
e8837cf1-73c6-434b-a573-810622eb701a
Wee, Leonard
2c6395df-0366-4d7d-902d-b84ab4ec9acb
Olde Damink, Steven W.M.
504fcca4-2739-494f-a203-70f0d756d3f7
Tweed, T.T.T.
e3054e11-d59a-40ec-9a1a-3075ec1fc4c4
Tummers, S.
b67042b3-f007-406d-8a19-b20e4f119bcd
van der Kroft, G.
be91999f-0d58-4620-9411-85a0f8bde1e3
Ligthart, M.A.P.
11082169-9e97-4fd8-b3eb-3213292b2a37
Aberle, M.R.
7176966d-917a-44db-abf7-bbabe95cb60f
Tim, L.
fa339743-fdfa-46bb-a6f9-e38185bc8a05
Bongers, B.C.
b08a9261-102a-42b3-b0c2-153e0439b894
Ubachs, J.
e678867f-c0cd-43b9-912f-b0722d991939
Kruitwagen, R.F.P.M.
dcb34269-3014-42cd-a669-cbf7a781274f
Pugh, S.
83010563-0865-446c-ba71-95e2a45d9562
Primrose, J.N.
d85f3b28-24c6-475f-955b-ec457a3f9185
Bridgewater, J.A.
22a97d4c-b9df-4f88-b413-d3193d2d14b7
Pucher, P.H.
88d1340c-f1df-448d-a816-84fdefca48a1
Curtis, N.J.
c66321fd-90f1-48ca-9019-377712a9e318
Dreyer, S.B.
612da26a-171d-4d66-9652-44e5eeaedda7

[Unknown type: UNSPECIFIED]

Record type: UNSPECIFIED

Abstract

Background: body composition assessment using abdominal computed tomography (CT) images is increasingly applied in clinical and translational research. Manual segmentation of body compartments on L3 CT images is time-consuming and requires significant expertise. Robust high-throughput automated segmentation is key to assess large patient cohorts and ultimately, to support implementation into routine clinical practice. By training a deep learning neural network (DLNN) with several large trial cohorts and performing external validation on a large independent cohort, we aim to demonstrate the robust performance of our automatic body composition segmentation tool for future use in patients.

Methods: L3 CT images and expert-drawn segmentations of skeletal muscle, visceral adipose tissue, and subcutaneous adipose tissue of patients undergoing abdominal surgery were pooled (n = 3,187) to train a DLNN. The trained DLNN was then externally validated in a cohort with L3 CT images of patients with abdominal cancer (n = 2,535). Geometric agreement between automatic and manual segmentations was evaluated by computing two-dimensional Dice Similarity (DS). Agreement between manual and automatic annotations were quantitatively evaluated in the test set using Lin’s Concordance Correlation Coefficient (CCC) and Bland-Altman’s Limits of Agreement (LoA).

Results: the DLNN showed rapid improvement within the first 10,000 training steps and stopped improving after 38,000 steps. There was a strong concordance between automatic and manual segmentations with median DS for skeletal muscle, visceral adipose tissue, and subcutaneous adipose tissue of 0.97 (interquartile range, IQR: 0.95-0.98), 0.98 (IQR: 0.95-0.98), and 0.95 (IQR: 0.92-0.97), respectively. Concordance correlations were excellent: skeletal muscle 0.964 (0.959-0.968), visceral adipose tissue 0.998 (0.998-0.998), and subcutaneous adipose tissue 0.992 (0.991-0.993). Bland-Altman metrics (relative to approximate median values in parentheses) indicated only small and clinically insignificant systematic offsets : 0.23 HU (0.5%), 1.26 cm2.m-2 (2.8%), -1.02 cm2.m-2 (1.7%), and 3.24 cm2.m-2 (4.6%) for skeletal muscle average radiodensity, skeletal muscle index, visceral adipose tissue index, and subcutaneous adipose tissue index, respectively. Assuming the decision thresholds by Martin et al. for sarcopenia and low muscle radiation attenuation, results for sensitivity (0.99 and 0.98 respectively), specificity (0.87 and 0.98 respectively), and overall accuracy (0.93) were all excellent.

Conclusion: we developed and validated a deep learning model for automated analysis of body composition of patients with cancer. Due to the design of the DLNN, it can be easily implemented in various clinical infrastructures and used by other research groups to assess cancer patient cohorts or develop new models in other fields.

Text
2023.04.23.23288981v3.full - Author's Original
Download (1MB)

More information

Published date: 22 January 2023

Identifiers

Local EPrints ID: 488955
URI: http://eprints.soton.ac.uk/id/eprint/488955
PURE UUID: 08c3a6b0-66c7-4c4d-a3fd-7b6e767c6e62
ORCID for Malcolm A. West: ORCID iD orcid.org/0000-0002-0345-5356
ORCID for J.N. Primrose: ORCID iD orcid.org/0000-0002-2069-7605

Catalogue record

Date deposited: 09 Apr 2024 17:20
Last modified: 13 Apr 2024 01:52

Export record

Altmetrics

Contributors

Author: David P.J. van Dijk
Author: Leroy F. Volmer
Author: Ralph Brecheisen
Author: Ross D. Dolan
Author: Adam S. Bryce
Author: David K. Chang
Author: Donald C. McMillan
Author: Jan H.M.B. Stoot
Author: Malcolm A. West ORCID iD
Author: Sander S. Rensen
Author: Andre Dekker
Author: Leonard Wee
Author: Steven W.M. Olde Damink
Author: T.T.T. Tweed
Author: S. Tummers
Author: G. van der Kroft
Author: M.A.P. Ligthart
Author: M.R. Aberle
Author: L. Tim
Author: B.C. Bongers
Author: J. Ubachs
Author: R.F.P.M. Kruitwagen
Author: S. Pugh
Author: J.N. Primrose ORCID iD
Author: J.A. Bridgewater
Author: P.H. Pucher
Author: N.J. Curtis
Author: S.B. Dreyer
Corporate Author: Body Composition Collaborative

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×