Skip to main content
Key Specialties
Cardiology Diabetology Endocrinology Gastroenterology Geriatrics and Gerontology Gynecology and Obstetrics Hematology Infectious Disease Internal Medicine Nephrology Neurology Oncology Pediatrics Respiratory Medicine Rheumatology Surgery
Congress Coverage
2024 ACC Congress 2023 ESMO Congress 2023 EASD Congress 2023 ERS Congress 2023 ESC Congress 2023 EHA Congress 2023 ASCO Annual Meeting
Clinical Collections
Advances in Alzheimer’s

At a glance: The ONWARDS insulin icodec trials

A round-up of the ONWARDS phase 3 clinical trials evaluating the novel weekly insulin icodec in people with diabetes.
ADVANCED SEARCH
Log in
Top
Magnetic Resonance Materials in Physics, Biology and Medicine
Published in: Magnetic Resonance Materials in Physics, Biology and Medicine 4/2020

01-08-2020 | Magnetic Resonance Imaging | Research Article

Clinical evaluation of fully automated thigh muscle and adipose tissue segmentation using a U-Net deep learning architecture in context of osteoarthritic knee pain

Authors: Jana Kemnitz, Christian F. Baumgartner, Felix Eckstein, Akshay Chaudhari, Anja Ruhdorfer, Wolfgang Wirth, Sebastian K. Eder, Ender Konukoglu

Published in: Magnetic Resonance Materials in Physics, Biology and Medicine | Issue 4/2020

Login to get access

Abstract

Objective

Segmentation of thigh muscle and adipose tissue is important for the understanding of musculoskeletal diseases such as osteoarthritis. Therefore, the purpose of this work is (a) to evaluate whether a fully automated approach provides accurate segmentation of muscles and adipose tissue cross-sectional areas (CSA) compared with manual segmentation and (b) to evaluate the validity of this method based on a previous clinical study.

Materials and methods

The segmentation method is based on U-Net architecture trained on 250 manually segmented thighs from the Osteoarthritis Initiative (OAI). The clinical evaluation is performed on a hold-out test set bilateral thighs of 48 subjects with unilateral knee pain.

Results

The segmentation time of the method is < 1 s and demonstrated high agreement with the manual method (dice similarity coeffcient: 0.96 ± 0.01). In the clinical study, the automated method shows that similar to manual segmentation (− 5.7 ± 7.9%, p < 0.001, effect size: 0.69), painful knees display significantly lower quadriceps CSAs than contralateral painless knees (− 5.6 ± 7.6%, p < 0.001, effect size: 0.73).

Discussion

Automated segmentation of thigh muscle and adipose tissues has high agreement with manual segmentations and can replicate the effect size seen in a clinical study on osteoarthritic pain.
Literature
1.
Issa RI, Griffin TM (2012) Pathobiology of obesity and osteoarthritis: integrating biomechanics and inflammation. Pathobiol Aging Age Relat Dis 2:1–7
2.
Øiestad BE, Juhl CB, Eitzen I, Thorlund JB (2015) Knee extensor muscle weakness is a risk factor for development of knee osteoarthritis. A systematic review and meta-analysis. Osteoarthritis Cartilage 23:171–177CrossRef
3.
Mobasheri A, Rayman MP, Gualillo O, Sellam J, Van Der Kraan P, Fearon U (2017) The role of metabolism in the pathogenesis of osteoarthritis. Nat Rev Rheumatol 13:302–311CrossRef
4.
Chang J, Liao Z, Lu M, Meng T, Han W, Ding C (2018) Systemic and local adipose tissue in knee osteoarthritis. Osteoarthr Cartil 26:864–871CrossRef
5.
Maly MR, Calder KM, Macintyre NJ, Beattie KA (2013) Relationship of intermuscular fat volume in the thigh with knee extensor strength and physical performance in women at risk of or with knee osteoarthritis. Arthritis Care Res (Hoboken) 65:44–52CrossRef
6.
Winby CR, Lloyd DG, Besier TF, Kirk TB (2009) Muscle and external load contribution to knee joint contact loads during normal gait. J Biomech 42:2294–2300CrossRef
7.
Ruhdorfer A, Dannhauer T, Wirth W, Hitzl W, Kwoh CK, Guermazi A, Hunter DJ, Benichou O, Eckstein F (2013) Thigh muscle cross-sectional areas and strength in advanced versus early painful osteoarthritis: an exploratory between-knee, within-person comparison in osteoarthritis initiative participants. Arthritis Care Res (Hoboken) 65:1034–1042CrossRef
8.
Ruhdorfer A, Wirth W, Dannhauer T, Eckstein F (2015) Longitudinal (4 year) change of thigh muscle and adipose tissue distribution in chronically painful vs painless knees—data from the Osteoarthritis Initiative. Osteoarthr Cartil 23:1348–1356CrossRef
9.
Dannhauer T, Sattler M, Wirth W, Hunter DJ, Kwoh CK, Eckstein F (2012) Comparison of muscle area and strength between oa knees with and without structural progression—data from the OA initiative. Osteoarthr Cartil 20:S221–S222CrossRef
10.
Kemnitz J, Wirth W, Eckstein F, Culvenor AG (2018) The role of thigh muscle and adipose tissue in knee osteoarthritis progression in women: data from the osteoarthritis initiative. Osteoarthr Cartil 26:1190–1195CrossRef
11.
Kumar D, Subburaj K, Lin W, Karampinos DC, McCulloch CE, Li X, Link TM, Souza RB, Majumdar S (2013) Quadriceps and hamstrings morphology is related to walking mechanics and knee cartilage MRI relaxation times in young adults. J Orthop Sports Phys Ther 43:881–890CrossRef
12.
Culvenor AG, Hamler FC, Kemnitz J, Wirth W, Eckstein F (2018) Brief report: loss of muscle strength prior to knee replacement: a question of anatomic cross-sectional area or specific strength? Arthritis Rheumatol 70:222–229CrossRef
13.
Dannhauer T, Ruhdorfer A, Wirth W, Eckstein F (2015) Quantitative relationship of thigh adipose tissue with pain, radiographic status, and progression of knee osteoarthritis: longitudinal findings from the osteoarthritis initiative. Invest Radiol 50:268–274CrossRef
14.
Visser AW, Ioan-Facsinay A, de Mutsert R, Widya RL, Loef M, de Roos A, le Cessie S, den Heijer M, Rosendaal FR, Kloppenburg M (2014) Adiposity and hand osteoarthritis: the Netherlands epidemiology of obesity study. Arthritis Res Ther 16:R19CrossRef
15.
Peterfy CG, Schneider E, Nevitt M (2008) The osteoarthritis initiative: report on the design rationale for the magnetic resonance imaging protocol for the knee. Osteoarthr Cartil 16:1433–1441CrossRef
16.
Gilles B, Magnenat-Thalmann N (2010) Musculoskeletal MRI segmentation using multi-resolution simplex meshes with medial representations. Med Image Anal 14:291–302CrossRef
17.
Prescott JW, Best TM, Swanson MS, Haq F, Jackson RD, Gurcan MN (2011) Anatomically anchored template-based level set segmentation: application to quadriceps muscles in MR images from the osteoarthritis initiative. J Digit imaging Off J Soc Comput Appl Radiol 24:28–43
18.
Le Troter A, Fouré A, Guye M, Confort-Gouny S, Mattei JP, Gondin J, Salort-Campana E, Bendahan D (2016) Volume measurements of individual muscles in human quadriceps femoris using atlas-based segmentation approaches. Magn Reson Mater Phy 29:245–257CrossRef
19.
Kemnitz J, Eckstein F, Culvenor AG, Ruhdorfer A, Dannhauer T, Ring-Dimitriou S, Sänger AM, Wirth W (2017) Validation of an active shape model-based semi-automated segmentation algorithm for the analysis of thigh muscle and adipose tissue cross-sectional areas. Magn Reson Mater Phy 30:489–503CrossRef
20.
Ghatas MP, Lester RM, Khan MR, Gorgey AS (2018) Semi-automated segmentation of magnetic resonance images for thigh skeletal muscle and fat using threshold technique after spinal cord injury. Neural Regen Res 13:1787–1795CrossRef
21.
Kemnitz J, Eckstein F, Culvenor AG, Ruhdorfer A, Dannhauer T, Ring-Dimitriou S, Sänger AM, Wirth W (2018) Validation of a 3D thigh muscle and adipose tissue segmentation method using statistical shape models. Osteoarhritis Cartil 26:457–458CrossRef
22.
23.
Andrews S, Hamarneh G (2015) The generalized log-ratio transformation: learning shape and adjacency priors for simultaneous thigh muscle segmentation. IEEE Trans Med Imaging 0062:1–1
24.
Karlsson A, Rosander J, Romu T, Tallberg J, Grönqvist A, Borga M, Dahlqvist Leinhard O (2015) Automatic and quantitative assessment of regional muscle volume by multi-atlas segmentation using whole-body water-fat MRI. J Magn Reson Imaging 41:1558–1569CrossRef
25.
Baudin P-Y, Azzabou N, Carlier PG, Paragios N (2012) Prior knowledge, random walks and human skeletal muscle segmentation. Med Image Comput Interv 7510:569–576
26.
27.
Lareau-Trudel E, Le TA, Ghattas B, Pouget J, Attarian S, Bendahan D, Salort-Campana E (2015) Muscle quantitative MR imaging and clustering analysis in patients with facioscapulohumeral muscular dystrophy type 1. PLoS ONE 10:1–16CrossRef
28.
Orgiu S, Lafortuna CL, Rastelli F, Cadioli M, Falini A, Rizzo G (2015) Automatic muscle and fat segmentation in the thigh from T1-Weighted MRI. J Magn Reson Imaging 43:601–610CrossRef
29.
Feng C, Zhao D, Huang M (2016) Image segmentation using CUDA accelerated non-local means denoising and bias correction embedded fuzzy c-means (BCEFCM). Signal Processing 122:164–189CrossRef
30.
Feng C, Zhao D, Huang M (2017) Image segmentation and bias correction using local inhomogeneous iNtensity clustering (LINC): a region-based level set method. Neurocomputing 219:107–129CrossRef
31.
32.
Ronneberger O, Fischer P, Brox T (2015) U-Net: convolutional networks for biomedical image segmentation. Lect Notes Comput Sci 9351:234–241CrossRef
33.
Litjens G, Kooi T, Bejnordi BE, Setio AAA, Ciompi F, Ghafoorian M, van der Laak JAWM, van Ginneken B, Sánchez CI (2017) A survey on deep learning in medical image analysis. Med Image Anal 42:60–88CrossRef
34.
Liu F, Zhou Z, Jang H, Samsonov A, Zhao G (2018) Deep convolutional neural network and 3D deformable approach for tissue segmentation in musculoskeletal magnetic resonance imaging. Magn Reson Med 79:2379–2391CrossRef
35.
Norman B, Pedoia BSV, Majumdar S (2018) Use of 2D U-Net convolutional neural networks for automated cartilage and meniscus segmentation of knee MR imaging data to determine relaxometry and morphometry. Radiology 288:177–185CrossRef
36.
37.
Ahmad E, McPhee JS, Degens H YM (2018) Automatic segmentation of MRI human thigh muscles: combination of reliable and fast framework methods for quadriceps, femur and marrow segmentation. 31–38
38.
Kemnitz J, Baumgartner CF, Wirth W, Eckstein F, Eder SK, Konukoglu E (2018) Combining heterogeneously labeled datasets for training segmentation networks. In: Shi Y, Suk HI, Liu M (eds) Machine learning in medical imaging. MLMI 2018. Lecture notes in computer science, vol 11046. Springer, Cham. https://​doi.​org/​10.​1007/​978-3-030-00919-9_​32
39.
Baumgartner CF, Koch LM, Pollefeys M, Konukoglu E (2018) An Exploration of 2D and 3D deep learning techniques for cardiac MR image segmentation. In: Pop M et al (eds) Statistical atlases and computational models of the heart. ACDC and MMWHS challenges. STACOM 2017. Lecture notes in computer science, vol 10663. Springer, Cham. https://​doi.​org/​10.​1007/​978-3-319-75541-0_​12
40.
Sattler M, Dannhauer T, Hudelmaier M, Wirth W, Sänger AM, Kwoh CK, Hunter DJ, Eckstein F, Sanger AM (2012) Side differences of thigh muscle cross-sectional areas and maximal isometric muscle force in bilateral knees with the same radiographic disease stage, but unilateral frequent pain—data from the osteoarthritis initiative. Osteoarthr Cart 20:532–540CrossRef
41.
Kingma DP, Ba JL (2015) ADAM: a method for stochastic optimization. Conf Pap ICLR 2015:1–15
42.
Ruhdorfer A, Dannhauer T, Wirth W, Cotofana S, Roemer FW, Nevitt M, Eckstein F (2014) Thigh muscle cross-sectional areas and strength in knees with early vs knees without radiographic knee osteoarthritis: a between-knee, within-person comparison. Osteoarthr Cartil 22:1634–1638CrossRef
43.
Eckstein F, Boeth H, Diederichs G, Wirth W, Hudelmaier M, Cotofana S, Hofmann-Amtenbrink M, Duda G (2014) Longitudinal change in femorotibial cartilage thickness and subchondral bone plate area in male and female adolescent vs. mature athletes. Ann Anat 196:150–157CrossRef
44.
Dannhauer T, Sattler M, Wirth W, Hunter DJ, Kwoh CK, Eckstein F (2014) Longitudinal sensitivity to change of MRI-based muscle cross-sectional area versus isometric strength analysis in osteoarthritic knees with and without structural progression: pilot data from the Osteoarthritis Initiative. Magn Reson Mater Phy 27:339–347CrossRef
45.
Ruhdorfer AS, Steidle E, Diepold J, Pogacnik Murillo AL, Dannhauer T, Wirth W, Eckstein F (2016) Inter- and intra-observer reliability of thigh muscle and adipose tissue cross-sectional area analysis from MR images. Osteoarthr Cartil 24:S415CrossRef
46.
Cotofana S, Hudelmaier M, Wirth W, Himmer M, Ring-Dimitriou S, Sanger AM, Eckstein F, Sänger AM (2010) Correlation between single-slice muscle anatomical cross-sectional area and muscle volume in thigh extensors, flexors and adductors of perimenopausal women. Eur J Appl Physiol 110:91–97CrossRef
47.
Ruhdorfer A (2013) Chronically painful vs. painless knees: Longitudinal (4 year) change in thigh muscle/adipose tissue distribution and isometric muscle strength. 1–17
48.
Ruhdorfer A, Wirth W, Eckstein F (2016) Longitudinal change in thigh muscle strength prior and concurrent to a minimal clinically important worsening or improvement in knee function—data from the Osteoarthritis Initiative. Arthritis Rheumatol 68:826–836CrossRef
Metadata
Title
Clinical evaluation of fully automated thigh muscle and adipose tissue segmentation using a U-Net deep learning architecture in context of osteoarthritic knee pain
Authors
Jana Kemnitz
Christian F. Baumgartner
Felix Eckstein
Akshay Chaudhari
Anja Ruhdorfer
Wolfgang Wirth
Sebastian K. Eder
Ender Konukoglu
Publication date
01-08-2020
Publisher
Springer International Publishing
Published in
Magnetic Resonance Materials in Physics, Biology and Medicine / Issue 4/2020
Print ISSN: 0968-5243
Electronic ISSN: 1352-8661
DOI
https://doi.org/10.1007/s10334-019-00816-5

Other articles of this Issue 4/2020

Magnetic Resonance Materials in Physics, Biology and Medicine 4/2020 Go to the issue

Research Article

Pilot study utilizing MRI 3D TGSE PASL (arterial spin labeling) differentiating clearance rates of labeled protons in the CNS of patients with early Alzheimer disease from normal subjects

Research Article

Performing IVIM-DWI using the multifunctional nanosystem for the evaluation of the antitumor microcirculation changes

Research Article

Comparison between synthetic and conventional magnetic resonance imaging in patients with multiple sclerosis and controls

Correction

Correction to: Development of a new phantom simulating extracellular space of tumor cell growth and cell edema for diffusion-weighted magnetic resonance imaging

Research Article

A 12-channel flexible receiver coil for accelerated tongue imaging

Research Article

Stability evaluation of Gd chelates for macromolecular MRI contrast agents