A1 Journal article (refereed)
DeepACSA : Automatic Segmentation of Cross-sectional Area in Ultrasound Images of Lower Limb Muscles Using Deep Learning (2022)
Ritsche, P., Wirth, P., Cronin, N. J., Sarto, F., Narici, M. V., Faude, O., & Franchi, M. V. (2022). DeepACSA : Automatic Segmentation of Cross-sectional Area in Ultrasound Images of Lower Limb Muscles Using Deep Learning. Medicine and Science in Sports and Exercise, 54(12), 2188-2195. https://doi.org/10.1249/MSS.0000000000003010
JYU authors or editors
Publication details
All authors or editors: Ritsche, Paul; Wirth, Philipp; Cronin, Neil J.; Sarto, Fabio; Narici, Marco V.; Faude, Oliver; Franchi, Martino V.
Journal or series: Medicine and Science in Sports and Exercise
ISSN: 0195-9131
eISSN: 1530-0315
Publication year: 2022
Publication date: 05/08/2022
Volume: 54
Issue number: 12
Pages range: 2188-2195
Publisher: Lippincott Williams & Wilkins
Publication country: United Kingdom
Publication language: English
DOI: https://doi.org/10.1249/MSS.0000000000003010
Publication open access: Not open
Publication channel open access:
Web address of parallel published publication (pre-print): https://doi.org/10.1101/2021.12.27.21268258
Abstract
Muscle anatomical cross-sectional area (ACSA) can be assessed using ultrasound and images are usually evaluated manually. Here, we present DeepACSA, a deep learning approach to automatically segment ACSA in panoramic ultrasound images of the human rectus femoris (RF), vastus lateralis (VL), gastrocnemius medialis (GM) and lateralis (GL) muscles.
Methods
We trained three muscle-specific convolutional neural networks (CNNs) using 1772 ultrasound images from 153 participants (age = 38.2 years, range: 13-78). Images were acquired in 10% increments from 30 to 70% of femur length for RF and VL and at 30 and 50% of muscle length for GM and GL. During training, CNN performance was evaluated using intersection-over-union scores. We compared the performance of DeepACSA to manual analysis and a semi-automated algorithm using an unseen test set.
Results
Comparing DeepACSA analysis of the RF to manual analysis with erroneous predictions removed (3.3%) resulted in intra-class correlation (ICC) of 0.989 (95% CI 0.983;0.992), mean difference of 0.20 cm2 (0.10;0.30) and standard error of the differences (SEM) of 0.33 cm2 (0.26,0.41). For the VL, ICC was 0.97 (0.96,0.968), mean difference was 0.85 cm2 (-0.4,1.31) and SEM was 0.92 cm2 (0.73,1.09) following removal of erroneous predictions (7.7%). After removal of erroneous predictions (12.3%), GM/GL muscles demonstrated an ICC of 0.98 (0.96,0.99), a mean difference of 0.43 cm2 (0.21,0.65) and a SEM of 0.41 cm2 (0.29,0.51). Analysis duration was 4.0 s standard deviation (SD) ± 0.43 for analysis of one image in our test set using DeepACSA.
Conclusions
DeepACSA provides fast and objective segmentation of lower limb panoramic ultrasound images comparable to manual segmentation. Inaccurate model predictions occurred predominantly on low-quality images, highlighting the importance of high image quality for accurate prediction.
Keywords: muscles; ultrasonography; imaging; segmentation; machine learning; deep learning; neural networks (information technology)
Contributing organizations
Ministry reporting: Yes
Reporting Year: 2022
JUFO rating: 3