3D Automated Segmentation of Lower Leg Muscles Using Machine Learning on a Heterogeneous Dataset
Quantitative MRI combines non-invasive imaging techniques to reveal alterations in muscle pathophysiology. Creating muscle-specific labels manually is time consuming and requires an experienced examiner. Semi-automatic and fully automatic methods reduce segmentation time significantly. Current machi...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-09-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/11/10/1747 |
| _version_ | 1797514871143661568 |
|---|---|
| author | Marlena Rohm Marius Markmann Johannes Forsting Robert Rehmann Martijn Froeling Lara Schlaffke |
| author_facet | Marlena Rohm Marius Markmann Johannes Forsting Robert Rehmann Martijn Froeling Lara Schlaffke |
| author_sort | Marlena Rohm |
| collection | DOAJ |
| description | Quantitative MRI combines non-invasive imaging techniques to reveal alterations in muscle pathophysiology. Creating muscle-specific labels manually is time consuming and requires an experienced examiner. Semi-automatic and fully automatic methods reduce segmentation time significantly. Current machine learning solutions are commonly trained on data from healthy subjects using homogeneous databases with the same image contrast. While yielding high Dice scores (DS), those solutions are not applicable to different image contrasts and acquisitions. Therefore, the aim of our study was to evaluate the feasibility of automatic segmentation of a heterogeneous database. To create a heterogeneous dataset, we pooled lower leg muscle images from different studies with different contrasts and fields-of-view, containing healthy controls and diagnosed patients with various neuromuscular diseases. A second homogenous database with uniform contrasts was created as a subset of the first database. We trained three 3D-convolutional neuronal networks (CNN) on those databases to test performance as compared to manual segmentation. All networks, training on heterogeneous data, were able to predict seven muscles with a minimum average DS of 0.75. U-Net performed best when trained on the heterogeneous dataset (DS: 0.80 ± 0.10, AHD: 0.39 ± 0.35). ResNet and DenseNet yielded higher DS, when trained on a heterogeneous dataset (both DS: 0.86), as compared to a homogeneous dataset (ResNet DS: 0.83, DenseNet DS: 0.76). In conclusion, a CNN trained on a heterogeneous dataset achieves more accurate labels for predicting a heterogeneous database of lower leg muscles than a CNN trained on a homogenous dataset. We propose that a large heterogeneous database is needed, to make automated segmentation feasible for different kinds of image acquisitions. |
| first_indexed | 2024-03-10T06:37:41Z |
| format | Article |
| id | doaj.art-8d9a33858eac4fb8af984826d3995dbb |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-10T06:37:41Z |
| publishDate | 2021-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-8d9a33858eac4fb8af984826d3995dbb2023-11-22T17:56:07ZengMDPI AGDiagnostics2075-44182021-09-011110174710.3390/diagnostics111017473D Automated Segmentation of Lower Leg Muscles Using Machine Learning on a Heterogeneous DatasetMarlena Rohm0Marius Markmann1Johannes Forsting2Robert Rehmann3Martijn Froeling4Lara Schlaffke5Department of Neurology, BG-University Hospital Bergmannsheil gGmbH, Ruhr-University Bochum, 44789 Bochum, GermanyDepartment of Neurology, BG-University Hospital Bergmannsheil gGmbH, Ruhr-University Bochum, 44789 Bochum, GermanyDepartment of Neurology, BG-University Hospital Bergmannsheil gGmbH, Ruhr-University Bochum, 44789 Bochum, GermanyDepartment of Neurology, BG-University Hospital Bergmannsheil gGmbH, Ruhr-University Bochum, 44789 Bochum, GermanyDepartment of Radiology, University Medical Centre Utrecht, 3584 Utrecht, The NetherlandsDepartment of Neurology, BG-University Hospital Bergmannsheil gGmbH, Ruhr-University Bochum, 44789 Bochum, GermanyQuantitative MRI combines non-invasive imaging techniques to reveal alterations in muscle pathophysiology. Creating muscle-specific labels manually is time consuming and requires an experienced examiner. Semi-automatic and fully automatic methods reduce segmentation time significantly. Current machine learning solutions are commonly trained on data from healthy subjects using homogeneous databases with the same image contrast. While yielding high Dice scores (DS), those solutions are not applicable to different image contrasts and acquisitions. Therefore, the aim of our study was to evaluate the feasibility of automatic segmentation of a heterogeneous database. To create a heterogeneous dataset, we pooled lower leg muscle images from different studies with different contrasts and fields-of-view, containing healthy controls and diagnosed patients with various neuromuscular diseases. A second homogenous database with uniform contrasts was created as a subset of the first database. We trained three 3D-convolutional neuronal networks (CNN) on those databases to test performance as compared to manual segmentation. All networks, training on heterogeneous data, were able to predict seven muscles with a minimum average DS of 0.75. U-Net performed best when trained on the heterogeneous dataset (DS: 0.80 ± 0.10, AHD: 0.39 ± 0.35). ResNet and DenseNet yielded higher DS, when trained on a heterogeneous dataset (both DS: 0.86), as compared to a homogeneous dataset (ResNet DS: 0.83, DenseNet DS: 0.76). In conclusion, a CNN trained on a heterogeneous dataset achieves more accurate labels for predicting a heterogeneous database of lower leg muscles than a CNN trained on a homogenous dataset. We propose that a large heterogeneous database is needed, to make automated segmentation feasible for different kinds of image acquisitions.https://www.mdpi.com/2075-4418/11/10/1747qMRImuscle segmentationmachine learning |
| spellingShingle | Marlena Rohm Marius Markmann Johannes Forsting Robert Rehmann Martijn Froeling Lara Schlaffke 3D Automated Segmentation of Lower Leg Muscles Using Machine Learning on a Heterogeneous Dataset Diagnostics qMRI muscle segmentation machine learning |
| title | 3D Automated Segmentation of Lower Leg Muscles Using Machine Learning on a Heterogeneous Dataset |
| title_full | 3D Automated Segmentation of Lower Leg Muscles Using Machine Learning on a Heterogeneous Dataset |
| title_fullStr | 3D Automated Segmentation of Lower Leg Muscles Using Machine Learning on a Heterogeneous Dataset |
| title_full_unstemmed | 3D Automated Segmentation of Lower Leg Muscles Using Machine Learning on a Heterogeneous Dataset |
| title_short | 3D Automated Segmentation of Lower Leg Muscles Using Machine Learning on a Heterogeneous Dataset |
| title_sort | 3d automated segmentation of lower leg muscles using machine learning on a heterogeneous dataset |
| topic | qMRI muscle segmentation machine learning |
| url | https://www.mdpi.com/2075-4418/11/10/1747 |
| work_keys_str_mv | AT marlenarohm 3dautomatedsegmentationoflowerlegmusclesusingmachinelearningonaheterogeneousdataset AT mariusmarkmann 3dautomatedsegmentationoflowerlegmusclesusingmachinelearningonaheterogeneousdataset AT johannesforsting 3dautomatedsegmentationoflowerlegmusclesusingmachinelearningonaheterogeneousdataset AT robertrehmann 3dautomatedsegmentationoflowerlegmusclesusingmachinelearningonaheterogeneousdataset AT martijnfroeling 3dautomatedsegmentationoflowerlegmusclesusingmachinelearningonaheterogeneousdataset AT laraschlaffke 3dautomatedsegmentationoflowerlegmusclesusingmachinelearningonaheterogeneousdataset |