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

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

doi:10.1007/s10334-019-00816-5

Abstract

ObjectiveSegmentation 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 methodsThe 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.ResultsThe 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).DiscussionAutomated 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.

Highlights

In the clinical evaluation study (Fig. 5), painful knees displayed significantly lower quadriceps cross-sectional areas (CSA) for analyses performed with both segmentation methods than painless contralateral knees (Table 3)
The current U-Net segmentation method was able to reproduce the results from a previous clinical study, in which we had observed that the quadriceps of limbs with frequently painful knees shows lower CSAs compared with contralateral knees without knee pain
The results from the current study showed high agreement (DSCs > 0.95) between the fully automated U-Net vs. manual segmentation approach for SCF, quadriceps, hamstrings, and femoral bone segmentations, independent of whether the algorithm was compared to the segmentations of the same or a different reader

Summary

Introduction

Thigh muscle deficits [1, 2] and accumulation of (local) adipose tissue [3,4,5] are important pathophysiological events in the context of the clinical science and management of1 3 Vol.:(0123456789)Magnetic Resonance Materials in Physics, Biology and Medicine (2020) 33:483–493There exist several semi-automated [16,17,18,19,20,21] and fully automated [22,23,24,25] tools for thigh tissue volume and CSA segmentation to overcome the challenges in capturing the complex morphology and texture of thigh muscle and adipose tissue that are complicated by considerable intersubject variability (Fig. 1) and potentially artefacts as intensity distortions. With more data becoming available and recent advances in machine learning and computing infrastructure, segmentation techniques based on deep convolutional neural networks (CNN) are emerging as the new state-of-the-art [32, 33]. For this reason, CNNs are recently examined for musculoskeletal tissue segmentation of the knee joint [34,35,36] and thigh muscle MRIs by Ahmad et al [37] and our group [38]. Ahmad et al explored five pre-trained fully convolutional networks (FCN)

Methods

Results

Discussion

Conclusion