Automated diagnosis of myositis from muscle ultrasound: Exploring the use of machine learning and deep learning methods

Abstract

Objective: To evaluate the use of ultrasound coupled with machine learning (ML) and deep learning (DL) techniques for automated or semi-automated classification of myositis.

Methods: Eighty subjects comprised of 19 with inclusion body myositis (IBM), 14 with polymyositis (PM), 14 with dermatomyositis (DM), and 33 normal (N) subjects were included in this study, where 3214 muscle ultrasound images of 7 muscles (observed bilaterally) were acquired. We considered three problems of classification including (A) normal vs. affected (DM, PM, IBM); (B) normal vs. IBM patients; and (C) IBM vs. other types of myositis (DM or PM). We studied the use of an automated DL method using deep convolutional neural networks (DL-DCNNs) for diagnostic classification and compared it with a semi-automated conventional ML method based on random forests (ML-RF) and "engineered" features. We used the known clinical diagnosis as the gold standard for evaluating performance of muscle classification.

Results: The performance of the DL-DCNN method resulted in accuracies ± standard deviation of 76.2% ± 3.1% for problem (A), 86.6% ± 2.4% for (B) and 74.8% ± 3.9% for (C), while the ML-RF method led to accuracies of 72.3% ± 3.3% for problem (A), 84.3% ± 2.3% for (B) and 68.9% ± 2.5% for (C).

Conclusions: This study demonstrates the application of machine learning methods for automatically or semi-automatically classifying inflammatory muscle disease using muscle ultrasound. Compared to the conventional random forest machine learning method used here, which has the drawback of requiring manual delineation of muscle/fat boundaries, DCNN-based classification by and large improved the accuracies in all classification problems while providing a fully automated approach to classification.

Fig 1. Example ultrasound images.

Examples of ultrasound images for both healthy and affected individuals are shown for each muscle group studied. Each row represents one muscle group. The first column contains images of healthy individuals, whereas the second column contains images of patients suffering from myositis. The third and fourth columns show the manual segmentations of muscle and fat tissues corresponding to these images as red (for muscle) and green (for subcutaneous fat) overlays. The muscle group/disease type represented by each row are as follows. A: biceps/DM. B: deltoid/PM. C: FCR/IBM. D: FDP/IBM. E: gastrocnemius/PM. F: rectus femoris/PM. G: tibialis anterior/IBM.

Fig 2. DCNN architecture.

This figure depicts the architecture of the AlexNet DCNN used in this study. The muscle images are input at left and the final class probabilities for categorization are output at right. Layers C1-C5 are convolutional layers, followed by fully connected layers (FC6 and FC7), and finally by the Softmax layer outputting the probabilities of the image corresponding to each disease. (For further architectural details, see the original AlexNet paper by Krizhevsky [44]).