Machine learning for medical ultrasound: status, methods, and future opportunities

Abstract

Ultrasound (US) imaging is the most commonly performed cross-sectional diagnostic imaging modality in the practice of medicine. It is low-cost, non-ionizing, portable, and capable of real-time image acquisition and display. US is a rapidly evolving technology with significant challenges and opportunities. Challenges include high inter- and intra-operator variability and limited image quality control. Tremendous opportunities have arisen in the last decade as a result of exponential growth in available computational power coupled with progressive miniaturization of US devices. As US devices become smaller, enhanced computational capability can contribute significantly to decreasing variability through advanced image processing. In this paper, we review leading machine learning (ML) approaches and research directions in US, with an emphasis on recent ML advances. We also present our outlook on future opportunities for ML techniques to further improve clinical workflow and US-based disease diagnosis and characterization.

Keywords: Deep learning; Elastography; Machine learning; Medical ultrasound; Sonography.

Fig. 1

Overview of Ultrasound processing system workflow.

Fig. 2

Example SWE color map overlaid on a B-mode ultrasound image.

Fig. 3

Conventional machine learning vs. deep learning.

Fig. 4

Supervised learning with deep neural networks.

Fig. 5

Example convolutional neural network (CNN).

Fig. 6a

Example pipeline of using SWE for liver fibrosis staging.

Fig. 6b

Proposed semi-automated SWE acquisition and analysis workflow.

Figure 7

Proposed framework for machine learning-based intelligent diagnostic assistant