Learning while Acquisition: Towards Active Learning Framework for Beamforming in Ultrasound Imaging

TL;DR

This paper introduces an active learning framework for ultrasound beamforming with neural networks, enabling real-time training during data acquisition and reducing the need for large pre-collected datasets.

Contribution

It presents the first active learning approach for neural network-based ultrasound beamforming that trains concurrently with data acquisition, improving efficiency and adaptability.

Findings

Training takes approximately 0.5 seconds per iteration.

The framework allows real-time model updates during data collection.

It reduces the reliance on large pre-existing datasets.

Abstract

In the recent past, there have been many efforts to accelerate adaptive beamforming for ultrasound (US) imaging using neural networks (NNs). However, most of these efforts are based on static models, i.e., they are trained to learn a single adaptive beamforming approach (e.g., minimum variance distortionless response (MVDR)) assuming that they result in the best image quality. Moreover, the training of such NNs is initiated only after acquiring a large set of data that consumes several gigabytes (GBs) of storage space. In this study, an active learning framework for beamforming is described for the first time in the context of NNs. The best quality image chosen by the user serves as the ground truth for the proposed technique, which trains the NN concurrently with data acqusition. On average, the active learning approach takes 0.5 seconds to complete a single iteration of training.