B-mode ultrasound (US) is often used to noninvasively measure skeletal muscle architecture, which contains human intent information. Extracted features from B-mode images can help improve closed-loop human–robotic interaction control when using rehabilitation/assistive devices. The traditional manual approach to inferring the muscle structural features from US images is laborious, time-consuming, and subjective among different investigators. This paper proposes a clustering-based detection method that can mimic a well-trained human expert in identifying fascicle and aponeurosis and, therefore, compute the pennation angle. The clustering-based architecture assumes that muscle fibers have tubular characteristics. It is robust for low-frequency image streams. We compared the proposed algorithm to two mature benchmark techniques: UltraTrack and ImageJ. The performance of the proposed approach showed higher accuracy in our dataset (frame frequency is 20 Hz), that is, similar to the human expert. The proposed method shows promising potential in automatic muscle fascicle orientation detection to facilitate implementations in biomechanics modeling, rehabilitation robot control design, and neuromuscular disease diagnosis with low-frequency data stream.

Robotic assistive or rehabilitative devices are promising aids for people with neurological disorders as they help regain normative functions for both upper and lower limbs. However, it remains challenging to accurately estimate human intent or residual efforts non-invasively when using these robotic devices. In this article, we propose a deep learning approach that uses a brightness mode, that is, B-mode, of ultrasound (US) imaging from skeletal muscles to predict the ankle joint net plantarflexion moment while walking. The designed structure of customized deep convolutional neural networks (CNNs) guarantees the convergence and robustness of the deep learning approach. We investigated the influence of the US imaging’s region of interest (ROI) on the net plantarflexion moment prediction performance. We also compared the CNN-based moment prediction performance utilizing B-mode US and sEMG spectrum imaging with the same ROI size. Experimental results from eight young participants walking on a treadmill at multiple speeds verified an improved accuracy by using the proposed US imaging + deep learning approach for net joint moment prediction. With the same CNN structure, compared to the prediction performance by using sEMG spectrum imaging, US imaging significantly reduced the normalized prediction root mean square error by 37.55% ($ p $ < .001) and increased the prediction coefficient of determination by 20.13% ($ p $ < .001). The findings show that the US imaging + deep learning approach personalizes the assessment of human joint voluntary effort, which can be incorporated with assistive or rehabilitative devices to improve clinical performance based on the assist-as-needed control strategy.

The present study examined the effects of different temperatures of protein-containing drink after exercise on subsequent gastric motility and energy intake in healthy young men. Twelve healthy young men completed three, 1-d trials in a random order. In all trials, the subjects ran on a treadmill for 30 min at 80% of maximum heart rate. In exercise + cold drink (2°C) and exercise + hot drink (60°C) trials, the subjects consumed 300 ml of protein-containing drink (0·34 MJ) at 2°C or 60°C over a 5-min period after exercise. In the exercise (i.e. no preload) trial, the subjects sat on a chair for 5 min after exercise. Then, the subjects sat on a chair for 30 min to measure their gastric motility with an ultrasound imaging system in all trials. Thereafter, the subjects consumed a test meal until they felt comfortably full. Energy intake in the exercise + hot drink trial was 14 % and 15 % higher than the exercise (P = 0·046, 95% CI 4·010, 482·538) trial and exercise + cold drink (P = 0·001, 95% CI 160·089, 517·111) trial, respectively. The frequency of the gastric contractions in the exercise + hot drink trial was higher than the exercise (P = 0·023) trial and exercise + cold drink (P = 0·007) trial. The total frequency of gastric contractions was positively related to energy intake (r = 0·386, P = 0·022). These findings demonstrate that consuming protein-containing drink after exercise at 60°C increases energy intake and that this increase may be related to the modulation of the gastric motility.

During early lactation, many dairy cows develop fatty liver, which is associated with decreased health and reproductive performance. Currently, fatty liver can be detected reliably only by using liver biopsy followed by chemical or histological analysis, which is not practical in most on-farm situations. We tested whether digital analyses of hepatic ultrasonograms can be used to detect non-invasively fatty liver and estimate liver triacylglycerol content. A total of 49 liver biopsies and ultrasonograms were taken from 29 dairy cows within 2 weeks postpartum. The usefulness of 17 first- or second-order parameters from digital analysis of B-mode ultrasonograms were evaluated by discriminant, correlation, and regression analyses. A group of linear combinations of the 17 parameters correctly classified 40 of 49 samples into normal liver as well as mild, moderate and severe fatty liver when cut-off values were 1%, 5% and 10% and correctly classified 45 of 49 samples when cut-off values were 5% and 10% triacylglycerol of wet weight. A linear combination of 16 image parameters estimated triacylglycerol concentrations of 38 of the 39 liver samples below the cut-off value of 10% within 2·5% of liver wet weight, and a linear combination of 3 parameters estimated triacylglycerol concentrations of the 10 liver samples above the cut-off value of 10% within 2% of liver wet weight. Therefore, ultrasound imaging followed by digital analysis of sonograms has potential to non-invasively detect fatty liver and estimate liver triacylglycerol content.