More Like this

1. Ultrasound Waveforms with and without Ringdown Artifacts

   https://doi.org/10.21227/z6v5-mf23  

   Sosnovskaya, Yana Sosnovskaya (January 2022, IEEE DataPort)

   Minimally-Invasive Surgeries can benefit from having miniaturized sensors on surgical graspers to provide additional information to the surgeons. One such potential sensor is an ultrasound transducer. At long travel distances, the ultrasound transducer can accurately measure its ultrasound wave's time of flight, and from it, classify the grasped tissue. However, the ultrasound transducer has a ringing artifact arising from the decaying oscillation of its piezo element, and at short travel distances, the artifact blends with the acoustic echo. Without a method to remove the artifact from the blended signal, this makes it impossible to measure the waveform's time of flight.It is possible to use both classical signal processing and deep learning methods to filter raw ultrasound signals, removing the ringing artifact from them, and from the filtered signals, to obtain the time of flight. In this dataset, two datasets are provided to train and test algorithms developed for filtering out the ringdown artifact, and for subsequently extracting the waveform's time of flight. All measured (raw) signals were collected the same experimental setup: an oscilloscope connected to an ultrasound driver to drive a transducer attached to a liquid water container, in an attempt to mimic tissue properties in a tightly controlled environment.The training dataset consists of two groups of signal pairs. The first group consists of 993 signal pairs, with each pair consisting of a raw ultrasound signal (with the acoustic echo blended with the ringing artifact), and a target filtered signal (with only the desired echo). Signals in the first group are sampled at the original sampling frequency of 500 MHz. The second group is like the first group, but with all signals downsampled by a factor of 26. This training dataset includes only travel distances from 2 cm to 4 cm, inclusively, because at these distances in water, the echo is sufficiently separated from the ringdown artifact to be manually extractable. The signal pairs are approximately equally distributed between the distances covered.The test dataset similarly consists of two groups of raw ultrasound signals. The first group consists of 270 signals, collected at 9 travel distances between 0.5 cm and 4.0 cm, with 30 signals per distance. It also includes the associated true times of flight for each distance. Signals in the first group are sampled at the original sampling frequency of 500 MHz. The second group is like the first group, but with all signals downsampled by a factor of 26. All signals in both datasets are aligned. 

   more » « less
2. Improving Control Precision and Motion Adaptiveness for Surgical Robot with Recurrent Neural Network

   Li, Yangming; Li, Shuai; Caballero, David; Miyasaka, Muneaki; Lewis, Andrew; Hannaford, Blake (September 2017, Proceedings of the ... IEEE/RSJ International Conference on Intelligent Robots and Systems)

   Surgical robot research is driven by the desire of improving surgical outcomes. This paper proposed a Recurrent Neural Network based controller to address two problems: 1) improving control precision, 2) increasing adaptiveness for robot motion (explained in Section I). RNN was adopted in this work mainly because 1) the problem formulation naturally matches RNN structure, 2) RNN has advantages as an biologi- cally inspired method. The proposed method was explained in detail and analysis shows that the proposed method is able to dynamically regulate outputs to increase the adaptiveness and the control precision. This paper uses Raven II surgical robot as an example to show the application of the proposed method, and the numeral simulation results from the proposed method and three other controllers show that the proposed method has improved precision, improved high robustness against noise and increased movement smoothness, and it keeps the manipulator links as far away as possible from physical boundaries, which potentially increases surgical safety and leads to improved surgical outcomes. 

   more » « less
3. Independent Component-Based Spatiotemporal Clutter Filtering for Slow Flow Ultrasound

   https://doi.org/10.1109/TMI.2019.2951465  

   Tierney, Jaime; Baker, Jennifer; Brown, Daniel; Wilkes, Don; Byram, Brett (January 2019, IEEE Transactions on Medical Imaging)

   (Early Access) Effective tissue clutter filtering is critical for non-contrast ultrasound imaging of slow blood flow in small vessels. Independent component analysis (ICA) has been considered by other groups for ultrasound clutter filtering in the past and was shown to be superior to principal component analysis (PCA)-based methods. However, it has not been considered specifically for slow flow applications or revisited since the onset of other slow flow-focused advancements in beamforming and tissue filtering, namely angled plane wave beamforming and full spatiotemporal singular value decomposition (SVD) (i.e., PCA-based) tissue filtering. In this work, we aim to develop a full spatiotemporal ICA-based tissue filtering technique facilitated by plane wave applications and compare it to SVD filtering. We compare ICA and SVD filtering in terms of optimal image quality in simulations and phantoms as well as in terms of optimal correlation to ground truth blood signal in simulations. Additionally, we propose an adaptive blood independent component sorting and selection method. We show that optimal and adaptive ICA can consistently separate blood from tissue better than principal component analysis (PCA)-based methods using simulations and phantoms. Additionally we demonstrate initial in vivo feasibility in ultrasound data of a liver tumor. 

   more » « less
4. On Neural Network Training from Noisy Data using a Novel Filtering Framework

   https://doi.org/10.2514/6.2020-1869  

   Deshpande, Vedang; Das, Niladri; Tadiparthi, Vaishnav; Bhattacharya, Raktim (January 2020, SciTech 2020 Forum)

   null (Ed.)

   We discuss a novel method to train a neural network from noisy data, using Optimal Transport based filtering. We show a comparative study of this methodology with three other filters: the Extended Kalman filter, the Ensemble Kalman filter, and the Unscented Kalman filter, that can also be used for the purpose of training a neural network. We empirically establish that Optimal Transport based filter performs better than the other three filters with respect to root mean square error measure, for non-Gaussian noise in the output. We demonstrate the efficacy of utilizing the Optimal Transport based filtering for neural network training in the context of predicting Mackey-Glass chaotic time series data. 

   more » « less
5. E-RNN: Design Optimization for Efficient Recurrent Neural Networks in FPGAs

   https://doi.org/10.1109/HPCA.2019.00028  

   Li, Zhe; Ding, Caiwen; Wang, Siyue; Wen, Wujie; Zhuo, Youwei; Liu, Chang; Qiu, Qinru; Xu, Wenyao; Lin, Xue; Qian, Xuehai; et al (February 2019, 2019 IEEE International Symposium on High Performance Computer Architecture (HPCA))

   Recurrent Neural Networks (RNNs) are becoming increasingly important for time series-related applications which require efficient and real-time implementations. The two major types are Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks. It is a challenging task to have real-time, efficient, and accurate hardware RNN implementations because of the high sensitivity to imprecision accumulation and the requirement of special activation function implementations. Recently two works have focused on FPGA implementation of inference phase of LSTM RNNs with model compression. First, ESE uses a weight pruning based compressed RNN model but suffers from irregular network structure after pruning. The second work C-LSTM mitigates the irregular network limitation by incorporating block-circulant matrices for weight matrix representation in RNNs, thereby achieving simultaneous model compression and acceleration. A key limitation of the prior works is the lack of a systematic design optimization framework of RNN model and hardware implementations, especially when the block size (or compression ratio) should be jointly optimized with RNN type, layer size, etc. In this paper, we adopt the block-circulant matrixbased framework, and present the Efficient RNN (E-RNN) framework for FPGA implementations of the Automatic Speech Recognition (ASR) application. The overall goal is to improve performance/energy efficiency under accuracy requirement. We use the alternating direction method of multipliers (ADMM) technique for more accurate block-circulant training, and present two design explorations providing guidance on block size and reducing RNN training trials. Based on the two observations, we decompose E-RNN in two phases: Phase I on determining RNN model to reduce computation and storage subject to accuracy requirement, and Phase II on hardware implementations given RNN model, including processing element design/optimization, quantization, activation implementation, etc. 1 Experimental results on actual FPGA deployments show that E-RNN achieves a maximum energy efficiency improvement of 37.4× compared with ESE, and more than 2× compared with C-LSTM, under the same accuracy. 

   more » « less