Self-Attention LSTM-FCN model for arrhythmia classification and uncertainty assessment.

This paper presents ArrhyMon, a self-attention-based LSTM-FCN model for arrhythmia classification from ECG signal inputs. ArrhyMon targets to detect and classify six different types of arrhythmia apart from normal ECG patterns. To the best of our knowledge, ArrhyMon is the first end-to-end classification model that successfully targets the classification of six detailed arrhythmia types and compared to previous work does not require additional preprocessing and/or feature extraction operations separate from the classification model. ArrhyMon's deep learning model is designed to capture and exploit both global and local features embedded in ECG sequences by integrating fully convolutional network (FCN) layers and a self-attention-based long and short-term memory (LSTM) architecture. Moreover, to enhance its practicality, ArrhyMon incorporates a deep ensemble-based uncertainty model that generates a confidence-level measure for each classification result. We evaluate ArrhyMon's effectiveness using three publicly available arrhythmia datasets (i.e., MIT-BIH, Physionet Cardiology Challenge 2017 and 2020/2021) to show that ArrhyMon achieves state-of-the-art classification performance (average accuracy 99.63%), and that confidence measures show close correlation with subjective diagnosis made from practitioners.

Deep reinforcement learning-based propofol infusion control for anesthesia: A feasibility study with a 3000-subject dataset.

In this work, we present a deep reinforcement learning-based approach as a baseline system for autonomous propofol infusion control. Specifically, design an environment for simulating the possible conditions of a target patient based on input demographic data and design our reinforcement learning model-based system so that it effectively makes predictions on the proper level of propofol infusion to maintain stable anesthesia even under dynamic conditions that can affect the decision-making process, such as the manual control of remifentanil by anesthesiologists and the varying patient conditions under anesthesia. Through an extensive set of evaluations using patient data from 3000 subjects, we show that the proposed method results in stabilization in the anesthesia state, by managing the bispectral index (BIS) and effect-site concentration for a patient showing varying conditions.

Predicting speech discrimination scores from pure-tone thresholds-A machine learning-based approach using data from 12,697 subjects.

Diagnostic tests for hearing impairment not only determines the presence (or absence) of hearing loss, but also evaluates its degree and type, and provides physicians with essential data for future treatment and rehabilitation. Therefore, accurately measuring hearing loss conditions is very important for proper patient understanding and treatment. In current-day practice, to quantify the level of hearing loss, physicians exploit specialized test scores such as the pure-tone audiometry (PTA) thresholds and speech discrimination scores (SDS) as quantitative metrics in examining a patient's auditory function. However, given that these metrics can be easily affected by various human factors, which includes intentional (or accidental) patient intervention, there are needs to cross validate the accuracy of each metric. By understanding a "normal" relationship between the SDS and PTA, physicians can reveal the need for re-testing, additional testing in different dimensions, and also potential malingering cases. For this purpose, in this work, we propose a prediction model for estimating the SDS of a patient by using PTA thresholds via a Random Forest-based machine learning approach to overcome the limitations of the conventional statistical (or even manual) methods. For designing and evaluating the Random Forest-based prediction model, we collected a large-scale dataset from 12,697 subjects, and report a SDS level prediction accuracy of 95.05% and 96.64% for the left and right ears, respectively. We also present comparisons with other widely-used machine learning algorithms (e.g., Support Vector Machine, Multi-layer Perceptron) to show the effectiveness of our proposed Random Forest-based approach. Results obtained from this study provides implications and potential feasibility in providing a practically-applicable screening tool for identifying patient-intended malingering in hearing loss-related tests.

Combining hepatic surface nodularity and serum tests better predicts hepatic fibrosis stages in chronic liver disease.

PURPOSE: Hepatic surface nodularity quantified on CT images has shown promising results in staging hepatic fibrosis in chronic hepatitis C. The aim of this study was to evaluate hepatic surface nodularity, serum fibrosis indices, and a linear combination of them for staging fibrosis in chronic liver disease, mainly chronic hepatitis B. METHODS: We developed a semiautomated software quantifying hepatic surface nodularity on CT images. Hepatic surface nodularity and serum fibrosis indices were assessed in the development group of 125 patients to generate 3 linear models combining hepatic surface nodularity with the aspartate aminotransferase to platelet ratio index, fibrosis-4 index, or platelet count in reference to the METAVIR scoring system. The models were validated in 183 patients. RESULTS: Hepatic surface nodularity and serum fibrosis indices all significantly correlated with fibrosis stages. For binary classifications into cirrhosis (F4), advanced fibrosis (≥ F3), and significant fibrosis (≥ F2), hepatic surface nodularity was significantly different across categories. The areas under the curve (AUCs) of the best model were 0.901, 0.872, and 0.794 for cirrhosis, advanced fibrosis, and significant fibrosis, respectively, higher than serum fibrosis indices alone (0.797-0.802, 0.799-0.818, and 0.761-0.773). In the validation group, the same model likewise showed higher AUCs (0.872, 0.831, and 0.850) compared to serum fibrosis indices (0.722-0.776, 0.692-0.768, and 0.695-0.769; p < 0.001 for F4). CONCLUSION: Hepatic surface nodularity combined with serum blood test could be a practical method to predict cirrhosis, advanced fibrosis, and significant fibrosis in chronic liver disease patients, providing higher accuracy than using serum fibrosis indices alone.

Disentangling the clinical data chaos: User-centered interface system design for trauma centers.

This paper presents a year-long study of our project, aiming at (1) understanding the work practices of clinical staff in trauma intensive care units (TICUs) at a trauma center, with respect to their usage of clinical data interface systems, and (2) developing and evaluating an intuitive and user-centered clinical data interface system for their TICU environments. Based on a long-term field study in an urban trauma center that involved observation-, interview-, and survey-based studies to understand our target users and their working environment, we designed and implemented MediSenseView as a working prototype. MediSenseView is a clinical-data interface system, which was developed through the identification of three core challenges of existing interface system use in a trauma care unit-device separation, usage inefficiency, and system immobility-from the perspectives of three staff groups in our target environment (i.e., doctors, clinical nurses and research nurses), and through an iterative design study. The results from our pilot deployment of MediSenseView and a user study performed with 28 trauma center staff members highlight their work efficiency and satisfaction with MediSenseView compared to existing clinical data interface systems in the hospital.

Kaleidoscopic fluorescent arrays for machine-learning-based point-of-care chemical sensing.

Multiplexed analysis allows simultaneous measurements of multiple targets, improving the detection sensitivity and accuracy. However, highly multiplexed analysis has been challenging for point-of-care (POC) sensing, which requires a simple, portable, robust, and affordable detection system. In this work, we developed paper-based POC sensing arrays consisting of kaleidoscopic fluorescent compounds. Using an indolizine structure as a fluorescent core skeleton, named Kaleidolizine (KIz), a library of 75 different fluorescent KIz derivatives were designed and synthesized. These KIz derivatives are simultaneously excited by a single ultraviolet (UV) light source and emit diverse fluorescence colors and intensities. For multiplexed POC sensing system, fluorescent compounds array on cellulose paper was prepared and the pattern of fluorescence changes of KIz on array were specific to target chemicals adsorbed on that paper. Furthermore, we developed a machine-learning algorithm for automated, rapid analysis of color and intensity changes of individual sensing arrays. We showed that the paper sensor arrays could differentiate 35 different volatile organic compounds using a smartphone-based handheld detection system. Powered by the custom-developed machine-learning algorithm, we achieved the detection accuracy of 97% in the VOC detection. The highly multiplexed paper sensor could have favorable applications for monitoring a broad-range of environmental toxins, heavy metals, explosives, pathogens.

Block-building performance test using a virtual reality head-mounted display in children with intermittent exotropia.

PURPOSE: To determine whether childhood intermittent exotropia (IXT) affects distance divergence and performance in block-building tasks within a virtual reality (VR) environment. METHODS: Thirty-nine children with IXT, aged 6-12 years, who underwent muscle surgery and 37 normal controls were enrolled. Children were instructed to watch the target moving away and perform a block-building task while fitted with a VR head-mounted display equipped with eye- and hand-movement tracking systems. The change in inter-ocular distance with binocular distance viewing, time to stack five cube blocks of different sizes in order, and distance disparities between the largest and farthest cubes were assessed. All children were evaluated at baseline and 3-month time points. RESULTS: The patients with IXT exhibited a larger distance divergence than did controls (p = 0.024), which was associated with greater distance angle of deviation and poorer distance control (r = 0.350, p = 0.001 and r = 0.349, p = 0.004). At baseline, the patients with IXT showed larger distance disparities in the block-building task than did controls in terms of the horizontal, vertical, and 3-dimensional (3-D) measurements (all ps < 0.050). Larger horizontal disparity was associated with greater distance angle of deviation (r = 0.383, p = 0.037). Three months after surgery, the horizontal and 3-D disparities in the patients with IXT improved significantly and were not comparably different compared with controls. CONCLUSIONS: These preliminary findings suggest that VR-based block-building task may be useful in testing possible deficits in visuo-motor skills associated with childhood IXT.

IB-MAC: Transmission Latency-Aware MAC for Electro-Magnetic Intra-Body Communications.

Intra-body Communication (IBC) is a communication method using the human body as a communication medium, in which body-attached devices exchange electro-magnetic (EM) wave signals with each other. The fact that our human body consists of water and electrolytes allows such communication methods to be possible. Such a communication technology can be used to design novel body area networks that are secure and resilient towards external radio interference. While being an attractive technology for enabling new applications for human body-centered ubiquitous applications, network protocols for IBC systems is yet under-explored. The IEEE 802.15.6 standards present physical and medium access control (MAC) layer protocols for IBC, but, due to many simplifications, we find that its MAC protocol is limited in providing an environment to enable high data rate applications. This work, based on empirical EM wave propagation measurements made for the human body communication channel, presents IB-MAC, a centralized Time-division multiple access (TDMA) protocol that takes in consideration the transmission latency the body channel induces. Our results, in which we use an event-based simulator to compare the performance of IB-MAC with two different IEEE 802.15.6 standard-compliant MAC protocols and a state-of-the art TDMA-based MAC protocol for IBC, suggest that IB-MAC is suitable for supporting high data rate applications with comparable radio duty cycle and latency performance.

Convolutional neural network-based classification system design with compressed wireless sensor network images.

With the introduction of various advanced deep learning algorithms, initiatives for image classification systems have transitioned over from traditional machine learning algorithms (e.g., SVM) to Convolutional Neural Networks (CNNs) using deep learning software tools. A prerequisite in applying CNN to real world applications is a system that collects meaningful and useful data. For such purposes, Wireless Image Sensor Networks (WISNs), that are capable of monitoring natural environment phenomena using tiny and low-power cameras on resource-limited embedded devices, can be considered as an effective means of data collection. However, with limited battery resources, sending high-resolution raw images to the backend server is a burdensome task that has direct impact on network lifetime. To address this problem, we propose an energy-efficient pre- and post- processing mechanism using image resizing and color quantization that can significantly reduce the amount of data transferred while maintaining the classification accuracy in the CNN at the backend server. We show that, if well designed, an image in its highly compressed form can be well-classified with a CNN model trained in advance using adequately compressed data. Our evaluation using a real image dataset shows that an embedded device can reduce the amount of transmitted data by ∼71% while maintaining a classification accuracy of ∼98%. Under the same conditions, this process naturally reduces energy consumption by ∼71% compared to a WISN that sends the original uncompressed images.

System for Collecting Biosignal Data from Multiple Patient Monitoring Systems.

OBJECTIVES: Biosignal data include important physiological information. For that reason, many devices and systems have been developed, but there has not been enough consideration of how to collect and integrate raw data from multiple systems. To overcome this limitation, we have developed a system for collecting and integrating biosignal data from two patient monitoring systems. METHODS: We developed an interface to extract biosignal data from Nihon Kohden and Philips monitoring systems. The Nihon Kohden system has a central server for the temporary storage of raw waveform data, which can be requested using the HL7 protocol. However, the Philips system used in our hospital cannot save raw waveform data. Therefore, our system was connected to monitoring devices using the RS232 protocol. After collection, the data were transformed and stored in a unified format. RESULTS: From September 2016 to August 2017, we collected approximately 117 patient-years of waveform data from 1,268 patients in 79 beds of five intensive care units. Because the two systems use the same data storage format, the application software could be run without compatibility issues. CONCLUSIONS: Our system collects biosignal data from different systems in a unified format. The data collected by the system can be used to develop algorithms or applications without the need to consider the source of the data.

Constructing an Open-Access Bio-Signal Repository from Intensive Care Units.

Bio-signals can be crucial evidence in detecting urgent clinical events. However, until now, access to this data was limited. We aim to construct and provide a new open bio-signal repository with data gathered from more than 40 intensive care unit (ICU) beds. For doing so, we completed the interfacing system with the patient monitors at the target beds and plan to expand this data set to more than 100 ICU beds. Once completed, we plan to publicly open the data to catalyze interesting clinical-event detection research.

Wireless sensing systems in clinical environments: improving the efficiency of the patient monitoring process.

Multiple studies suggest that the level of patient care may decline in the future because of a larger aging population and medical staff shortages. Wireless sensing systems that automate some of the patient monitoring tasks can potentially improve the efficiency of patient workflows, but their efficacy in clinical settings is an open question. This article examines the potential of wireless sensor network (WSN) technologies to improve the efficiency of the patient-monitoring process in clinical environments. MEDiSN, a WSN designed to continuously monitor the vital signs of ambulatory patients, is designed. The usefulness of MEDiSN is validated with test bed experiments and results from a pilot study performed at the Emergency Department, Johns Hopkins Hospital. Promising results indicate that MEDiSN can tolerate high degrees of human mobility, is well received by patients and staff members, and performs well in real clinical environments.