CigTrak: Smartwatch-Based Accurate Online Smoking Puff & Episode Detection with Gesture-Focused Windowing for CNN.

Wearable-based motion sensing solutions are capable of automatically detecting and tracking individual smoking puffs and/or episodes to aid the users in their journey of smoking cessation. But they are either obtrusive to use, perform with a low accuracy, or have questionable ability of running fully on a low-power device like a smartwatch, all affecting their widespread adoption. We propose 'CigTrak', a novel pipeline for an accurate smoking puff and episode detection using 6-DoF motion sensor on a smartwatch. A multi-stage method for puff detection is devised, comprising a novel kinematic analysis of puffing motion enabling temporal localization of puff. A Convolutional Neural Network (CNN)-backed model uses this candidate puff as an input instance by re-sampling it to required input size for the final decision. Clusters of detected puffs are further used to detect episodes. Data from 13 subjects was used for evaluating puff detection, and 9 subjects for evaluating episode detection. CigTrak achieved a high subject-independent performance for puff detection (F1-score 0.94) and free-living episode detection (F1-score 0.89), surpassing state of the art performance. CigTrak was also implemented fully online on two different smartwatches for testing a real-time puff detection.Clinical Relevance- Cigarette smoking affects physical & mental well-being of a person, and is the leading cause of preventable diseases, adversely affecting cardiac and respiratory systems. With many adults wanting to quit smoking [1], a reliable way of auto-journaling of smoking activities can greatly aid in cessation efforts through self-help, and reduce burden on healthcare industry. CigTrak, with its high accuracy in detecting smoking puffs and episodes, and capability of running fully on a smartwatch, can be readily used for this purpose.

An Efficient Service-based System for Hierarchical Human Activity Sensing.

In this paper, we propose an end-to-end system, based on SEnsing as Service (SEAS) model, which processes continuous mobility data from multiple sensors on the client edge-device by optimizing the on-device processing pipelines. Thus, reducing the cost of data transfer and CPU usage. We also propose a classification algorithm as a part of the system to recognize Activities of Daily Living (ADL). The results indicate that our proposed system recognizes ADLs with considerable accuracy and flexibility.Clinical relevance- Measurement of Activities of Daily Living has a high correlation with independent living measures for elderly people [1] and post-event rehabilitation where an event may be heart-attack [2], stroke [3], surgical intervention [4], or trauma [5] etc.

Automated Shoulder Implant Manufacturer Detection using Encoder Decoder based Classifier from X-ray Images.

Total shoulder arthroplasty is the process of replacing the damaged ball and socket joint in the shoulder with a prosthesis made with polyethylene and metal components. The prosthesis helps to restore the normal range of motion and reduce pain, enabling the patient to return to their daily activities. These implants may need to be replaced over the years due to damage or wear and tear. It is a tedious and time-consuming process to identify the type of implant if medical records are not properly maintained. Artificial intelligence systems can speed up the treatment process by classifying the manufacturer and model of the prosthesis. We have proposed an encoder-decoder based classifier along with the supervised contrastive loss function that can identify the implant manufacturer effectively with increased accuracy of 92% from X-ray images overcoming the class imbalance problem.

Cardiac Anomaly Detection from Cine MRI Images using Physiological Features and Random Forest Classifier.

Computer-aided diagnosis (CAD) with cine MRI is a foremost research topic to enable improved, faster, and more accurate diagnosis of cardiovascular diseases (CVD). However, current approaches that use manual visualization or conventional clinical indices can lack accuracy for borderline cases. Also, manual visualization of 3D/4D MR data is time-consuming and expert-dependent. We try to simplify this process by creating an end-to-end automated CAD system that segments the critical substructures of the heart. The new domain-related physiological features are then calculated from the segmented regions. These features are fed to a random forest classifier that identifies the anomaly. We have obtained a very high accuracy when testing this end-to-end approach on the Automated Cardiac Diagnosis challenge (ACDC) dataset (4 pathologies, 1 normal). To prove the generalizability of the method we have blind-tested this approach on M&Ms-2 dataset which is a multi-center, multi-vendor, and multi-disease dataset with better than 90% accuracy.

3D Cardiac Substructures Segmentation from CMRI using Generative Adversarial Network (GAN).

Cardiac magnetic resonance imaging (CMRI) improves the diagnosis of cardiovascular diseases by providing images at high spatio-temporal resolution helping physicians in providing correct treatment plans. Segmentation and identification of various substructures of the heart at different cardiac phases of end-systole and end-diastole helps in the extraction of ventricular function information such as stroke volume, ejection fraction, myocardium thickness, etc. Manual delineation of the substructures is tedious, time-consuming, and error-prone. We have implemented a 3D GAN that includes 3D contextual information capable of segmenting and identifying the substructures at different cardiac phases with improved accuracy. Our method is evaluated on the ACDC dataset (4 pathologies, 1 healthy group) to show that the proposed out-performs other methods in literature with less amount of data. Also, the proposed provided a better Dice score in segmentation surpassing other methods on a blind-tested M&Ms dataset.

Synthetic PPG Signal Generation to Improve Coronary Artery Disease Classification: Study With Physical Model of Cardiovascular System.

This paper presents a novel approach of generating synthetic Photoplethysmogram (PPG) data using a physical model of the cardiovascular system to improve classifier performance with a combination of synthetic and real data. The physical model is an in-silico cardiac computational model, consisting of a four-chambered heart with electrophysiology, hemodynamic, and blood pressure auto-regulation functionality. Starting with a small number of measured PPG data, the cardiac model is used to synthesize healthy as well as PPG time-series pertaining to coronary artery disease (CAD) by varying pathophysiological parameters. A Variational Autoencoder (VAE) structure is proposed to derive a statistical feature space for CAD classification. Results are presented in two perspectives namely, (i) using artificially reduced real disease data and (ii) using all the real disease data. In both cases, by augmenting with the synthetic data for training, the performance (sensitivity, specificity) of the classifier changes from (i) (0.65, 1) to (1, 0.9) and (ii) (1, 0.95) to (1, 1). The proposed hybrid approach of combining physical modelling and statistical feature space selection generates realistic PPG data with pathophysiological interpretation and can outperform a baseline Generative Adversarial Network (GAN) architecture with a relatively small amount of real data for training. This proposed method could aid as a substitution technique for handling the problem of bulk data required for training machine learning algorithms for cardiac health-care applications.

Resource Constrained CVD Classification Using Single Lead ECG On Wearable and Implantable Devices.

Electrocardiogram (ECG) is one of the fundamental markers to detect different cardiovascular diseases (CVDs). Owing to the widespread availability of ECG sensors (single lead) as well as smartwatches with ECG recording capability, ECG classification using wearable devices to detect different CVDs has become a basic requirement for a smart healthcare ecosystem. In this paper, we propose a novel method of model compression with robust detection capability for CVDs from ECG signals such that the sophisticated and effective baseline deep neural network model can be optimized for the resource constrained micro-controller platform suitable for wearable devices while minimizing the performance loss. We employ knowledge distillation-based model compression approach where the baseline (teacher) deep neural network model is compressed to a TinyML (student) model using piecewise linear approximation. Our proposed ECG TinyML has achieved ~156x compression factor to suit to the requirement of 100KB memory availability for model deployment on wearable devices. The proposed model requires ~5782 times (estimated) less computational load than state-of-the-art residual neural network (ResNet) model with negligible performance loss (less than 1% loss in test accuracy, test sensitivity, test precision and test F1-score). We further feel that the small footprint model size of ECG TinyML (62.3 KB) can be suitably deployed in implantable devices including implantable loop recorder (ILR).

Precise Bleeding and Red lesions localization from Capsule Endoscopy using Compact U-Net.

Wireless capsule endoscopy is a non-invasive and painless procedure to detect anomalies from the gastrointestinal tract. Single examination results in up to 8 hrs of video and requires between 45 - 180 mins for diagnosis depending on the complexity. Image and video computational methods are needed to increase both efficiency and accuracy of the diagnosis. In this paper, a compact U-Net with lesser encoder-decoder pairs is presented, to detect and precisely segment bleeding and red lesions from endoscopy data. The proposed compact U-Net is compared with the original U-Net and also with other methods reported in the literature. The results show the proposed compact network performs on par with the original network but with faster training and lesser memory consumption. Also, the proposed model provided a dice score of 91% outperforming other methods reported on a blind tested WCE dataset with no images from this set used for training.

Heart Region Segmentation using Dense VNet from Multimodality Images.

Cardiovascular diseases (CVD) have been identified as one of the most common causes of death in the world. Advanced development of imaging techniques is allowing timely detection of CVD and helping physicians in providing correct treatment plans in saving lives. Segmentation and Identification of various substructures of the heart are very important in modeling a digital twin of the patient-specific heart. Manual delineation of various substructures of the heart is tedious and time-consuming. Here we have implemented Dense VNet for detecting substructures of the heart from both CT and MRI multimodality data. Due to the limited availability of data we have implemented an on-the-fly elastic deformation data augmentation technique. The result of the proposed has been shown to outperform other methods reported in the literature on both CT and MRI datasets.

Anatomical Landmark Detection using Deep Appearance-Context Network.

Accurate identification of anatomical landmarks is a crucial step in medical image analysis. While deep neural networks have shown impressive performance on computer vision tasks, they rely on a large amount of data, which is often not available. In this work, we propose an attention-driven end-to-end deep learning architecture, which learns the local appearance and global context separately that helps in stable training under limited data. The experiments conducted demonstrate the effectiveness of the proposed approach with impressive results in localizing landmarks when evaluated on cephalometric and spine X-ray image data. The predicted landmarks are further utilized in biomedical applications to demonstrate the impact.

Walking Pole Gait to Reduce Joint Loading post Total Knee Athroplasty: Musculoskeletal modeling Approach.

Excessive knee contact loading is precursor to osteoarthritis and related knee ailment leading to knee athroplasty. Reducing contact loading through gait modifications using assisted pole walking offers noninvasive process of medial load offloading at knee joint. In this paper, we evaluate the efficacy of different configuration of pole walking for reducing contact force at the knee joint through musculoskeletal (MSK) modeling. We have developed a musculoskeletal model for a subject with knee athroplasty utilizing in-vivo implant data and computed tibio-femoral contact force for different pole walking conditions to evaluate the best possible configuration for guiding rehabilitation, correlated with different gait phases. Effect of gait speed variation on knee contact force, hip joint dynamics and muscle forces are simulated using the developed MSK model. Results indicate some interesting trend of load reduction, dependent on loading phases pertaining to different pole configuration. Insights gained from the simulation can aid in designing personalized rehabilitation therapy for subjects suffering from Osteoarthritis.

Computational Model for Therapy Optimization of Wearable Cardioverter Defibrillator: Shockable Rhythm Detection and Optimal Electrotherapy.

Wearable cardioverter defibrillator (WCD) is a life saving, wearable, noninvasive therapeutic device that prevents fatal ventricular arrhythmic propagation that leads to sudden cardiac death (SCD). WCD are frequently prescribed to patients deemed to be at high arrhythmic risk but the underlying pathology is potentially reversible or to those who are awaiting an implantable cardioverter-defibrillator. WCD is programmed to detect appropriate arrhythmic events and generate high energy shock capable of depolarizing the myocardium and thus re-initiating the sinus rhythm. WCD guidelines dictate very high reliability and accuracy to deliver timely and optimal therapy. Computational model-based process validation can verify device performance and benchmark the device setting to suit personalized requirements. In this article, we present a computational pipeline for WCD validation, both in terms of shock classification and shock optimization. For classification, we propose a convolutional neural network-"Long Short Term Memory network (LSTM) full form" (Convolutional neural network- Long short term memory network (CNN-LSTM)) based deep neural architecture for classifying shockable rhythms like Ventricular Fibrillation (VF), Ventricular Tachycardia (VT) vs. other kinds of non-shockable rhythms. The proposed architecture has been evaluated on two open access ECG databases and the classification accuracy achieved is in adherence to American Heart Association standards for WCD. The computational model developed to study optimal electrotherapy response is an in-silico cardiac model integrating cardiac hemodynamics functionality and a 3D volume conductor model encompassing biophysical simulation to compute the effect of shock voltage on myocardial potential distribution. Defibrillation efficacy is simulated for different shocking electrode configurations to assess the best defibrillator outcome with minimal myocardial damage. While the biophysical simulation provides the field distribution through Finite Element Modeling during defibrillation, the hemodynamic module captures the changes in left ventricle functionality during an arrhythmic event. The developed computational model, apart from acting as a device validation test-bed, can also be used for the design and development of personalized WCD vests depending on subject-specific anatomy and pathology.

Selective Discrimination of VOCs Applying Gas Sensing Kinetic Analysis over a Metal Oxide-Based Chemiresistive Gas Sensor.

Semiconducting metal oxide-based gas sensors have inadequate selectivity as they are responsive toward a variety of gases. Here, we report the implementation of gas sensing kinetic analysis of the sensor to identify the tested volatile organic compounds (VOCs) (2-propanol, formaldehyde, methanol, and toluene) precisely. A single chemiresistive sensor was employed having tin oxide-based hollow spheres as the sensing material, which were obtained by chemical synthesis. The gas sensing measurements were conducted in a dynamic manner where the sensor displayed excellent response with high sensitivity. Eley-Rideal model was adopted to obtain the kinetic properties of the gas sensing phenomenon through theoretical fitting of response transient curves and their corresponding kinetic parameters. The calculated characteristic kinetic properties were further examined to discriminate among different VOCs. The approach of using gas sensing kinetic analysis for multiple gas discrimination is an attractive solution to mitigate the problem of cross-sensitivity for resistive gas sensors.

Parameter Estimation of Hemodynamic Cardiovascular Model for Synthesis of Photoplethysmogram Signal.

Synthesis of accurate, personalize photoplethysmogram (PPG) signal is important to interpret, analyze and predict cardiovascular disease progression. Generative models like Generative Adversarial Networks (GANs) can be used for signal synthesis, however, they are difficult to map to the underlying pathophysiological conditions. Hence, we propose a PPG synthesis strategy that has been designed using a cardiovascular system, modeled through the hemodynamic principle. The modeled architecture is composed of a two-chambered heart along with the systemic-pulmonic blood circulation and a baroreflex auto-regulation mechanism to control the arterial blood pressure. The comprehensive PPG signal is synthesized from the cardiac pressure-flow dynamics. In order to tune the modeled cardiac parameters with respect to a measured PPG data, a novel feature extraction strategy has been employed along with the particle swarm optimization heuristics. Our results demonstrate that the synthesized PPG is accurately followed the morphological changes of the ground truth (GT) signal with an RMSE of 0.003 occurring due to the Coronary Artery Disease (CAD) which is caused by an obstruction in the artery.

Surgical tool segmentation and localization using spatio-temporal deep network.

Laparoscopic cholecystectomy surgery is a minimally invasive surgery to remove the gallbladder, where surgical instruments are inserted through small incisions in the abdomen with the help of a laparoscope. Identification of tool presence and precise segmentation of tools from the video is very important in understanding the quality of the surgery and training budding surgeons. Precise segmentation of tools is required to track the tools during real-time surgeries. In this paper, a new pixel-wise instance segmentation algorithm is proposed, which segments and localizes the surgical tool using spatio-temporal deep network. The performance of the proposed has been compared with the state-of-the-art image-based instance segmentation method using the Cholec80 dataset. It is also compared with methods in the literature using frame-level presence detection and spatial detection with good results.

Detection of Counterfeit Medicines Using Hyperspectral Sensing.

Worldwide revenue of pharmaceutical market is more than 1200 billion USD [1] and that of counterfeit medicines is around 200 billion USD [2][3]. Counterfeit medicines can be detected by technical experts using visual inspection or through sophisticated lab and relevant methods. However, such methods require time, sample preparation and technical expertise with lab setup. These methods are not feasible and scalable to be used in the field by the general public. The objective of our research work was to detect counterfeit medicines using simpler and faster method using hyperspectral sensing. In this experiment, a visible - near infrared (350nm - 1050nm) hyperspectral device was used to capture spectral signature of the medicines. We used 24 medicine tablets of different companies. To imitate counterfeit medicines, tablet powders were adulterated by adding different levels of calcium carbonate. Spectral signatures were captured from original stage to all stages of adulterations and analyzed using machine learning (multilayer perceptron classifier). Result shows that we are able to achieve more than 90% classification accuracy. Portable hyperspectral sensing combined with medicines spectral database can be a good field level test method for detection of counterfeit medicines, as it is very fast, easy to use and does not require technical expertise.

Artificial Intelligence and Machine Learning: A New Disruptive Force in Orthopaedics.

Orthopaedics as a surgical discipline requires a combination of good clinical acumen, good surgical skill, a reasonable physical strength and most of all, good understanding of technology. The last few decades have seen rapid adoption of new technologies into orthopaedic practice, power tools, new implants, CAD-CAM design, 3-D printing, additive manufacturing just to name a few. The new disruption in orthopaedics in the current time and era is undoubtedly the advent of artificial intelligence and robotics. As these technologies take root and innovative applications continue to be incorporated into the main-stream orthopedics, as we know it today, it is imperative to look at and understand the basics of artificial intelligence and what work is being done in the field today. This article takes the form of a loosely structured narrative review and will introduce the reader to key concepts in the field of artificial intelligence as well as some of the directions in application of the same in orthopaedics. Some of the recent work has been summarised and we present our viewpoint at the conclusion as to why we must consider artificial intelligence as a disrupting positive influence on orthopaedic surgery.

PI-Sole: A Low-Cost Solution for Gait Monitoring Using Off-The-Shelf Piezoelectric Sensors and IMU.

In this paper, viability of low-cost off-the-shelf Piezoelectric ceramic disc elements is explored for an insole-based gait monitoring system, `PI-Sole' (Piezo In-Sole). Piezoelectric elements can sense dynamic changes in pressure in a closed-loop environment with good sensitivity and a wide measurement range. In this paper, method to enable these elements to continuously sense plantar pressure while walking is proposed, making them a very cost-efficient alternative to the widely used Force Sensing Resistors (FSR) and pressure plates for monitoring human gait. However, piezoelectric elements show hysteresis in their force response, inducing a drift in calculated pressure which increases with time. A novel and effective method to perform detrending of the signal is also presented utilizing stride contexts from a 6-DoF Inertial Measurement Unit (IMU) and the same is utilized to perform zero-correction in the pressure data. 3-D trajectories of strides are calculated using the IMU, and parameters like stride length, stride height etc. are further derived. In order to test the validity of our proposed methods, important kinetic parameters like Vertical Ground Reaction Force (VGRF) and Center of Pressure (CoP) are calculated using PI-Sole and compared to the ones calculated using FSR's in multiple prior works. Applicability of PI-Sole is demonstrated further by depicting and analysing characteristic differences between a heel-strike toe-off stance type, and a flat-strike stance type, the latter being one of the primary symptoms in many cases of pathological gait, including Parkinsonian gait. Important artefacts from foot's height profile while walking are analysed for both stance types in context of standard gait events. We report a mean error of 2.8cm in stride length calculation, and a mean accuracy of 94.5% in calculating swing/stance duration of gait cycles.

Early Detection of Mild Cognitive Impairment using Pervasive Sensing.

Aging in place and independent living for the elderly has gained importance, and so has instrumented homes for ambient assisted living (AAL). In this paper we explore the feasibility of using passive sensors to provide insights into the cognitive and physical well-being of the subject. We derive a novel clustering based tactics to check heterogeneity in terms of movement behaviour among patients, and then provide our feasibility study on detection of mild cognitive impairment based on the results of the clustering.