Unlocking Insights in IoT-Based Patient Monitoring: Methods for Encompassing Large-Data Challenges

Hospital is not the home: Lessons from implementing remote technology to support acute inpatient and transitional care in the home in the US and UK

UNSTRUCTURED

The pandemic, patient preference and economic opportunity are shifting acute care from hospital to home, supported by the transformation in remote monitoring technology. Monitoring patients with digital medical devices gives unprecedented insight into their physiology. However, this technology does not exist in a vacuum. Distinguishing pathology from physiological variability, user error, or device limitations is challenging. In a hospital, patients are monitored in a contrived environment. Monitoring at home instead captures activities of daily living alongside patients' trajectory of disease and recovery. Both settings make for "noisy" data. However, we are familiar with hospital noise, accounting for it in our practice and perceptions of normal. Home monitoring as a diagnostic intervention introduces a new set of downstream consequences, dependent on device, cadence of collection, adherence, duration, alarm thresholds, and escalation criteria. We must accept greater ambiguity and contextualize vital signs. All devices balance accuracy with acceptability, compromises are inevitable and perfect data should not be expected. Alarms must be specific as well as sensitive, balancing clinical risk with capacity for response. By setting expectations around data from the home, we can smooth the adoption of remote monitoring and accelerate the transition of acute care.

Identification of a Person in a Trajectory Based on Wearable Sensor Data Analysis

Human trajectories can be tracked by the internal processing of a camera as an edge device. This work aims to match peoples' trajectories obtained from cameras to sensor data such as acceleration and angular velocity, obtained from wearable devices. Since human trajectory and sensor data differ in modality, the matching method is not straightforward. Furthermore, complete trajectory information is unavailable; it is difficult to determine which fragments belong to whom. To solve this problem, we newly proposed the SyncScore model to find the similarity between a unit period trajectory and the corresponding sensor data. We also propose a Likelihood Fusion algorithm that systematically updates the similarity data and integrates it over time while keeping other trajectories in mind. We confirmed that the proposed method can match human trajectories and sensor data with an accuracy, a sensitivity, and an F1 of 0.725. Our models achieved decent results on the UEA dataset.

Emerging Technologies in Edge Computing and Networking

The global evolution of the Internet is experiencing a notable and inevitable change towards a convergent scenario known as the Internet of Things (IoT), where a large number of devices with heterogeneous characteristics and requirements have to be interconnected to serve different verticals, such as smart cities, intelligent transportation systems, smart grids, (ITS) or e-health [...].