On the use of simulation in robotics: Opportunities, challenges, and suggestions for moving forward.

The last five years marked a surge in interest for and use of smart robots, which operate in dynamic and unstructured environments and might interact with humans. We posit that well-validated computer simulation can provide a virtual proving ground that in many cases is instrumental in understanding safely, faster, at lower costs, and more thoroughly how the robots of the future should be designed and controlled for safe operation and improved performance. Against this backdrop, we discuss how simulation can help in robotics, barriers that currently prevent its broad adoption, and potential steps that can eliminate some of these barriers. The points and recommendations made concern the following simulation-in-robotics aspects: simulation of the dynamics of the robot; simulation of the virtual world; simulation of the sensing of this virtual world; simulation of the interaction between the human and the robot; and, in less depth, simulation of the communication between robots. This Perspectives contribution summarizes the points of view that coalesced during a 2018 National Science Foundation/Department of Defense/National Institute for Standards and Technology workshop dedicated to the topic at hand. The meeting brought together participants from a range of organizations, disciplines, and application fields, with expertise at the intersection of robotics, machine learning, and physics-based simulation.

Closing the Digital Health Evidence Gap: Development of a Predictive Score to Maximize Patient Outcomes.

Background: Clinical studies of telemedicine (TM) programs for chronic illness have demonstrated mixed results across settings and populations. With recent uptake in use of digital health modalities, more precise patient classification may improve outcomes, efficiency, and effectiveness. Objective: The purpose of the research was to develop a predictive score that measures the influence of patient characteristics on TM interventions. The central hypothesis is that disease type, illness severity, and the social determinants of health influence outcomes, including resource utilization, and can be precisely characterized. Methods: The retrospective study evaluated the feasibility of creating a patient "Telemedicine ImPact" (TIP) score derived from a Virginia Medicare and Medicaid claims data set. Claims were randomly selected, stratified by disease type, and matched by illness severity into a TM intervention group (N = 7,782) and a nontelemedicine "usual care" control cohort (N = 7,981). The individual records were then summarized into 15,762 cases with 80% of the cases used to develop, train, and test four predictive models (hospital utilization, readmissions, total utilization, and mortality) using 10-fold cross-validation. Results: Bayesian supervised machine learning achieved reference model performance index area under the curve for receiver operating characteristic (AUC/ROC) ≥0.85. Posterior probabilities for each outcome model were generated on a "hold-back" set of 3,082 cases. Robust parametric statistical methods enabled dimension reduction, model validation, and derivation of a reliable composite scaled score that quantified the overall health risk for each case. The TM intervention cohort demonstrated higher total utilization (representing the sum of inpatient, outpatient, and prescription use) and lower mean inpatient utilization than the usual standard of care. This finding suggests TM-based care may shift the composition of health resource utilization, reducing hospitalizations while increasing outpatient services, adjusted for patient differences. Conclusions: The creation of a patient score using machine learning to predict the effect of TM on outcomes is feasible. Adoption of the TIP score may reduce variability in results by more precisely accounting for the effects of patient characteristics on health outcomes and utilization. More consistent outcome prediction may lead to greater support for digital health.

Remote Health Monitoring for Older Adults and Those with Heart Failure: Adherence and System Usability.

BACKGROUND: Remote health monitoring technology has been suggested as part of an early intervention and prevention care model. Older adults with a chronic health condition have been shown to benefit from remote monitoring but often have challenges with complex technology. The current study reports on the usability of and adherence with an integrated, real-time monitoring system over an extended period of time by older adults with and without a chronic health condition. MATERIALS AND METHODS: Older adults 55 years of age and over with and without heart failure participated in a study in which a telehealth system was used for 6 months each. The system consisted of a wireless wristwatch-based monitoring device that continuously collected temperature and motion data. Other health information was collected daily using a weight scale, blood pressure cuff, and tablet that participants used for health surveys. Data were automatically analyzed and summarized by the system and presented to study nurses. RESULTS: Forty-one older adults participated. Seventy-one percent of surveys, 75% of blood pressure readings, and 81% of daily weight measurements were taken. Participants wore the watch monitor 77% of the overall 24/7 time requested. The weight scale had the highest usability rating in both groups. The groups did not otherwise differ on device usage. CONCLUSIONS: The findings indicate that a health monitoring system designed for older adults can and will be used for an extended period of time and may help older adults with chronic conditions reside longer in their own homes in partnership with the healthcare system.

Differentiating Walking from other Activities of Daily Living in Older Adults Using Wrist-based Accelerometers.

There is a significant body of literature demonstrating that accelerometers placed at various locations on the body can provide the data necessary to recognize walking. Most of the literature, however, either does not consider accelerometers placed at the wrist, or suggests that the wrist is not the ideal location. The wrist, however, is probably the most socially-acceptable location for a monitoring device. This study evaluates the possibility of using wrist accelerometers to recognize walking in the elderly during everyday life to evaluate the amount of time spent walking and, moreover, potentially recognize changes in stability that might lead to falls. Thirty elderly individuals aged 65 years and older were asked to wear a wrist accelerometer for four hours each while simultaneously being video recorded as they went about their normal daily activities. Accelerometer data were then analyzed using both frequency- and time-domain analyses. Particular attention was given to methods capable of being calculated on the wrist device so that future work will not require streaming large amounts of data from the device to the central server. Frequency based analysis to characterize walking in the test set yielded results of 98% area under the receiver operating characteristic curve (AUC). Using a time-series algorithm limited to features calculable on the wrist device, moreover, achieved an AUC of 90%. A small, socially-acceptable, wrist-based device, therefore, can successfully be used to differentiate walking from other activities of daily living in older adults. These findings may enable improved gait monitoring and efforts in falls prevention.

Metrics for assessing the reliability of a telemedicine remote monitoring system.

OBJECTIVE: The goal of this study was to assess using new metrics the reliability of a real-time health monitoring system in homes of older adults. MATERIALS AND METHODS: The "MobileCare Monitor" system was installed into the homes of nine older adults >75 years of age for a 2-week period. The system consisted of a wireless wristwatch-based monitoring system containing sensors for location, temperature, and impacts and a "panic" button that was connected through a mesh network to third-party wireless devices (blood pressure cuff, pulse oximeter, weight scale, and a survey-administering device). To assess system reliability, daily phone calls instructed participants to conduct system tests and reminded them to fill out surveys and daily diaries. Phone reports and participant diary entries were checked against data received at a secure server. RESULTS: Reliability metrics assessed overall system reliability, data concurrence, study effectiveness, and system usability. Except for the pulse oximeter, system reliability metrics varied between 73% and 92%. Data concurrence for proximal and distal readings exceeded 88%. System usability following the pulse oximeter firmware update varied between 82% and 97%. An estimate of watch-wearing adherence within the home was quite high, about 80%, although given the inability to assess watch-wearing when a participant left the house, adherence likely exceeded the 10 h/day requested time. In total, 3,436 of 3,906 potential measurements were obtained, indicating a study effectiveness of 88%. CONCLUSIONS: The system was quite effective in providing accurate remote health data. The different system reliability measures identify important error sources in remote monitoring systems.

Using principal component analysis to aid bayesian network development for prediction of critical care patient outcomes.

BACKGROUND: Predicting an intensive care unit patient's outcome is highly desirable. An end goal is for computational techniques to provide updated, accurate predictions about changing patient condition using a manageable number of physiologic parameters. METHODS: Principal component analysis was used to select input parameters for critical care patient outcome models. Vital signs and laboratory values from each patient's hospital stay along with outcomes ("Discharged" vs. "Deceased") were collected retrospectively at a Level I Trauma-Military Medical Center in the southwest; intensive care unit patients were included if they had been admitted for burn, infection, or hypovolemia during a 5-year period ending October 2007. Principal component analysis was used to determine which of the 24 parameters would serve as inputs in a bayesian network developed for outcome prediction. RESULTS: Data for 581 patients were collected. Pulse pressure, heart rate, temperature, respiratory rate, sodium, and chloride were found to have statistically significant differences between Discharged and Deceased groups for "Hypovolemia" patients. For "Burn" patients, pulse pressure, hemoglobin, hematocrit, and potassium were found to have statistically significant differences. For a "Combined" group, heart rate, temperature, respiratory rate, sodium, and chloride had statistically significant differences. A bayesian network based on these results, developed for the Combined group, achieved an accuracy of 75% when predicting patient outcome. CONCLUSIONS: Outcome prediction for critical care patients is possible. Future work should explore model development using additional temporal data and should include prospective validation. Such technology could serve as the basis of real-time intelligent monitoring systems for critical patients.

Using Bayesian networks and rule-based trending to predict patient status in the intensive care unit.

Multivariate Bayesian models trained with machine learning, in conjunction with rule-based time-series statistical techniques, are explored for the purpose of improving patient monitoring. Three vital sign data streams and known outcomes for 36 intensive care unit (ICU) patients were captured retrospectively and used to train a set of Bayesian net models and to construct time-series models. Models were validated on a reserved dataset from 16 additional patients. Receiver operating characteristic (ROC) curves were calculated. Area under the curve (AUC) was 91% for predicting improving outcome. The model's AUC for predicting declining outcome increased from 70% to 85% when the model was indexed to personalized baselines for each patient. The rule-based trending and alerting system was accurate 100% of the time in alerting a subsequent decline in condition. These techniques promise to improve the monitoring of ICU patients with high-sensitivity alerts, fewer false alarms, and earlier intervention.