Using machine learning to improve bovine tuberculosis control in herd level outbreaks.

Bovine tuberculosis (bTB), a chronic disease of cattle, is caused by the Mycobacterium bovis infection. Despite having a serious social and economic impact in the United Kingdom and Ireland, there is no antemortem gold standard diagnostic test. Tuberculin skin tests (CICT) are commonly used as a control measure with the interferon gamma (IFN-γ) assay being applied in certain circumstances. This paper utilizes data gathered describing tuberculin regression in reactors (test positive cattle) following the CICT at 72 ± 4 h post injection in herds with large bTB outbreaks. The work then applies machine learning techniques (Decision Trees, Bagging Trees and Random Forests, alongside several balancing approaches) to predict which cattle were likely to be truly infected with tuberculosis, enabling identification of atypical breakdowns that require extra investigation and providing a mechanism for quality assurance of the existing CICT bTB surveillance scheme. The analysis showed that Random Forests (RF) trained using SMOTE balancing had the joint best performance and accuracy (0.90). The importance of the two components of the interferon gamma assay within the RF model also indicated that varying the assay threshold for large outbreaks would be beneficial. Furthermore, the combined use of the RF and IFN- γ models could lead to the improved detection of infection within breakdown herds, reducing the scale and duration of outbreaks. An additional use of these models would be for quality assuring the current bTB surveillance based on CICT and post mortem inspection. Quality control is well recognized as an essential component of a disease surveillance/eradication programme.Clinical Relevance- Bovine tuberculosis remains a disease that is hard to control on a national level. The use of the machine learning model could lead to significant improved detection of infection within breakdown herds, reducing the scale and duration of outbreaks. Advanced modelling, such as this, has the potential to strengthen the efficacy of disease surveillance and the eradication strategy and can meaningfully contribute to animal disease national control plans.

Evaluation of a novel real-time adaptive assist-as-needed controller for robot-assisted upper extremity rehabilitation following stroke.

Rehabilitation therapy plays an essential role in assisting people with stroke regain arm function. Upper extremity robot therapy offers a number of advantages over manual therapies, but can suffer from slacking behavior, where the user lets the robot guide their movements even when they are capable of doing so by themselves, representing a major barrier to reaching the full potential of robot-assist rehabilitation. This is a pilot clinical study that investigates the use of an electromyography-based adaptive assist-as-needed controller to avoid slacking behavior during robotic rehabilitation for people with stroke. The study involved a convenience sample of five individuals with chronic stroke who underwent a robot therapy program utilizing horizontal arm tasks. The Fugl-Meyer assessment (FM) was used to document motor impairment status at baseline. Velocity, time, and position were quantified as performance parameters during the training. Arm and shoulder surface electromyography (EMG) and electroencephalography (EEG) were used to assess the controller's performance. The cross-sectional results showed strong second-order relationships between FM score and outcome measures, where performance metrics (path length and accuracy) were sensitive to change in participants with lower functional status. In comparison, speed, EMG and EEG metrics were more sensitive to change in participants with higher functional status. EEG signal amplitude increased when the robot suggested that the robot was inducing a challenge during the training tasks. This study highlights the importance of multi-sensor integration to monitor and improve upper-extremity robotic therapy.

Mixed Reality for Veterinary Medicine: Case Study of a Canine Femoral Nerve Block.

The femoral nerve blockage is a procedure that aims to provide anesthesia to the hip, anterior thigh, and stifle. This procedure presents several challenges when performed in veterinary patients with diverse anatomy and physiology. Successful use of this technique will improve a dog's recovery time after surgery in comparison to the commonly used epidural block. A mixed reality application to guide practitioners in the femoral nerve block procedure was developed in Unity and Visual Studio. A 3D model for use within the application was created from pictures of a cadaver leg using photogrammetry software. The Microsoft HoloLens headset provides the mixed reality hardware platform. This paper presents the workflow used in developing the mixed reality application and custom 3D model, as well as initial results with respect to the utility of the application in guiding an anesthesiologist in the procedure of the femoral nerve block.

Health Informatics for Neonatal Intensive Care Units: An Analytical Modeling Perspective.

The effective use of data within intensive care units (ICUs) has great potential to create new cloud-based health analytics solutions for disease prevention or earlier condition onset detection. The Artemis project aims to achieve the above goals in the area of neonatal ICUs (NICU). In this paper, we proposed an analytical model for the Artemis cloud project which will be deployed at McMaster Children's Hospital in Hamilton. We collect not only physiological data but also the infusion pumps data that are attached to NICU beds. Using the proposed analytical model, we predict the amount of storage, memory, and computation power required for the system. Capacity planning and tradeoff analysis would be more accurate and systematic by applying the proposed analytical model in this paper. Numerical results are obtained using real inputs acquired from McMaster Children's Hospital and a pilot deployment of the system at The Hospital for Sick Children (SickKids) in Toronto.

Heart rate variability as an indicator for morphine pharmacokinetics and pharmacodynamics in critically ill newborn infants.

Morphine is the commonest drug used for analgesia in newborn infants. It is a natural opioid that acts as an agonist at the mu and kappa receptors, which are receptors for analgesia and sedation. Morphine pharmacokinetics and pharmacodynamics (PKPD) for the newborn infant population are not well understood. The objective of this study is to use morphine PKPD parameters to estimate morphine plasma concentrations to be correlated with heart rate variability in the neonatal population.

Integration of drug dosing data with physiological data streams using a cloud computing paradigm.

Many drugs are used during the provision of intensive care for the preterm newborn infant. Recommendations for drug dosing in newborns depend upon data from population based pharmacokinetic research. There is a need to be able to modify drug dosing in response to the preterm infant's response to the standard dosing recommendations. The real-time integration of physiological data with drug dosing data would facilitate individualised drug dosing for these immature infants. This paper proposes the use of a novel computational framework that employs real-time, temporal data analysis for this task. Deployment of the framework within the cloud computing paradigm will enable widespread distribution of individualized drug dosing for newborn infants.

Dogs mean arterial pressure and heart rate responses during high propofol plasma concentrations estimated by a pharmacokinetic model.

Propofol total intravenous anesthesia should provide stability of the cardiovascular system. In this study, mean arterial pressure and heart rate were evaluated in eight healthy dogs anesthetized with increasing rates of propofol. The cerebral state index (CSI) was studied as an additional parameter. Although the estimated propofol plasma concentration reached a maximal value of 15.3 µg ml(-1), no hypotension or bradycardia were observed. Exploration of each animal's data revealed high inter-individual variability regarding mean arterial pressure and heart rate. Considering the logarithmic of the concentration, a moderate depressant effect of propofol on mean arterial pressure was revealed in five dogs but the effect was not followed on heart rate.

Anaesthesia synchronization software: target controlled infusion system evaluation.

Target Controlled Infusion (TCI) systems are based in drug Pharmacokinetic (Pk) and Pharmacodynamic (Pd) models implemented in an algorithm to drive an infusion device. Several studies had compare manual titration of anesthesia and TCI system use; some studies evaluate the performance of the control algorithms for TCI systems, and a considerable number of studies assess the performance of Pk/Pd models implemented into TCI systems. This study presents a set of tests to validate the performance of a TCI system as a computer-aided. The goal of the current study was to assess the performance of the TCI system, Anaesthesia Synchronization Software (ASYS), on clinical set up to evaluate communication consistence (computer - infusion device) and controller performance in real time. These measures provided quantitative and qualitative evidences of software robustness and accuracy to be used at clinical environment.

Target Controlled Infusion algorithms for anesthesia: theory vs practical implementation.

Target Controlled Infusion (TCI) systems are based in drug pharmacokinetic and pharmacodynamic models implemented in an algorithm to drive an infusion pump. Infusion control algorithms have been designed, implemented and validated for several anesthetic drugs, devices and controllers. The maintenance phase in these algorithms is represented by an equation that compensates the loss of drug from the central compartment and maintains the set target concentration. The goal of the current study was to improve existing TCI software with a new method for the maintenance phase. We compared and analyzed two different methods to find the more efficient method for the maintenance phase in an open-loop control TCI system.

Brain monitoring in dogs using the cerebral state index during the induction of anaesthesia via target-controlled infusion of propofol.

The aim of this study was to evaluate the correlation between the cerebral state index (CSI) and the estimated propofol plasma concentrations in dogs during induction of anaesthesia. Fifteen healthy dogs undergoing scheduled routine surgical procedures were enrolled in this study. Target controlled infusion (TCI) software, based on the pharmacokinetic model for propofol, was used to control the syringe pump and to estimate plasma propofol concentrations (PropCp) and the CSI values every five-seconds. Three electrodes placed in the centre of the forehead, on the left side of the forehead and on the left mastoid were used to collect the electroencephalographic (EEG) signal converted by the cerebral state monitor into the CSI. The cerebral electrical changes induced by increasing propofol concentrations appear to be detected by CSI monitoring in dogs. The negative correlation between CSI and PropCp demonstrates that the CSI could be used to assess electrical brain activity in dogs during the induction of anaesthesia with propofol.

Automation in anesthesia: computer controlled propofol infusion and data acquisition.

This work presents the improvement of a software for a clinical setup, the Anesthesia Synchronization Software (ASYS). The first version presented the data acquisition from cerebral monitors and a partial control with Target Controlled Infusion (TCI) system. Based on pharmacokinetic models, the effect-site and plasma concentrations can be related with the drug dose infused and vice versa. This later version comes with the implementation of the hemodynamic monitor always present in operating rooms and with the full TCI system developed to determine the infusion rates of the drug which are given as commands to the infusion pumps.

Nonlinear modeling of cerebral state index in dogs.

The Cerebral State Index (CSI) is an electroencephalogram derived signal representing the depression of central nervous activity produced by anesthetic drugs. In this study, a nonlinear model was developed to describe the CSI tendency during general anesthesia in dogs, by evaluating the effect of the anesthetic drug propofol. The model was based on a compartmental and Hill Equation structure with individually identified parameters. The clinical data of 14 dog surgeries were collected and used for modeling and testing. The model presented good results, following the CSI trend. A model for drug-effect for veterinarian anesthesia is an important step when developing advisory, educational and control systems. The overall aim is to improve animal safety and comfort.

Synchronization software for automation in anesthesia.

This work presents the development of a software for data acquisition and control (ASYS) on a clinical setup. Similar to the industrial Supervisory Control And Data Acquisition (SCADA) the software assembles a Target Controlled Infusion (TCI) monitoring and supervisory control data in real time from devices in a surgical room. The software is not a full controller since the TCI systems comprehend permanent interaction from the anesthesiologist. Based on pharmacokinetic models, the effect-site and plasma concentrations can be related with the drug dose infused and vice versa. The software determines the infusion rates of the drug which are given as commands to the infusion pumps. This software provides the anesthesiologist with a trustworthy tool for managing a safe and balanced anesthesia. Since it also incorporates the acquisition and display of patients brain signals.