Development of a Hand Motion-based Assessment System for Endotracheal Intubation Training.

Endotracheal intubation (ETI) is a procedure to manage and secure an unconscious patient's airway. It is one of the most critical skills in emergency or intensive care. Regular training and practice are required for medical providers to maintain proficiency. Currently, ETI training is assessed by human supervisors who may make inconsistent assessments. This study aims at developing an automated assessment system that analyzes ETI skills and classifies a trainee into an experienced or a novice immediately after training. To make the system more available and affordable, we investigate the feasibility of utilizing only hand motion features as determining factors of ETI proficiency. To this end, we extract 18 features from hand motion in time and frequency domains, and also 12 force features for comparison. Subsequently, feature selection algorithms are applied to identify an ideal feature set for developing classification models. Experimental results show that an artificial neural network (ANN) classifier with five hand motion features selected by a correlation-based algorithm achieves the highest accuracy of 91.17% while an ANN with five force features has only 80.06%. This study corroborates that a simple assessment system based on a small number of hand motion features can be effective in assisting ETI training.

Study of blood glucose and insulin infusion rate in real-time in diabetic rats using an artificial pancreas system.

Artificial pancreas system (APS) is an emerging new treatment for type 1 diabetes mellitus. The aim of this study was to develop a rat APS as a research tool and demonstrate its application. We established a rat APS using Medtronic Minimed Pump 722, Medtronic Enlite sensor, and the open artificial pancreas system as a controller. We tested different dilutions of Humalog (100 units/ml) in saline ranged from 1:3 to 1:20 and determined that 1:7 dilution works well for rats with ~500g bodyweight. Blood glucose levels (BGL) of diabetic rats fed with chow diet (58% carbohydrate) whose BGL was managed by the closed-loop APS for the total duration of 207h were in euglycemic range (70-180 mg/dl) for 94.5% of the time with 2.1% and 3.4% for hyperglycemia (>180mg/dl) and hypoglycemia (<70 mg/dl), respectively. Diabetic rats fed with Sucrose pellets (94.8% carbohydrate) for the experimental duration of 175h were in euglycemic range for 61% of the time with 35% and 4% for hyperglycemia and hypoglycemia, respectively. Heathy rats fed with chow diet showed almost a straight line of BGL ~ 95 mg/dl (average 94.8 mg/dl) during the entire experimental period (281h), which was minimally altered by food intake. In the healthy rats, feeding sucrose pellets caused greater range of BGL in high and low levels but still within euglycemic range (99.9%). Next, to study how healthy and diabetic rats handle supra-physiological concentrations of glucose, we intraperitoneally injected various amounts of 50% dextrose (2, 3, 4g/kg) and monitored BGL. Duration of hyperglycemia after injection of 50% dextrose at all three different concentrations was significantly greater for healthy rats than diabetic rats, suggesting that insulin infusion by APS was superior in reducing BGL as compared to natural insulin released from pancreatic ß-cells. Ex vivo studies showed that islets isolated from diabetic rats were almost completely devoid of pancreatic ß-cells but with intact α-cells as expected. Lipid droplet deposition in the liver of diabetic rats was significantly lower with higher levels of triacylglyceride in the blood as compared to those of healthy rats, suggesting lipid metabolism was altered in diabetic rats. However, glycogen storage in the liver determined by Periodic acid-Schiff staining was not altered in diabetic rats as compared to healthy rats. A rat APS may be used as a powerful tool not only to study alterations of glucose and insulin homeostasis in real-time caused by diet, exercise, hormones, or antidiabetic agents, but also to test mathematical and engineering models of blood glucose prediction or new algorithms for closed-loop APS.

Preliminary Experience With Inertial Movement Technology to Characterize Endotracheal Intubation Kinematics.

BACKGROUND: Endotracheal intubation (ETI) is an important emergency intervention. Only limited data describe ETI skill acquisition and often use bulky technology, not easily transitioned to the clinical setting. In this study, we used small, portable inertial detection technology to characterize intubation kinematic differences between experienced and novice intubators. METHODS: We performed a prospective study including novice (<10 prior clinical ETI) and experienced (>100 clinical ETI) emergency providers. We tracked upper extremity motion with roll, pitch, and yaw using inertial measurement units (IMU) placed on the bilateral hands and wrists of the intubator. Subject performed 6 simulated emergency intubations on a mannequin. Using machine learning algorithms, we determined the motions that best discriminated experienced and novice providers. RESULTS: We included data on 12 novice and 5 experienced providers. Four machine learning algorithms (artificial neural network, support vector machine, decision tree, and K-nearest neighbor search) were applied. Artificial neural network had the greatest accuracy (95% confidence interval) for discriminating between novice and experienced providers (91.17%, 90.8%-91.5%) and was the most parsimonious of the tested algorithms. Using artificial neural network, information from 5 movement features (right hand, roll amplitude; right hand, pitch amplitude; right hand, yaw standard deviation; left hand, yaw standard deviation; left hand, pitch frequency of peak amplitude) was able discriminated experienced from novice providers. CONCLUSIONS: Novice and experienced providers have different ETI movement patterns and can be distinguished by 5 specific movements. Inertial detection technology can be used to characterize the kinematics of emergency airway management.

Tracking changes of the parameters of glucose-insulin homeostasis during the course of obesity in B6D2F1 mice.

Obesity is one of the primary causes of type 2 diabetes mellitus (T2DM). To better understand how obesity impairs glucose-insulin homeostasis, we tracked fasting blood glucose and insulin levels and the key components of glucose-insulin homeostasis for 7 months in high fat diet (HFD; 45% fat) fed mice (n = 8). Every 2 weeks we measured body weight, fasting blood glucose and insulin levels, and estimated 5 key rate constants of glucose-insulin homeostasis using the methods established previously (Heliyon 3: e00310, 2017). Mice gained weight steadily, more than doubling their weights after 7 months (23.6 ± 0.5 to 52.3 ± 1.4 g). Fasting (basal) insulin levels were elevated (221.3 ± 16.7 to 1043.1 ± 90.5 pmol l-1) but fasting blood glucose levels unexpectedly returned to the baseline levels (152.8 ± 7.0 to 152.0 ± 7.2 mg/dl) despite significantly elevated levels (216.8 ± 44.9 mg/dl, average of 3 highest values for 8 mice) during the experimental period. After 7 months of HFD feeding, the rate constants for insulin secretion (k1), insulin-independent glucose uptake (k3), and insulin concentration where liver switches from glucose uptake to release (Ipi) were significantly elevated. Insulin-dependent glucose uptake (k2) and rate constant of liver glucose transfer (k4) were lowered but no statistical significance was reached. The novel and key finding of this study is the wide range of fluctuations of the rate constants during the course of obesity, reflecting the body's compensatory responses against metabolic alterations caused by obesity.

Neural network-based model predictive control for type 1 diabetic rats on artificial pancreas system.

Artificial pancreas system (APS) is a viable option to treat diabetic patients. Researchers, however, have not conclusively determined the best control method for APS. Due to intra-/inter-variability of insulin absorption and action, an individualized algorithm is required to control blood glucose level (BGL) for each patient. To this end, we developed model predictive control (MPC) based on artificial neural networks (ANNs), which combines ANN for BGL prediction based on inputs and MPC for BGL control based on the ANN (NN-MPC). First, we developed a mathematical model for diabetic rats, which was used to identify individual virtual subjects by fitting to empirical data collected through an APS, including BGL data, insulin injection, and food intake. Then, the virtual subjects were used to generate datasets for training ANNs. The NN-MPC determines control actions (insulin injection) based on BGL predicted by the ANN. To evaluate the NN-MPC, we conducted experiments using four virtual subjects under three different scenarios. Overall, the NN-MPC maintained BGL within the normal range about 90% of the time with a mean absolute deviation of 4.7 mg/dl from a desired BGL. Our findings suggest that the NN-MPC can provide subject-specific BGL control in conjunction with a closed-loop APS. Graphical abstract ᅟ.

Estimation of insulin secretion, glucose uptake by tissues, and liver handling of glucose using a mathematical model of glucose-insulin homeostasis in lean and obese mice.

Destruction of the insulin-producing ß-cells is the key determinant of diabetes mellitus regardless of their types. Due to their anatomical location within the islets of Langerhans scattered throughout the pancreas, it is difficult to monitor ß-cell function and mass clinically. To this end, we propose to use a mathematical model of glucose-insulin homeostasis to estimate insulin secretion, glucose uptake by tissues, and hepatic handling of glucose. We applied the mathematical model by Lombarte et al. (2013) to compare various rate constants representing glucose-insulin homeostasis between lean (11% fat)- and high fat diet (HFD; 45% fat)-fed mice. Mice fed HFD (n = 12) for 3 months showed significantly higher body weights (49.97 ± 0.52 g vs. 29.86 ± 0.46 g), fasting blood glucose levels (213.08 ± 10.35 mg/dl vs. 121.91 ± 2.26 mg/dl), and glucose intolerance compared to mice fed lean diet (n = 12). Mice were injected with 1 g/kg glucose intraperitoneally and blood glucose levels were measured at various intervals for 120 min. We performed simulation using Arena™ software based on the mathematical model and estimated the rate constants (9 parameters) for various terms in the differential equations using OptQuest™. The simulated data fit accurately to the observed data for both lean and obese mice, validating the use of the mathematical model in mice at different stages of diabetes progression. Among 9 parameters, 5 parameters including basal insulin, k2 (rate constant for insulin-dependent glucose uptake to tissues), k3 (rate constant for insulin-independent glucose uptake to tissues), k4 (rate constant for liver glucose transfer), and Ipi (rate constant for insulin concentration where liver switches from glucose release to uptake) were significantly different between lean- and HFD-fed mice. Basal blood insulin levels, k3, and Ipi were significantly elevated but k2 and k4 were reduced in mice fed a HFD compared to those fed a lean diet. Non-invasive assessment of the key components of glucose-insulin homeostasis including insulin secretion, glucose uptake by tissues, and hepatic handling of glucose may be helpful for individualized drug therapy and designing a customized control algorithm for the artificial pancreas.