Driver Monitoring of Automated Vehicles by Classification of Driver Drowsiness Using a Deep Convolutional Neural Network Trained by Scalograms of ECG Signals

Driver drowsiness is one of the leading causes of traffic accidents. This paper proposes a new method for classifying driver drowsiness using deep convolution neural networks trained by wavelet scalogram images of electrocardiogram (ECG) signals. Three different classes were defined for drowsiness based on video observation of driving tests performed in a simulator for manual and automated modes. The Bayesian optimization method is employed to optimize the hyperparameters of the designed neural networks, such as the learning rate and the number of neurons in every layer. To assess the results of the deep network method, heart rate variability (HRV) data is derived from the ECG signals, some features are extracted from this data, and finally, random forest and k-nearest neighbors (KNN) classifiers are used as two traditional methods to classify the drowsiness levels. Results show that the trained deep network achieves balanced accuracies of about 77% and 79% in the manual and automated modes, respectively. However, the best obtained balanced accuracies using traditional methods are about 62% and 64%. We conclude that designed deep networks working with wavelet scalogram images of ECG signals significantly outperform KNN and random forest classifiers which are trained on HRV-based features.

Driver Monitoring of Automated Vehicles by Classification of Driver Drowsiness using a Deep Convolutional Neural Network Trained by Scalograms of ECG Signals

Driver drowsiness is one of the leading causes of traffic accidents. This paper proposes a new method for classifying driver drowsiness using deep convolution neural networks trained by wavelet scalogram images of electrocardiogram (ECG) signals. Three different classes were de-fined for drowsiness based on video observation of driving tests performed in a simulator for manual and automated modes. The Bayesian optimization method is employed to optimize the hyperparameters of the designed neural networks, such as the learning rate and the number of neurons in every layer. To assess the results of the deep network method, Heart Rate Variability (HRV) data is derived from the ECG signals, some features are extracted from this data, and finally, random forest and k-nearest neighbors (KNN) classifiers are used as two traditional methods to classify the drowsiness levels. Results show that the trained deep network achieves balanced accuracies of about 77% and 79% in the manual and automated modes, respectively. However, the best obtained balanced accuracies using traditional methods are about 62% and 64%. We conclude that designed deep networks working with wavelet scalogram images of ECG signals significantly outperform KNN and random forest classifiers which are trained on HRV-based features.

Adaptive Support Ventilation (ASV). Beneficial or not?

Ventilators functions and features have evolved with the advancement of technology along with the addition of microprocessors. It is important to understand and examine the benefits and risks associated with these advanced automated modes. Adaptive Support Ventilation (ASV) is a mode that is unique to the Hamilton Medical ventilators, thereby limiting the number of clinicians who have experience with using this mode. ASV can make changes to respiratory rate and tidal volume and adjusting the driving pressure in the absence of a professional. ASV changes ventilator strategies when it detects changes to a patient’s lung dynamics. The scope of ASV mode is not universally understood. Respiratory therapists may feel their position would be threatened with the use of smart automated modes. This paper will aim to review the literature on the ASV mode of ventilation. The literature review will address the following research questions to broaden the understanding of the risks and benefits of the ASV mode. 1) Is the ASV mode effective for weaning patients? 2) Is ASV a safe mode of ventilation for patients with COPD and ARDS? 3) Is ASV a safe mode of ventilation with changes in lung dynamics? 4) Does ASV impact the bedside respiratory therapist? Conclusions: ASV appears to be at least effective or even more superior to other modes especially during weaning off mechanical ventilation, and in other forms of respiratory failure. More studies in different clinical conditions and head-to-head with other modes. Keywords: ASV, COPD, ARDS, Weaning

Computer-Controlled Instrument System for Sequential Clinical Chemical Testing. II. Evaluation of Instrumental Performance

The instrumental system described [Clin. Chem.19, 1114 (1973)] is evaluated in manual, partially automated, and totally automated modes to measure system and individual component performance. Excellent accuracy and precision were observed, hence the system is judged suitable for most analytical applications. Results are: wavelength accuracy and reproducibility in automated mode: ±0.004 nm and ±0.1 nm, respectively; photometric accuracy and precision: ca. 0.8% relative and ±0.007 A, respectively, in both the manual and automated modes; and ratemeter accuracy and precision: ±0.4% relative and ±0.873 (SD) mV/min (or mA/ min), respectively, for standard synthetic ramps and ±1.2% relative and ±2.2 (SD) mV/min, respectively, under actual laboratory conditions for rates in the range of 1 to 200 mV/min. Automated experiments are made without human intervention after the samples are loaded.