Neural and Kernal Methods for Therapeutic Drug Monitoring

Recently, important advances in dosage formulations, therapeutic drug monitoring (TDM), and the emerging role of combined therapies have resulted in a substantial improvement in patients’ quality of life. Nevertheless, the increasing amounts of collected data and the non-linear nature of the underlying pharmacokinetic processes justify the development of mathematical models capable of predicting concentrations of a given administered drug and then adjusting the optimal dosage. Physical models of drug absorption and distribution and Bayesian forecasting have been used to predict blood concentrations, but their performance is not optimal and has given rise to the appearance of neural and kernel methods that could improve it. In this chapter, we present a complete review of neural and kernel models for TDM. All presented methods are theoretically motivated, and illustrative examples in real clinical problems are included.

A Recurrent Probabilistic Neural Network for EMG Pattern Recognition

In the field of pattern recognition, probabilistic neural networks (PNNs) have been proven as an important classifier. For pattern recognition of EMG signals, the characteristics usually used are: (1) amplitude, (2) frequency, and (3) space. However, significant temporal characteristic exists in the transient and non-stationary EMG signals, which cannot be considered by traditional PNNs. In this article, a recurrent PNN, called recurrent log-linearized Gaussian mixture network (R-LLGMN), is introduced for EMG pattern recognition, with the emphasis on utilizing temporal characteristics. The structure of R-LLGMN is based on the algorithm of a hidden Markov model (HMM), which is a routinely used technique for modeling stochastic time series. Since R-LLGMN inherits advantages from both HMM and neural computation, it is expected to have higher representation ability and show better performance when dealing with time series like EMG signals. Experimental results show that R-LLGMN can achieve high discriminant accuracy in EMG pattern recognition.

The Role of Neural Networks in Computerized Classification of the Electrocardiogram

Electrical forces generated by the heart are transmitted to the skin through the body’s tissues. These forces can be recorded on the body’s surface and are represented as an electrocardiogram (ECG). The ECG can be used to detect many cardiac abnormalities. Traditionally, ECG classification algorithms have used rule based techniques in an effort to model the thought and reasoning process of the human expert. However, the definition of an ultimate rule set for cardiac diagnosis has remained somewhat elusive, and much research effort has been directed at data driven techniques. Neural networks have emerged as a strong contender as the highly non-linear and chaotic nature of the ECG represents a well-suited application for this technique. This study presents an overview of the application of neural networks in the field of ECG classification, and, in addition, some preliminary results of adaptations of conventional neural classifiers are presented. From this work, it is possible to highlight issues that will affect the acceptance of this technique and, in addition, identify challenges faced for the future. The challenges can be found in the intelligent processing of larger amounts of ECG information which may be generated from recording techniques such as body surface potential mapping.

A Survey on Various Applications of Artificial Neural Networks in Selected Fields of Healthcare

This chapter gives an overview of artificial neural networks as instruments for processing miscellaneous biomedical signals. A variety of applications are illustrated in several areas of healthcare. The structure of this chapter is rather oriented on medical fields like cardiology, gynecology, or neuromuscular control than on types of neural nets. Many examples demonstrate how neural nets can support the diagnosis and prediction of diseases. However, their content does not claim completeness due to the enormous amount and exponentially increasing number of publications in this field. Besides the potential benefits for healthcare, some remarks on underlying assumptions are also included as well as problems which may occur while applying artificial neural nets. It is hoped that this review gives profound insight into strengths as well as weaknesses of artificial neural networks as tools for processing biomedical signals.

Computational Fluid Dynamics and Neural Network for Modeling and Simulations of Medical Devices

This chapter describes the utilization of computational fluid dynamics (CFD) with neural network (NN) for analysis of medical devices. First, the concept of mathematical modeling and its use for solving engineering problems is presented followed by an introduction to CFD with a brief summary of the numerical techniques currently available. A brief introduction to the standard optimization strategies for NN and the various methodologies in use are also presented. A case study of the design and optimization of scaffolds for tissue engineering heart valve using the combined CFD and NN approach is presented and discussed. This chapter concludes with a discussion of the advantages and disadvantages of the combined NN and CFD techniques and their future potential prospective.

The Use of Artificial Neural Networks for Objective Determination of Hearing Threshold Using the Auditory Brainstem Response

We have analyzed high and low level auditory brainstem response data (550 waveforms over a large age range; 126 were repeated sessions used in correlation analysis), by extracting time, frequency, and phase features and using these as inputs to ANN and decision tree classifiers. A two stage process is used. For responses with a high poststimulus to prestimulus power ratio indicative of high level responses, a classification accuracy of98% has been achieved. These responses are easily classified by the human expert. For lower level responses appropriate to hearing threshold, additional features from time, frequency, and phase have been used for classification, providing accuracies between 65% and 82%. These used a dataset with repeated recordings so that correlation could be employed. To increase classification accuracy, it may be necessary to combine the relevant features in a hybrid model.

Movement Pattern Recognition Using Neural Networks

This chapter provides an overview of artificial neural network applications for the detection and classification of various gaits based on their typical characteristics. Gait analysis is routinely used for detecting abnormality in the lower limbs and also for evaluating the progress of various treatments. Neural networks have been shown to perform better compared to statistical techniques in some gait classification tasks. Various studies undertaken in this area are discussed with a particular focus on neural network’s potential in gait diagnostics. Examples are presented to demonstrate the suitability of neural networks for automated recognition of gait changes due to aging from their respective gait patterns and their potential for identification of at-risk or non-functional gait.

Artificial Neural Networks in EEG Analysis

The artificial neural networks (ANNs) are regularly employed in EEG signal processing because of their effectiveness as pattern classifiers. In this chapter, four specific applications will be studied: On a day to day basis, ANNs can assist in identifying abnormal EEG activity in patients with neurological diseases such as epilepsy, Huntington’s disease, and Alzheimer’s disease. The ANNs can reduce the time taken for interpretation of physiological signals such as EEG, respiration, and ECG recorded during sleep. During an invasive surgical procedure, the ANNs can provide objective parameters derived from the EEG to help determine the depth of anesthesia. The ANNs have made significant contributions toward extracting embedded signals within the EEG which can be used to control external devices. This rapidly developing field, which is called brain-computer interface, has a large number of applications in empowering handicapped individuals to independently operate appliances, neuroprosthesis, or orthosis.

Myoelectric Teleoperation of a Dual-Arm Manipulator Using Neural Networks

In this chapter, an advanced intelligent dual-arm manipulator system teleoperated by EMG signals and hand positions is described. This myoelectric teleoperation system employs a probabilistic neural network, so called log-linearized Gaussian mixture network (LLGMN), to gauge the operator’s intended hand motion from EMG patterns measured during tasks. In addition, an event-driven task model using Petri net and a non-contact impedance control method are introduced to allow a human operator to maneuver a couple of robotic manipulators intuitively. A set of experimental results demonstrates the effectiveness of the developed prototype system.

Overview of Artificial Neural Networks and their Applications in Healthcare

Artificial neural network (ANN) is one of the main constituents of the artificial intelligence techniques. Like in many other areas, ANN has made a significant mark in the domain of healthcare applications. In this chapter, we provide an overview of the basics of neural networks, their operation, major architectures that are widely employed for modeling the input-to-output relations, and the commonly used learning algorithms for training the neural network models. Subsequently, we briefly outline some of the major application areas of neural networks for the improvement and well being of human health.