A wrapper framework for feature selection and ELM weights optimization for FMG-based sign recognition.

Force myography (FMG) is increasingly gaining importance in gesture recognition because of it's ability to achieve high classification accuracy without having a direct contact with the skin. In this study, we investigate the performance of a bracelet with only six commercial force sensitive resistors (FSR) sensors for classifying many hand gestures representing all letters and numbers from 0 to 10 in the American sign language. For this, we introduce an optimized feature selection in combination with the Extreme Learning Machine (ELM) as a classifier by investigating three swarm intelligence algorithms, which are the binary grey wolf optimizer (BGWO), binary grasshopper optimizer (BGOA), and binary hybrid grey wolf particle swarm optimizer (BGWOPSO), which is used as an optimization method for ELM for the first time in this study. The findings reveal that the BGWOPSO, in which PSO supports the GWO optimizer by controlling its exploration and exploitation using inertia constant to improve the convergence speed to reach the best global optima, outperformed the other investigated algorithms. In addition, the results show that optimizing ELM with BGWOPSO for feature selection can efficiently improve the performance of ELM to enhance the classification accuracy from 32% to 69.84% for classifying 37 gestures collected from multiple volunteers and using only a band with 6 FSR sensors.

A PI-fuzzy logic controller for the regulation of blood glucose level in diabetic patients.

This manuscript investigates different fuzzy logic controllers for the regulation of blood glucose level in diabetic patients. While fuzzy logic control is still intuitive and at a very early stage, it has already been implemented in many industrial plants and reported results are very promising. A fuzzy logic control (FLC) scheme was recently proposed for maintaining blood glucose level in diabetics within acceptable limits, and was shown to be more effective with better transient characteristics than conventional techniques. In fact, FLC is based on human expertise and on desired output characteristics, and hence does not require precise mathematical models. This observation makes fuzzy rule-based technique very suitable for biomedical systems where models are, in general, either very complicated or over-simplistic. Another attractive feature of fuzzy techniques is their insensitivity to system parameter variations, as numerical values of physiological parameters are often not precise and usually vary from patient to another. PI and PID controllers are very popular and are efficiently used in many industrial plants. Fuzzy PI and PID controllers behave in a similar fashion to those classical controllers with the obvious advantage that the controller parameters are time dependant on the range of the control variables and consequently, result in a better performance. In this manuscript, a fuzzy PI controller is designed using a simplified design scheme and then subjected to simulations of the two common diabetes disturbances--sudden glucose meal and system parameter variations. The performance of the proposed fuzzy PI controller is compared to that of the conventional PID and optimal techniques and is shown to be superior. Moreover, the proposed fuzzy PI controller is shown to be more effective than the previously proposed FLC, especially with respect to the overshoot and settling time.

A fuzzy logic based closed-loop control system for blood glucose level regulation in diabetics.

In this study, a closed-loop system to control the plasma glucose level in patients with diabetes mellitus type 1 is proposed. This control scheme is based on fuzzy logic control theory to maintain a normoglycaemic average of 4.5 mmol 1(-1) and the normal conditions for free plasma insulin concentration in severe initial state; in particular, when the diabetic patient is subjected to a glucose meal disturbance or fluctuations in the measured glucose level due to error in the measuring instrument. The proposed controller has demonstrated superiority over other conventional controlling therapies. While fuzzy logic controllers have shown promising results in many fields, a comparative study is presented with well-known conventional controllers such as Proportional-Integral-Derivative (PID) and continuous insulin infusion control strategies. The simulated results, for the proposed controller, are presented and discussed.

A semiclosed-loop optimal control system for blood glucose level in diabetics.

In this manuscript, the well-known Bergman nonlinear mathematical model of the plasma glucose/insulin interaction is adopted and a semi closed-loop optimal technique is proposed for the correction of hyperglycemia in diabetic subjects. Computer simulations are used to evaluate the effectiveness of the proposed technique and to demonstrate its superiority in controlling hyperglycemia over other existing algorithms.

Statistical modelling of human blood carbon dioxide partial pressure.

This paper deals with statistical modelling of blood carbon dioxide partial pressure pCO2. The clinical measurements were carried out in the Arab Centre for Heart and Special Surgery in Amman. The statistical analysis of the results obtained demonstrates that pCO2 for arterial, venous and capillary blood have histograms approaching normal ones. Moreover, the experimental data suggest that the blood pCO2 can be expressed by a linear regression model. Making use of this regression model, the blood pCO2 population can indirectly be obtained with accuracy fulfilling clinical requirements. Therefore, this approach will not only lead to less invasive methods but will also result in decreasing the cost of projected analyser for blood gas analysis. The implementation of this proposed regression equation ensures the effectiveness of the model as an indirect method for pCO2 measurements.

The concentric-ring array for ultrasound hyperthermia: combined mechanical and electrical scanning.

While two-dimensional phased arrays can be electronically focused and steered in three dimensions without physically moving the applicator, they generally require a relatively large number of small transducer elements and, consequently, complex drive electronics. A configuration that does not require a large number of elements is that of a concentric-ring array. The field conjugation method can be used to produce a focal spot (or multiple spots) along the array axis. The resulting focal regions are very small and need to be steered transversely to heat tumours of typical size. However, steering the focused beam away from the array axis results in annular heating patterns which are often associated with undesired secondary foci (hot spots). In this paper, a method based on combining electrical and mechanical scanning using a concentric-ring applicator is presented. Advantages of the new method over the mechanically scanned fixed-focus transducers, currently in use, are pointed out. Computer simulations are conducted to investigate the possibility of heating different size tumours by appropriately combining the two scanning techniques. The bioheat transfer equation is solved numerically and temperature distributions associated with relevant heating patterns are presented and discussed. The simulations demonstrate the possibility of the combined technique to produce useful heating patterns which cannot be produced by either technique separately.

NxN square-element ultrasound phased-array applicator: simulated temperature distributions associated with directly synthesized heating patterns.

Computer simulations to demonstrate the possibility of heating small tumors by appropriately driving the electronic control circuitry of an nxn square-element ultrasonic phased array are conducted. The synthesis method consists of simultaneously focusing the ultrasonic beam at different points uniformly distributed along the tumor periphery and, hence, involves no scanning. It is demonstrated that by combining the multiple focusing feature with a new field phasing concept, typical undesired hot spots can be eliminated. The tissue thermal response to this heating modality is investigated by iteratively solving the three-dimensional steady-state bioheat equation. Temperature distributions associated with different directly synthesized power deposition patterns are simulated and discussed.

A field conjugation method for direct synthesis of hyperthermia phases-array heating patterns.

A phased-array field conjugation method is investigated as a means for synthesizing directly many ultrasound field patterns useful for tumor heating. For virtually any ultrasound phased array, the method permits the computation of element driving amplitude and phase distributions appropriate for synthesizing directly diffuse heating patterns without the need for mechanical or electrical scanning. Moreover, the proposed method offers the possibility of creating simultaneously, at different sites, more than one focus, which can then be scanned electronically. This attractive feature eliminates the need for operating at high spatial-peak temporal-peak focal intensities, a potential problem associated with conventional scanning, while achieving a desired heating pattern. The method is applied to two different applicator configurations: a concentric-ring and a square nxn array. Computer simulations of different heating patterns, synthesized using the field conjugation method, are presented. Important practical design parameters, such as the size and number of the array elements, are discussed.

The concentric-ring phased-array hyperthermia applicator: problems associated with directly synthesized annular heating patterns.

The authors investigate the synthesis of different size annular patterns and examine the relative intensities produced in the ultrasonic field proximal and distal to the focal plane. Computer simulations using a concentric-ring array are presented and discussed. The size and position of the synthesized annular patterns are varied by adjusting the phase and amplitude of the driving signal to each array element. It is demonstrated that annular patterns need to be larger than a limiting size (R(s)) to avoid excessive energy accumulation along the array axis. Analysis and computer simulations demonstrate that smaller size annual patterns (r/=R(s)) show that while therapeutic temperatures are reached at the periphery, the tumor center temperature cannot be evenly raised.

A cylindrical-section ultrasound phased-array applicator for hyperthermia cancer therapy.

A phased-array applicator geometry for deep localized hyperthermia is presented. The array consists of rectangular transducer elements forming a section of a cylinder that conforms to the body portals in the abdominal and pelvic regions. Focusing and scanning properties of the cylindrical-section array are investigated in homogeneous lossy media using appropriate computer simulations. The characteristic focus of this array is shown to be spatially limited in both transverse and longitudinal directions with intensity gain values suitable for deep hyperthermia applications. The ability of the cylindrical-section phased array to generate multiple foci using the field conjugation method is examined. The effect of the grating lobes on the power deposition pattern of the scanned field is shown to be minimal. Steady-state temperature distributions are simulated using a three-dimensional thermal model of the normal tissue layers surrounding a tumor of typical volume. The advantages and the limitations of this array configuration are discussed.