The on-line and off-line parallel information processing systems for rat's long-term locomotor activity, electroencephalogram, electrooculogram and electromyogram are proposed in this study. They provide the users with many functions, which will help them to study the sleep pattern, biological rhythms and bioassay etc. To use the on-line information processing system, which mainly determines the physiological state of the rate in real-time, it is only necessary to set some parameters which will determine the operational conditions, such as date, starting time, data sampling interval, unit processing time and duration of unit experimental procedure. Every process is, then, performed automatically and all the necessary data for off-line processing, which can equivalently perform on-line processing under different conditions and make statistical analysis, are stored in a disk file simultaneously. In addition to these, by automatic calculation of the proper reference values used for calculating ratios, logical combinations of which determine the physiological state in advance or when necessary, even during the determination process of physiological state, misdetermination caused by the differences of individual rats, operation technique of implanting electrodes and their gradual displacement can be avoided. Some experiments are also carried out to verify the proposed data processing systems.

View full abstract

A new method of estimating the number and the size of active motor units by processing mass electromyogram (EMG) has been devised. This method is based on a model of mass EMG generation, which is founded on the knowledge of the mode of motor unit activities. This model is described as follows : (1) Mass EMG is the sum of all active motor units action potential trains. (2) An input of each motor unit is a statistically independent random pulse train. (3) Motor units are devided into groups by their threshold force for recruitment. N_j denotes the number of motor units belonging to group j, K_j denotes the size, and f_j (P_i) denotes the firing rate. (4) The firing rate f_j (P_i) is the function of force.
By using the theory of the shot noise, the number N_j and the size Kj of motor units of each group are expressed as a function of the second and fourth moments of mass EMG (m₂ and m₄) and the firing rate f_j (P_i). This estimation starts from the lowest threshold force group, using m₂, m₄, and f_j (P_i).
This method has been applied to the human brachialis muscle and the human extensor digitrum communis muscle. The estimated results agree with the size principle and the physiological knowledge of the relation between the threshold force and the number of motor units. This agreement comfirms the propriety of this estimation method.
The estimation accuracy, that is, the relation between the observed period of mass EMG, which means the integration time for the calculation of moments, and the deviation of estimated values, is elucidated theoretically.

View full abstract

This study has been attempted to evaluate the parallel conductor hypothesis (Nyboer, 1950) for the measurement of limb blood flow by electrical impedance or admittance plethysmography using a compensation technique. The principle of this technique is based on the theory that the admittance change in a limb segment immersed in an electrolyte solution disappears when the resistivity of the solution is equal to that of the blood flowing into the segment (ρ_b), if the segment is regarded as a parallel conductor model. The relationship between the admittance change and volume of the blood flow in the segment can be precisely analyzed by this technique.
In the forearms of 10 healthy subjects, ρ_b values have been compared with the resistivity of the electrolyte solution (ρ_s0), when the admittance variation following venous occlusion almost completely disappeared. These two values showed only a slight difference. From the difference between ρ_b and ρ_s0, the error which will be caused by modelling a limb segment as a parallel conductor was calculated to be within ±2%. This suggests that the parallel conductor hypothesis would be acceptable for the measurement of blood flow in the limb segment. Using the results the resistivity constant (k=ΔV/V₀ vs. ΔY/Y₀) and the form factor (x) in each limb segment have also been calculated. Based on these data the difference between the parallel and the random-pooling model is discussed.

View full abstract

A mathematical insulin secretion model has been constructed by analysing the glucose-induced insulin secretion dynamics with the aid of control theory.
The following equation was used to express the relationship between insulin secretion, i (t) and the glucose concentration, x (τ);
i (t) =∫^t_-∞C_px (τ) e^{- (t-τ) /T₁}dτ+∫^t_-∞C_px' (τ) e^{- (t-τ) /T₂}dτ
where x' (τ) is the rate of change in glucose concentration, T₁ and T₂ are the constants in time lag for proportional and derivative components respectively, C_p and C_d are the coefficients for proportional and derivative components respectively.
The validity of the model has been studied by comparing the physiological experimental data obtained from the isolated rat pancreatic islets perifusion system with the simulation results of the equation. Step-wise or ramp-mode glucose stimulation was applied to induce insulin secretion.
The following results were obtained :
1) The mathematical model fitted well with physiological experimental results.
2) T₁ and T₂ was proved to be 12 min. and 2 min. respectively.
3) Relationship between C_p and glucose concentration was linear within the physiological glucose levels.
4) It has been shown that Cd had the negative linear correlation with the rate of change in glucose concentration.
5) When the rate of change in glucose concentration was negative, the presence of inhibitory factor on insulin secretion was proven.
In summary, glucose-induced insulin secretion was revealed to be as the function of a proportional plus derivative response to glucose input with the first order time lag. This phenomenon suggests that pancreatic β-cell recognizes both blood glucose concentration and the rate of change in glucose concentration, then secretes insulin in response to them. The validity of the model was also recognized in vivo state clinically, when used as the basis for determining insulin infusion rates of artificial beta cell.

View full abstract

This paper describes the improvement of response on transcutaneous gas tension measurement by mass spectrometry. Gas diffusion through the skin, the membrane of a transcutaneous sampler and a connecting tubing to a mass spectrometer, is analyzed theoretically. The theoretical result shows that thinner membrane with lower-permeability and shorter connecting tubing of larger inner diameter are able to improve the in vivo transient response of the apparatus without decreasing transcutaneous gas tension. A polypropylene membrane 25μm thick and the connecting tubing of 1.9 mm inner diameter and 9 cm length were used in the experiment instead of a Teflon membrane 120μm thick and a commercially developed connecting tubing of 0.5 mm inner diameter and 180 cm length. Then in vitro response and in vivo response were reduced to 20% and 60% of the previous value respectively without decreasing transcutaneous gas tension. This result suggests the manner of selecting the membrane and the connecting tubing for measurement.

View full abstract

To analyse the dynamic behavior of the whole cardiovascular system, the transient responses of the system parameters to sudden hemodynamic change have been formulated as developed from the theory of the circulatory equilibrium by Guyton. The time constant T in the transient venous pressure change has been given by T=C_v/ (G_VR+G_CO), where C_v is the systemic venous compliance, and G_VR and G_CO are the gradient (conductances) of the venous return curve and the cardiac output curve, respectively.
The in-vivo measurement of these parameters was carried out in the acute and chronic experiments in dogs by the technique of right heart bypass using an artificial pump. A transient change in the central venous pressure was induced by exact stepwise changes in the cardiac output function of the artificial pump. The time course of the consequent venous pressure was almost a monoexponential function of time which allowed accurate measurement of the time constant T. The gradient of the venous return curve G_VR was calculated from the differences in cardiac output (ΔCO) and venous pressure (ΔP_v) between the two equilibrating conditions before and after the transient change by G_VR=-ΔCO/ΔP_v. Together with the known value of the gradient of cardiac output function of the artificial pump G_CO, the systemic venous compliance C_v was estimated from the above equation. The mean values of T and C_v measured under anesthetized states were 3.9± 0.69 (SD) s and 1.79±0.13 (SD) ml/mmHg/kg, and those measured under conscious states were 4.1± 0.84 (SD) s and 1.60±0.22 (SD) ml/mmHg/kg, respectively. The results obtained from the conscious states showed significant decreases in the values of venous compliance against those obtained from the anesthetized conditions (p<0.05). The C, values measured by this transient method were comparable with those measured with the steady method previously reported by other investigators.
It is concluded that the time constant of the venous pressure change is a good index to represent the transient response of the circulatory system and that the transient method presented here is a reliable technique to estimate systemic venous compliance in conscious animals.

View full abstract

The adaptive correlating filter (ACF) is effective to detect P-waves in ECG's buried in noise or overlapped with the QRS or T-waves. In this paper, an index has been applied to the evaluation of three templates (i. e., a Gaussian waveform, a typical P-waveform obtained from normal ECG and segment data involving a P-wave taken out from each ECG being processed) used in the P-wave segmentation procedure as a pre-procedure of the P-wave recognition program with the ACF. The index was highly reasonable in the sense that the derivation of the index was based on the performance characteristics of the ACF.
As a result of the evaluation using 14 cases of ECG's with A-V block, the Gaussian template was shown to be superior to the other templates.

View full abstract

A computer-based system for calculating stroke volume, heart rate and cardiac output automatically from the transthoracic impedance variation has been developed. A beat-by-beat determination has been performed on several healthy subjects and a patient with atrial fibrillation either during breath holding maneuver or during spontaneous breathing at rest. A mini-computer was used to measure the systolic peak height of the first derivative of impedance change (dZ/dt) min, the ventricular ejection time ET and the basic impedance of the thorax Z₀ for each cardiac cycle. The ET was determined on the dZ/dt wave from the time point corresponding to.15 (dZ/dt) min, to the most positive peak following systole. The respiratory drift appearing on the dZ/dt during breathing was corrected by determining the base line for each cardiac cycle, using as reference the dZ/dt values at ECG-R wave which was obtained during breath holding maneuver. Stroke volume was, then calculated using the formula of Kubicek et al. (1966). Heart rate was determined from R-R interval of ECG. Cardiac minute volume was obtained by multiplying the heart rate by stroke volume. A comparison between the computed stroke volumes and those obtained from a manual calculation showed good agreement.

View full abstract