For an automatic supervision of respiration, the following themes are mainly investigated to synthesize an stable and safe control system of respiration: 1) To find appropriate a mathematical model for the description of a respiratory system which is readily applicable to the synthesis of its control system, 2) To determine an appropriate method for the assurance of stability of the control system, 3) To develop an appropriate programmable respirator which is suitable for its automatic manipulation. For the effective study of the above themes, a new type of programmable respirator is first developed, which can be controlled by outside signals. Secondly, an appropriate mathematical model of input-output relation about ventilation and alveolar CO₂-concentration is determined. Thirdly, to overcome an occasional emergence of its unstability in the controlling process, the control system is designed according to an adaptive pole placement method. Then, healthy young students are subjected to control experiments to clarify the feature of the proposed method, resulting in good control of alveolar CO₂-concentration even when air mask is not tight enough or when a subject has a cough during its control process.

View full abstract

Aim: It has been reported that impairment of left ventricular (LV) relaxation decreases early diastolic coronary blood flow rate (CBF). On the one hand, recent studies showed that mechanical vibration during diastole (diastolic vibration) accelerates LV relaxation rate instantly both in clinical settings and in experimental animals. Therefore, we examined whether diastolic vibration could increase diastolic CBF in regionally ischemic heart. Methods: In 9 open chest canine hearts, we narrowed the perfusion line to reduce total CBF to 50% of control. Applying 50Hz, 2mm amplitude diastolic vibration to the epicardium of the LV, we measured CBF, coronary perfusion pressure (CPP) and LV pressure. Results: Diastolic vibration shortened the time constant of the LV pressure fall (p<0.01), increased diastolic CBF (p<0.01) and decreased coronary vascular resistance (R) (p<0.01). The percentages of increase in diastolic CBF (p<0.05) and decrease in R (p<0.01) were larger in ischemia than in control. The percentages of increase in diastolic CBF (p<0.05) and decrease in R (p<0.01) were smaller with papaverine HCl injection (i. e; at attenuated vascular tone) than without injection. Diastolic vibration did not decrease systolic CBF. Conclusions: Diastolic vibration increased diastolic coronary blood flow rate and decreased coronary vascular resistance. We speculated that this was caused by the mechanism that diastolic vibration accelerates LV relaxation rate and dilates the coronary vasculature directly.

View full abstract

Instantaneous time-varying spectral analysis of three kinds of evoked responses (ABR: auditory brain-stem response, MLR: auditory middle latency response, SVR: auditory slow vertex response) were carried out. The conventional spectral analysis is to be considered as evoked response as a function of frequency (called time-invariant spectrum). The time-varying spectral analysis represents both the time and frequency characteristics of the evoked response at the same time. The results of time-varying spectrum (TVS) of averaged evoked responses (300 times) were as follows: The dominant frequency ranges of the TVS of normal ABR occurred at around 866-945Hz (frequency range A), 472-551Hz (range B), and 236-315Hz (range C). The main spectral range of wave I were distributed in the frequency range A and B; wave II, range A; wave-III, range A and B; wave IV, range A; wave V, range A and C. The TVS of normal MLR was concentrated at around 45-56Hz in frequency. The TVS showed its peaks at around the latencies of each component of MLR. The TVS of normal SVR had the two dominant frequency ranges (9.8-12.2Hz and 2.4-4.9Hz). The component with higher frequency range (9.8-12.2Hz) showed shorter latency (from P 1 to P 2), and it's magnitude was about 40% compared with that of the lower frequency range (2.4-4.9Hz). The relative magnitude at the lower frequency range showed increased, in accordance with the components of longer latency, from P 2 to N 2.

View full abstract

We proposed a new type Neural Network model by which we could detect a several kinds of important characteristic waves in EEG that is needed for the diagnosis of sleep stages. In this paper, we compared the proposed method with two conventional type Neural Network medels and likelihood ratio method as to the performance of detecting the characteristic waves from sleep EEG. Experimental results indicated that the proposed Neural Network model had much more capacity for detecting the isolated and transient characteristic waves than other conventinal methods and little depended on the duration of characteristic waves. Analyzing the connection weights of the Neural Network model revealed that XOR operation was carried out for detecting the isolated characteristic waves.

View full abstract

In recent years, magnetic stimulation has emerged as a useful method for stimulating nerves in clinical site, because magnetic stimulation is noninvasive and less painful than electric stimulation. In magnetic stimulation, it is important to estimate induced currents produced in a volume conductor. It is, however, difficult to predict distributions of induced currents in an inhomogeneous medium like as the human brain. There is no paper has shown the accurate distributions of induced currents in such that medium. In this paper, we measured spatial distributions of currents generated in the tank model of an assumed human brain filled with saline solution, when the tank model was exposed to pulsed magnetic fields. We used a commercial magnetic stimulator. The induced currents were measured with a current probe which is made of ultra-mini coaxial cable. We discussed the focality of the induced current and an effect of dura on the induced current in the vicinity. In the half-spherical tank, the part of the cross section of figure-of-8 coil becomes the point of highest current density, while the part corresponding to an edge of a coil also becomes the point of higher current density. In the tank with skull, the distribution of the current is also compressed near the bone and the current flows along the surface of inside skull. The value of current density perpendicular to the skull decreases about 20%. On the other hand, the current density perpendicular to the dura decreases about 15% rather than without dura.

View full abstract

In paraplegic standing using functional electrical stimulation (FES), closed-loop control systems have been studied to maintain knee joint in extended position with least muscle contractile force. In this case, the selection of the controlled variable is important. Muscle force is considered to be useful as a controlled variable. However, there is no clinical method of directly measuring muscle force during motion. Hence, we used the tension of the patellar ligament as a controlled variable which reflects the contractile force of the knee extensor (the quadriceps femoris). In this study, a new method is proposed to estimate the tension of the patellar ligament. We estimated the tension of the patellar ligament by measuring the patellar ligament reaction force (PLRF) using the force transducer pushing the patellar ligament through skin. The following results were obtained in paraplegic experiments. The linear relationship existed between PLRF and the knee extension torque, and the transient response of the PLRF was similar to that of the knee extension torque. The PLRF had the good reproducibility, after a few prior stimulation to the muscle. The closed-loop control using PLRF maintained the knee extension torque above the desired value.

View full abstract