The progress in the field of organ transplantation and artificial organ has been remarkable in the recent years. The number of clinical cases is increasing daily, and the clinical renal transplantation will soon be performed in more than 2000 occasions. One human life has been supported by artificial kidneys for over eight years. Many hearts and livers have been transplanted, and artificial auxiliary left ventricles have been functioning in human bodies. The ancient dream of replacement of organs has almost been realized.
In reviewing the history of development of organ substitutes, the current status in reference to various organs has been described as precisely as possible, along with the expected development and problems in the future.

View full abstract

This paper describes the recognition method of random mass patterns corrupted by additive Gaussian noise. In our previous papers this method was applied to the recognition of abnormal shadows in chest roentgenograms and found to work effectively. In this paper we theoretically study the method from the more general viewpoint.
We first construct simple models of input patterns, which consist of (1) some irregular figures whose uniform densities are known and whose shapes, sizes and positions are unknown t (random mass patterns, abbr. RMP) and (2) additive Gaussian noise with 0 mean and a known covariance function.
We next derive a recognition method for these patterns employing the likelihood_ ratio test. Performance of the method is estimated by calculation of error probabilities and. computer simulation with the use of artificial RMP's and normal random numbers. Furthermore, the decision function used here is shown to be equivalent to the two-dimensional smoothing by a suitable weight function with the explained physical meaning.
The whole recognition procedures including the method for design of the recognition system presented here will be applicable to processing of various photographs conveying medical information.

View full abstract

The concept of equivalent cardiac multipole components or electrocardiographic lead field function which is an analytical aspect of tensor cardiography was studied in its generalized form in an inhomogeneous medium of finite extent. As a result it became clear that the cardiac multipole components could not be determined by the ECG body surface potential alone.
This implies that the ideal lead field in tensor cardiography requires to place the electrodes not only on the body surface but also on the inhomogeneous phase boundaries.
The impact of inhomogeneity of the medium on the apparent cardiac multipole components computed from the body surface potential alone was shown to be significant even in the first term of a simple two-dimensional model.
A numerical method to compute the cardiac multipole components from the ECG surface potential and geometry of the outer and phase boundary surfaces was proposed to quantitate physicians' diagnosis on ECG.

View full abstract

There have been rather few accounts in the literature of quantitative relationship between the frequency of electro stimulation of the cardiac vagus nerve and the induced change in the heart rate. We have conducted some experiments further on this relationship and observed that the heart beat interval fluctuates periodically although the stimulation frequency less than about 5 pulses/sec is kept constant. The difference between the maximum and the minimum of the heart beat intervals was about 50-100 ms. The frequency of this periodic variation in the heart beat interval is well approximated by the difference between the vagal stimulation frequency (F_v) and the heart beat frequency (F_h), or 2 F_h, depending on F_v being around F_h and 2 F_h respectively. When F_v was close to F_h or 2 F_h, the above-mentioned variation was not observed. When the periodic variation was observed, the interval between each stimulus and the following P of EKG fluctuated also periodically with the same frequency. The relationship between this stimulus-P interval and the heart beat interval in which the stimulus was given was studied and a chronotropic function curve was introduced. This curve denotes the extent of delay of P when continuous stimuli were applied to the vagus nerve. This curve was approximated by a linear superposition of the chronotropic function curves corresponding to each stimulus.
From these results it was concluded that the variation in the heart beat interval is produced because P waves take periodically changing values of the chronotropic effect which is produced by superposition of effects of each stimulus.

View full abstract

The ultrasound cardiogram of the tricuspid valve has been hardly recorded, while that of the mitral valve is easily done.
It is probable that the tricuspid valve makes an untoward angle with the ultrasound beam from the anterior chest wall so that the ultrasound beam may be little reflected from the valve in a direction to the transducer. The tricuspid valve is situated 4-5 cm. behind the sternum. Therefore in detecting its motion, the ultrasound beam should be sent obliquely to an anteroposterior direction from the vicinity of the left sternal edge to the retrosternal area. area. When the transducer is kept obliquely to the chest wall in manipulation, 1) the reproducibility and fine adjustment of the site and directing of the transducer are difficult to maintain, because the site and direction should be adjusted depending mainly on the memory of the operator, 2) the direction of the transducer is so unstable that anatomical relations between the ultrasound beam and the valve may be changed even due to a slight movement of the chest wall as in respiration.
Such discrepancies seem to result in difficulty of detecting an echo from the tricuspid valve.
The transducer was immersed in a water-filled vinyl bag placed on the chest wall so that it could be applied to desired site and direction without direct contact to the chest wall. It was made concave so that the ultrasound beam could pass through an intercostal space and be focused in a thin beam.
The present equipment enabled one to obtain a satisfactory ultrasound cardiogram of the tricuspid valve in about half of the healthy subjects examined.

View full abstract

There has been little or no statistical data processing study of the basal body temperature curves. The pattern recognition acquired from these curves, the prediction of the length of future menstrual cycles, and the prediction of estimated value of the basal body temperature on everyday in future cycles reported statistically. In order to solve these problems, computing the autocorrelation coefficients on the observed values of the basal body temperature in the past cycles, we classified the pattern of the basal body temperature for each of the women. The autocorrelogram represented the characteristic of menstrual cycle of each woman, who exhibited the individual difference on her cycle.
The prediction of the estimated value of the basal body temperature on everyday in future cycle rather than the prediction of the length of future cycle alone has been of great importance clinically. For the purpose of the prediction, using the autocorrelation function and the weighting function on the BBT curves, we predicted the estimated value of the everyday basal body temperature in the next cycle of menstruation. The graph predicted as to the next cycle by the estimated value from the past cycle and the graph recorded with respect to the measured value in the next cycle showed a similarity with some accuracy.

View full abstract

We studied the sensitivity and accuracy of a pH-measuring radio telemetering capsule under various conditions.
The results obtained were as follows :
1) Temperature had some influence on the function of the capsule : a change of 1°C caused a change in the pH recording of about 0.1.
2) The capsule responded so rapidly to pH change of the solution that its time lag was negligible.
3) The capsule maintained correct pH values in the solution containing such various. ions as those in the human gastric juice and also under pressure changes corresponding to, the Intragastric pressure changes.
4) The capsule, being calibrated within a range of±0.25°C of the temperature in the solution tested with, responded with an accuracy of about 0.1 pH unit during at least 2 hours.
5) On the basis of the in vitro evaluation of the pH-measuring capsule, we confirmed its possible clinical merit in determination of gastric pH in vivo.

View full abstract