Recently the problems of electromagnetic interference in the hospital are serious. An optical biotelemetry technique using indirect light transmission was considered to be one of the solutions. However with this technique, the discrimination of the different transmitters has been difficult. To solve this problem, we have applied the spread spectrum (SS) technique to the optical biotelemetry. An experimental system was developed which consists of two transmitters and one receiver. In a transmitter, original signal is multiplied with a pseudo-noise code in time domain so as to have extremely wide frequency spectrum. And the resultant signal is transmitted out in light. In the receive, the received signal is multiplied with the same pseudo-noise code as the one in the transmitter. Using different types of the pseudo-noise code for each transmitter, many channels can be multiplexed in common space. This technique is less susceptible than other techniques to the interference due to extrinsic noises, multi-path propagation and signal of other channels. Using this system, some fundamental characteristics of this technique have been investigated. Internal noises of the system and threshold values for satisfactory transmission were measured. Finally, simultaneous transmission of ECG's from two transmitters were performed. The quality of transmitted signal was generally good having no cross-talk between two channels. With this technique the usefulness of optical biotelemetry in clinical use will expanded.

View full abstract

The action potential of the ventricle of the carp was measured by the optical method using laser light and WW781. WW781 is a voltage sensitive dye and fluorescence is changed by the membrane potential. The purposes of this study is to obtain basic data of the optical properties of WW781 and to measure the spike phase of the action potential of the ventricle for multi sites optical monitoring using a laser scanning system. In order to select light source, absorption spectra of WW781 in ethanol and in Ringer's solution were measured. As the absorption band was from 550nm to 635nm, He-Ne laser emitting at 632.8nm was selected. Fluorescence emission pumped by He-Ne laser was only detected from the dye in ethanol and wavelength of fluorescence maximum was 655nm. The ventricle was removed from the carp and was stained with a solution of WW781 (0.1mg/ml in Ringer's solution) for 30min. The staining period was determined by comparison of fluorescence emission spectra obtained from the ventricles which were stained for several periods. Only reflection of pumping light was observed from the ventricle stained for short period. It was necessary a staining period of 30min to get large fluorescence. After staining, absorption band of WW781 was from 590nm to 660nm. Time course of fluorescence emmision from the stained ventricle was recorded. In order to check light scattering arisen by contraction of the ventricle, we monitored time course of reflected light from the surface of the ventricle at the same time. Fluorescence emission intensity increased 2% during the spike phase. However the reflected light intensity did not change. Because the contraction occurred after the spike phase, both optical intensities changed similarly. He-Ne laser is a useful light source for optical measuring of membrane potential with WW781.

View full abstract

We have developed a system to automatically adjust the display window width and level (WWL) for MR images using neural networks. There were three main points in the development of our system as follows: 1) We defined an index for the clarity of a displayed image, and called “EW” EW is a quantitative measure of the clarity of an image displayed in a certain WWL, and can be derived from the different between gray-level with the WWL adjusted by a human expert and with a certain WWL. 2) We extracted a group of six features from a gray-level histogram of a displayed image. We designed two neural networks which are able to learn the relationship between these features and the desired output (teaching signal), “EQ, ” which is normalized to 0 to 1.0 from EW. Two neural networks were used to share the patterns to be learned; one learns a variety of patterns with less accuracy, and the other learns similar patterns with accuracy. Learning was performed using a back-propagation method. As a result, the neural networks after learning are able to provide a quantitative measure, “Q, ” of the clarity of images displayed in the designated WWL. 3) Using the “Hill climbing” method, we have been able to determine the best possible WWL for a displaying image. We have tested this technique for MR brain images. The results show that this system can adjust WWL comparable to that adjusted by a human expert for the majority of test images. The neural network is effective for the automatic adjustment of the display window for MR images, We are now studying the application of this method to MR images of another regions.

View full abstract

In the present paper, a simple QRS detection algorithm is proposed aiming at applications of QRS intervals and its variations to evaluate mental state. The pre-processing part of the proposed algorithm is a digital filter which has only three coefficients but has the characteristics of notch filter for the powerline interference and the bandpass filter for QRS frequency content. The decision rule is a simple one using two thresholds. The noise sensitivities were measured for a gold standard ECG; synthesized ECG corrupted with five different types of synthesized noise, which was introduced by Friesen et al. (1990). The performance of the proposed algorithm was compared with Engelese's algorithm which showed the best performance in the Friesen's study. Both algorithms made no false positive nor false negative for ECG corrupted with 50Hz powerline interference and with simulated baseline drift due to respiration. The proposed algorithm had more tolerance to abrupt baseline shift than the latter. Two algorithms had almost same performance for EMG corrupted noise. The proposed algorithm was tested using actual ECG measured under various mental and physical task conditions, which showed errors less than 0.12%. It is concluded that the proposed QRS detection algorithm has a good ability considering its memory and calculation load.

View full abstract

This paper proposed a method for analysing acceleration characteristics of voluntary movement of a hand that pursued a visual moving target. We calculated acceleration of a hand movement from measured position data by a numerical differentiation with some modification, and analysed the acceleration information both directly and by using a proposed mathematical model. As for a visual moving target signal, we adopted the trajectory which corresponded to an impulse signal of an acceleration of the target movement. By using the proposed method we analysed the data for 31 subjects categorized into 3 groups; 9 patients with celebellar ataxia, 6 patients with Parkinson's disease and 16 normal adults. We obtained the following results both by direct analysis and by using the mathematical model analysis. Normal adults and patients were clearly distinguished by parameters of the dead time of a hand motion. Cerebellar ataxia and Parkinson's disease were shown to be distinguished by parameters which expressed quickness and damping of the impulse response of acceleration.

View full abstract

We considered that an analysis of regional image (ARI) on endoscopic ultrasonography (EUS) would be useful for differential diagnosis of gastric cancer. However, quantitative measures of regional ultrasonographic image alter as a function of position within the field of view, even in case of an uniform ensemble of scatters. It is important to investigate an optimum extent on ultrasonography for ARI. Therefore, we tested the influence of distance and azimuth from EUS to region of interest (ROI). We have used a phantom and resected specimen in this study. The results showed that there was an optimum extent for ARI. We conculded that it is necessary to adjust EUS's position in EUS examination.

View full abstract

A transmission system of control data from outside of human body into inside to change the operation conditions of implanted electronic apparatus is presented. Since the system is to be operated in the atmospheric noises caused by electromagnetic field existing in the surroundings of human life, the media, frequency and modulating type of the data transmission were examined. So pulse amplitude modulated ultrasonic waves of 1MHz, 600bit/s, was selected. The tranmitter consists of key board, microcomputer, modulator and sending transducer, while the receiver consists of wave receiver, demodulator, microcomputer and control circuit. New data input on the key board are converted into asynchronous format signal by the microcomputer, and pulse modulates the amplitude of ultrasonic carrier waves. The received signal is, meanwhile, converted to control signal by the microcomputer to be input into implanted apparatus. For application of this system to cardiac pacemaker, the transmitted data signal is composed of 8-bit opening flag, 8-bit pulse rate data, 4-bit pulse voltage data, 4-bit pulse with data and 8-bit closing flag. Those datas are sent 6 times repeatedly so as to prevent transmission error. Received data are compared with one another, and, when all the data are same, the microcomputer accepts them as the new data and transmitted into implanted apparatus to change operation condition. An experimental transmission test of this system took place with 25mm thick beef meet, likening to human body. In this test, lead zicronate titanate ceramic transducer of 40mm dia. and 1MHz resonance frequency were applied to wave transmitter and receiver, and the operation conditions of pacemaker could be changed to such extent as 42-124pulse/min of pulse rate, 1.0-4.0V of pulse voltage and 1.0-3.0ms of pulse width.

View full abstract