JPH01204644A - Device for treating blood flowing sound in cranium - Google Patents

Abstract

PURPOSE:To estimate the existence and position of a morbidly changed site in a cranium with high accuracy by detecting a blood flowing sound and obtaining a coherence function between channels and removing asynchronous noise components based on the changing pattern in time of the peak value thereof and extracting a blood flowing sound component. CONSTITUTION:Blood flowing sound detecting means 6a, 6b detect a blood flowing sound generated from an morbidly changed site in the cranium of an examined body P and the result is taken into an A/D converter 9 via preamplifiers 8a, 8b. Electrodes 10a, 10b, 10c detect the heart beat of the examined body P and the result is taken into the A/D converter 9 via an electrocardiograph 11. These are converted into digital signals by the A/D converter 9 and taken into a wave-form data analyzing portion 13 via a buffer memory 12. In the analyzing portion 13, a first means 13a obtains a coherence function indicating the degree of relation between each channel from the detected result of the blood flowing sound and, based on the changing pattern in time of the peak value thereof, a second means 13b removes asynchronous noise components such as respiratory sound to extract a blood flowing sound component.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Field of Industrial Application) The present invention is a non-invasive method for detecting abnormal blood flow sounds caused by cranial lesions such as cerebral aneurysms and arteriovenous malformations from outside the cranium. death,
The present invention relates to an intracranial blood flow sound processing device that makes it possible to estimate the presence or absence and position of the lesion site by analyzing this.

(Prior art) It is well known that intracranial masses and arteriovenous malformations, which are the main causes of subarachnoid hemorrhage, emit abnormal blood flow sounds. Methods for non-invasively estimating the location of venous malformations from outside the skull have been reported (e.g. CLINI-CAL NEURO3CHI).
NCE Vol. 3 No. 8 (1985) P
, P 86B-867). This method detects abnormal blood flow sounds using a blood flow sound detection means (detector) placed outside the skull, and analyzes this detection signal.
As the blood flow sound detecting means, one shown in FIG. 7 or the like is used. That is, this blood flow sound detection means 6 has an acceleration detector 1 disposed inside a case body 2 formed in a cylindrical shape with a bottom, supports this acceleration detector 1 with a stabilizer 5, and connects the acceleration detector 1 and the case. It is constructed by attaching a spring 3 between the body 2 and the bottom of the body 2, and by abutting the acceleration detector 1 against the head 4 of the subject (usually a patient) against the biasing force of the spring 3, blood is removed. It attempts to detect flowing sound.

By attaching a plurality of such blood flow sound detection means 6 to the subject's head 4 and processing and analyzing the detection results, presence or absence and position of a cranial lesion site are estimated.

(Problem to be Solved by the Invention) However, blood flow sounds are very small, and the blood flow sound detection means for detecting abnormal blood flow sounds requires not only blood flow sounds but also other noises such as breathing sounds and saliva inhalation sounds. This has caused a problem in that even if the detection signal of the blood flow sound detection means is analyzed, the presence or absence and position of a cranial lesion cannot be easily estimated.

SUMMARY OF THE INVENTION Therefore, the present invention aims to provide an intracranial blood flow sound processing device that can estimate the presence or absence and position of a cranial lesion site with high accuracy based on abnormal blood flow sounds.

[Structure of the Invention] (Means for Solving the Problems) The present invention has a plurality of blood flow sound detection means for detecting intracranial blood flow sounds of a subject, and the detection results of the blood flow sound detection means are In the intracranial blood flow sound processing device that analyzes intracranial blood flow sound based on the detection result of the blood flow sound detection means, a first means for calculating a coherence function indicating the degree of association between each channel from the detection result of the blood flow sound detection means; and second means for extracting the blood flow sound component by removing the asynchronous noise component based on the temporal change pattern of the peak value of the coherence function.

(Function) In the present invention, a coherence function indicating the degree of association between each channel is obtained from the detection results of a plurality of blood flow sound detection means, and asynchronous detection is performed based on the temporal change pattern of the peak value of the obtained coherence function. The blood flow sound component is extracted by removing the noise component, and by doing so, the presence or absence of a cranial lesion site and the position estimation are improved with high accuracy.

(Example) Hereinafter, the present invention will be described in detail based on an example shown in the drawings.

FIG. 1 shows a state in which an intracranial blood flow sound processing device according to an embodiment of the present invention is attached to a subject (usually a patient) P.

As shown in the figure, this intracranial blood flow sound processing device includes blood flow sound detection means 5a, 6b, and a preamplifier 8a.

8b, A/D converter 9, electrodes 10a, 10b.

10c, an electrocardiograph 11, a buffer memory 12, and a waveform data analysis section 13.

The blood flow sound detecting means 6a, 5b are for detecting the blood flow sound generated from the cranial lesion site of the subject P, and are equipped with an acceleration detector 1 as shown in FIG. 7, for example. Applicable. Although a large number of blood flow sound detection means are actually arranged on the head of the subject P, only two are shown in FIG.

The detection results of the blood flow sound detection means 6a and 6b are detected by the preamplifier 8a.
, 8b, respectively, to the A/D converter 9. Further, the electrodes 10a, 10b, and 10c are used to detect the heartbeat of the subject P, and the detection results are taken into the A/D converter 9 via the electrocardiograph 11.

The outputs of the preamplifiers (3a, 8b are converted into digital signals by the A/D converter 9 and then sent to the buffer 7 memory 12.
The waveform data analysis unit 13 receives the data via the waveform data analysis section 13.

This waveform data analysis section 13 analyzes waveform data, and functionally includes a first means 13a and a second means 13'.
It has b. The first means 13a determines a coherence function indicating the degree of association between each channel from the detection results of the blood flow sound detection means 6a and 6b, and the second means 13b determines the peak value of this coherence function. The blood flow sound component is extracted by removing asynchronous noise components such as breathing sounds based on the temporal change pattern of the noise.

Hereinafter, the operation of the intracranial blood flow sound processing apparatus shown in FIG. 1 will be explained in detail.

As shown in Figure 2, the intracranial blood flow velocity V is uniform, especially near the 11is ring, which is a frequent site of vascular lesions such as cerebral aneurysms, for example, at the origin of the middle cerebral artery. The electrocardiographic waveform (ECG) R wave occurs at time A, which is very slow, and after time td has elapsed, at time B, it is very fast.

On the other hand, abnormal blood flow noise in the skull is caused by the eddy sound of blood flowing through a cerebral aneurysm, and the turbulent sound at localized narrowings in blood vessels such as arteriosclerosis. These eddy flow sounds and turbulence sounds are difficult to generate around time A where the intracranial blood flow velocity (2) is slow, but they tend to occur around time B where the blood flow velocity V is high, and the power of the abnormal blood flow sound increases. However, unnecessary noises other than blood flow sounds are uniformly generated at a certain level regardless of whether the intracranial blood flow velocity V is slow at time A or fast, and is mixed into the blood flow sound detection means 6a and 5b. do.

Here, in a time phase in which abnormal blood flow sound does not occur, there is no significant signal and most of the noise components are present, so the correlation between multiple channels becomes weak and the value of the coherence function between any two channels is small. On the other hand, in the phase in which abnormal blood flow sounds occur, contrary to the above, significant signals increase, so the value of the coherence function becomes close to 1. Therefore, abnormal blood flow sounds can be extracted using the pattern of temporal changes in the coherence function.

The electrocardiographic waveform (ECG) shown in FIG. 3 was obtained from the electrocardiograph 11 shown in FIG. The waveforms of channels 1 to n are blood flow sounds detected by n blood flow sound detection means placed on the head of the subject P. The blood flow noise of channels 1 to channel n is converted into digital data by the A/D converter 9, but the time td1 has elapsed from point A when the intracranial blood flow velocity is slow and there is very little abnormal blood flow noise. Buffer 7 memory 12 as digital data at t1
The digital data at time t2 after time td2 has elapsed from time point A, where the blood flow velocity is high and a lot of abnormal blood flow noise is included, is also stored in the buffer 7 memory 12.

FIG. 4 shows the intracranial blood flow waveforms of two arbitrary channels (channels 1 and 2 in this case) from the time point of the QRS triage of the electrocardiogram. As the time phase from the QRS trigger time, the delay time is Q, td2, td3, ... to the time interval tl, t.
2. t3. For example, the coherence function of two-channel data of ... is as shown in Fig. 5, and its peak values a, b,
6. d increases or decreases depending on each time phase.

In this case, when asynchronous noise such as breathing sounds is not mixed, there is no significant signal in the blood flow sound between fd2 and t3 when the blood flow velocity is high in a normal person, as shown by the dotted line in Figure 6. The values of the coherence function (Coh) at each time phase are small and do not vary greatly.

However, when there is abnormal blood flow sound, the coherence function becomes 1 between and around td2 to t3, as shown by the solid line in the same figure, because the signal is significant in areas where the blood flow velocity is high.
The value is close to . Furthermore, since asynchronous noise such as breathing sounds is uniformly mixed into each channel, a pattern as shown by the broken line in FIG. 6 is obtained, regardless of the intracranial blood flow velocity pattern.

In the waveform data analysis section 13, first, the first means 13a
The coherence function between each channel (for example, between two arbitrary channels) is determined by sequentially shifting the time phase from the synchronized pulse of the electrocardiogram waveform, and then the second means 13b calculates the temporal peak value of the coherence function. Based on the conversion pattern, asynchronous noise components such as breathing sounds are removed, and blood flow sound components are extracted. Based on this blood flow sound component, features of turbulent flow and thin flow sound are extracted.

It goes without saying that the present invention is not limited to the above-mentioned embodiments, and that various modifications can be made.

[Effects of the Invention] As detailed above, according to the present invention, a coherence function indicating the degree of association between each channel is obtained from the detection results of a plurality of blood flow sound detection means, and the time of the peak value of this coherence function is calculated. Since asynchronous noise components such as breathing sounds are removed based on the physical change pattern, blood flow sound components can be accurately extracted, and the presence and location of cranial lesions can be estimated with high accuracy. It has the excellent effect of being able to

[Brief explanation of the drawing]

Fig. 1 is a block diagram of an apparatus according to an embodiment of the present invention, Fig. 2 is a characteristic diagram of blood flow velocity, Fig. 3 is a diagram showing the relationship between electrocardiographic waveforms and blood flow velocity, and Figs. The figure is a waveform diagram for explaining the operation of the apparatus of this embodiment, and FIG. 7 is a diagram illustrating the configuration of the blood flow sound detecting means. 6a, 6b... Blood flow sound detection means, 7... Subject, 1
3... Waveform data processing section, 13a... First means, 13b... Second means. Agent Patent Attorney Nori Ken Chika Ken Shu Kondo Takeshi Figure 3

Claims (1)

[Claims] An intracranial blood flow sound processing device that includes a plurality of blood flow sound detection means for detecting intracranial blood flow sounds of a subject and analyzes the intracranial blood flow sounds based on the detection results of the blood flow sound detection means. In,
A first means for determining a coherence function indicating the degree of association between each channel from the detection result of the blood flow sound detection means, and removing an asynchronous noise component based on a temporal change pattern of the peak value of the determined coherence function. and second means for extracting blood flow sound components by extracting blood flow sound components.