Geliştirilmiş Bir Ultrasonik Darbeli Doppler Kan Akış Ölçme Düzeninde Hata Kaynaklarının Analizi

Abstract

Bu çalışmada, darbeli ultrasonik Doppler metodu kullanarak, gerçekleştirilen akış ölçerlere etki eden hata kaynaklarının teorik analizleri yapılarak deneysel sonuçlarla karşılaş tırılmıştır. Sistem sonucunu etkileyen hatalar iki temel başlık altında incelenmiştir. Bunlardan birincisi fizyolojik sistemden kaynaklanan hatalar, ikin cisi ise elektronik sistemden kaynaklanan hatalardır. Hastalarda doku yapılarının farklılık göstermesinden dolayı, farklı hastalarda aynı damarda yapılan ölçmeler de farklı neticeler vermektedir. Ayrı ca dokuları belirli yapıda ya da yapılarda kabul etmek de mümkün değildir. Bu durum ise nicel ölçme yapmayı engellemektedir. Fakat bütün bu olumsuz durumlara rağmen, genel bir doku modeli ele alına rak, Doppler ölçme sonucuna olan etkisi bulunup değiştirilen model parametreleriyle de yine ölçme sonucuna olan olumsuz katkıları göz lenebilir. Buna ait çalışma, bir çok araştırmacının kabul ettiği doku modeli ele alınarak bir bilgisayar simülasyonu ile yapılmıştır. Elektronik sistemden kaynaklanan hataların analizine ait de neysel sonuçlar, halen klinik çalışmalarda kullanılmakta olan 4 MHz lik bir darbeli Doppler akış ölçerine ilave elektronik donanım ge liştirilerek, geliştirilen bu akış ölçer üzerinden elde edilmiştir, ilave elektronik donanımları değişik darbe tekrarlama frekansının seçilebilmesi, gönderilen darbedeki titreşim sayısının seçilebilme si, örnek alma kapı genişliğinin ayarlanabilir olması, Doppler işa retinin 90° faz farklı elde edilmesine imkân vermesi ve değişik band genişliğinde darbeli Doppler dönüştürücülerini kullanabilmeye imkân veren empedans uydurma devresini içermektedir. Elektronik sistemde olması gereken bazı parametreler gerçek Doppler işaretinin elde edilmesini engellemektedir. Bu tezde girişim gürültüsü, Doppler filtresi, frekans örtüşmesi, çift yanband oluşumu ve örnek hacimde birden fazla damar olması halinde meydana gelen etkilerin analizleri yapılarak, deneysel sonuçlarla karşılaştırılmış ve bu alanda litera türde görülen önemli bir boşluk doldurulmuş bulunmaktadır. Bu çalışma,, şimdiye kadar darbeli Doppler sistemlerinin nicel ölçme yapmadığını ortaya koymakta, ancak yukarıda bahsedilen hata kaynaklarının dikkate alınmasıyla nicel ölçme yapılabileceğini be lirtmektedir.

The cardiovascular system is one of the major systems of the human body. Its main purpose is to distribute blood to all parts of the body. Quantitave measurements of blood flow using ultrasonic Doppler techniques have considerable importance in clinical measurement. Since pulse wave Doppler has range resolution, and therefore can provide the information at a particular site of the vessel lumen, it is capable of measuring dynamic cardiovascular systems mode effectively than continuous wave Doppler. unfortunately because of its complexity, it requires a skilled operator to obtain blood flow information. For this reason, researchers have worked to reduce the dependence of PW Doppler on the operator. This can basically be achieved by developing more sophisticated PW Doppler systems. In this thesis, the principles of operation, design, measurement constrains, and applications of a pulsed ultrasound Doppler velocimeter are described. The elements necessary to construct a directional pulsed Doppler are described and signal processing methods required are stated. In the first section, a brief introduction of pulse wave Doppler unit is stated. No attempt has been made to provide a complete review of Doppler ultrasound or all of the limitations of Doppler flow measurement methods. One of the major accuracay problems with the current Doppler flow measurement techniques is that volume flow is estimated, not measured. The typical Doppler flowmeter measures the average flow velocity along the beam direction at a particular distance from the transducer. Both the vessel size and the measurement angle are unknown. The vessel size is estimated from a knowledge of an average size for the particular vessel being measured or by pulse-echo scanning methods. Since arterial vessel walls are constantly moving (due to the pulsatile flow or blood), it is difficult to determine the position of the vessel at a given time. Also, since the paths for each measurement are different, the two measurements could have different biases due to the different tissue effects. Since each patient has different size blood vessels and it is difficult to measure the Doppler angle with precision, quantitative comparisons between normal and abnormal flows are nearly impossible. -viii- In the second section, basic principles of ultrasonic Doppler system is introduced. Doppler flow systems depend on the interaction of impinging sonic energy with moving blood. It is known that the primary source of ultrasonic scattering is the red blood cells. At low concentrations of red blood cells, less than 10 %, the scat tering is a linear function of hematocrit. At higher concentrations, it was described a more complex relationship using a frequency of 5 MHZ. The Doppler frequency shift associated with the scattering from an interface, or a single cell, can be determined under certain conditions, since the frequency is defined as the rate of change of phase, the frequency of the backscattered energy is changed by a constant proportional to the velocity of the scatterer. This frequencies, is termed the Doppler shift. To easiest way of characterizing the backscattered Doppler signal is via a vector or phaser type of description. This electrical phaser approach will be helpful in understanding the relationships between the various components of the returned signal and its amplitude and modulation format. It will also be helpful in understanding the types of signal detection which can be used to drive the Doppler information. ultrasound travelling through the intervening tissue to the region of the blood vessel is absorbed at a rate proportional to the frequency while the backscattered signal intensity varies with the fourth power of the frequency. Based on this, it would appear that there is some optimal relationship between the ultrasound frequency and the depth of penetration for an optimal signal-to-noise ratio for a Doppler device. Seeking to maximize the signal-to-noise ratio by proper selection of the center frequency calls for a compromise between maximizing the scattering cross section, minimizing attenuation losses, and adjusting, system bandwidth. Recognizing that the wavelength of ultrasound is orders of magnitude greater than the dimensions of the scattering blood cells, the power level of the scattered sound follows the Rayleigh fourth power law assuming single scattering. In biological soft tissues the attenuation coefficient varies linearly with frequency. The rate of attenuation depends on the tissue type and ranges from 0.2 dB/MHz/cm to more than 2 dB/MHz/cm. It has been demonstrated that the performance of pulsed wave Doppler system depends on the average power level transmitted into the vessel of interest. The average power may be from 20 to 300 mW/crn^. The last part of second section is devoted to the theoretical analysis of color Doppler flow mapping. In this analysis, two parameters (mean flow velocity and flow turbulance) are derived by using statistical analysis of returned echoes. -ix- In the third section, the design of pulsed wave Doppler flowmeter is described. There are two levels of information relating to the design, application, and measurement constraints of pulsed Doppler flowmeters. First, the overall principle of operation of a pulsed Doppler blood flowmeter is described. A functional diagram of the complete system required to produce a Doppler signal is illustrated. Various design compromises and error sources are listed in order to facilitate user discrimination of the accuracy and applicability of the obtained blood velocity data. The first iteration in a two-tiered construct has been written to be easily accesible to the clinician desiring to understand the fundamentals and limits of pulsed Doppler system. Mathematical details have been de-emphasized, with the focus on salient features. Second, this section expands on the analytic details of pulsed Doppler designs and their electronic implementation. A pulsed Doppler flowmeter can be analyzed as a group of interacting blocks. A narrow band square wave signal of fundamental frequency f0 is generated by oscillator. This continuous signal is gated to yield a short pulse which is repeated at a lower frequency known as the pulse repetition frequency (hence the name pulsed Doppler). The gated pulse is amplified by a power amplifier so that a piezoelectric transducer can be driven to yield a sufficient ultrasonic signal to penetrate and return from tissue and blood chambers. Reflected and scattered ultrasonic energy is returned to the same piezoelectric transducer from which it is transmitted inducing a corresponding electric. signal which can be sensed from the transducer. This signal, which now includes components which have been shifted in frequency if the reflecting medium was moving, may require amplification by as much as 40 dB (100 times) by a radio-frequency amplifier, since it is usually quite small. This amplified signal is multiplied (mixed) with the original oscillator signal (a synchronous detector) in a process known as demodulation. For directionality of the system, a quadrature phase detector is addet in order to discriminate whether the flow is toward or away from the transducer. Demodulation yields a signal containing sum frequencies : fQ+(fo - *d)» an<3 difference frequencies f0-(fo ± fd)« This signal is sampled by the sample gate at the pulse repetition frequency and bandpass filtered to obtain a signal known as the raw Doppler signal. These signals can siroply be audibly monitored or can be post-processed by a variety of modalities to yield desired analog information readily interpretable by the investigator. In this section, ultrasonic transducer which is suitable for the present system is also constructed and tested. Transducers are used to convert electrical energy into acoustical energy and vice versa. For medical applications, transducers constructed of piezoelectric materials are most commonly used. The design or even selection of a particular transducer must be dictated, ultimately, for a specific applicatoin. The active element in a typical ultrasonic transducer is a. -x- thin plate fabricated from piezoelectric material. Such a plate functions as a resonator in a thickness expander mode for the generation and detection of longitudinal waves. Broadband transducers are used in applications ranging from imaging with short ultrasonic pulses to quantitative measurement of the phase velocity and attenuation over a continuous range of frequencies. Accurate pulse wave Doppler blood flow measurements require a transducer which is specially designed for the particular application. The transducer is shock excited by a short high voltage sinusoidal burst. This will provide a wideband ultrasonic pulse and therefore the best axial resolution and energy level. In the fifth section, errors introduced by the tissue attenuation are investigated. Most current clinical blood flow measurement techniques are based on qualitative comparisons of measurements from the same patient. This is primarily because the variation of the vascular system from patient can be significant. The thicknes and characteristics of the intervening tissue between the skin surface and the flow to be measured affect the accuracy of Doppler frequency estimate. Different tissues have differenet frequency dependent attenuations and will therefore affect the measurement of the mean frequency of the received echo which is used to estimate the Doppler shift. The presence of more tissues between the transducer and the flowing blood in a particular patient may not only increase the variance of the Doppler frequency estimate (because of a lower signal-to-noise ratio due to a longer tissue path length) but may also bias the estimate in an unpredictable way. This is because the received ultrasonic echo spectrum can have a different power spectrum than that of the transmitted signal. In this fifth section, errors that affect the mean Frequency Doppler signal processor are analyzed. The mean frequency Doppler signal processor is of interest for quantitative measurement of blood flow, particularly in peripheral vessels. It is thus desirable that their effects on its performance be quantified where possible. This section analyses the effects of interfering noise, more than one vessel falling within the Doppler sample volume, Doppler filtering, frequency aliasing, and double-sidebanding of the Doppler signal. The analysis applies to either frequency offset or non- offset Doppler systems and a variety of blood velocity distributions is considered. It is shown that, in a number of instances, the errors can be predicted and therefore corrected. Experimental results are presented confirming the theoretical analysis. The relationship between the received Doppler-shif t frequency and the blood velocity depends on : i) the fundamental insonation frequency} ii) the ultrasonic velocity in tissue; and iii) the value of the angle between the ultrasonic beam axis and the flow axis. Of these variables, the first two are readly determined. The angle of insonation noninvasive investigations is not so easily measured -xi- though, and this leads to considerable errors in the estimation of mean blood velocity from the Doppler spectrum. In clinical measurements this error may be flow waveforms in the arterial system in ways which minimize this error. Such characterizations are of particular value in the classification of arterial diseases. The measurement of volume flow requires an accurate measure of the vessel lumen at the point of insonation. Pulsed Doppler systems can provide depth information and, if the angle of insonation is known, can be used to compute the volume flow rate. Even if the vessel/ultrasound angle is known, other criteria must be satisfied if the mean velocity is to be measured accurately. First, the distribution of scatterers must be uniform within the insonated vessel cross section. Second, the ultrasonic beam must be uniform across the entire vessel cross section. Finally, the entire Doppler- shift spectrum must be processed without the introduction of artefacts. The concentration of red cells can be considered to be uniform and, while there is a cell-free zone immediately adjoining the vessel wall, it is too small to introduce significant error. The second condition will not be met if the ultrasonic beam width is less than the vessel diameter as a parabolic flow profile will lead to an overestimation of the mean velocity when the beam covers only the centre of vessel and a corresponding underestimation when the beam is off centre. It is customary to search for the strongest signal (which will be produced when the beam axis intersects the vessel axis). While this can lead to errors in the mean frequency value, the peak frequency is not subject to the same errors. The uniformity of the insonating beam is affected by the progressive loss of intensity with depth caused by absorption and scattering, and this may be significant in large arteries when using high ultrasonic frequencies. The third condition requires that the Doppler-frequency spectrum be processed in its entirety without the introduction of systematic, technique may introduce a systematic error, as is the case with the zero-crossing detector. The Doppler signals may not have an adequate signal-to-noise ratio, which is particularly likely to be the case in a fully developed parabolic and turbulent flow where the scattered signalstrength is distributed over a wide bandwidth. Also, the necessity of high pass filtering to remove high- intensity low-frequency reflected signals, arising from probe movement and vessel wall pulsation, can cause error either by swamping the wanted signal if the filtering is inadequate or by removing low-velocity information if the filtering is excessive. Since the filter requirements are mutually exclusive, there will inevitably be some error from this source. The thesis is ended with a brief conclusion and recommendation