JPS63242226A - Local brain blood flow measuring apparatus - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

Detailed Description of the Invention [Objective of the Invention] (Industrial Application Field) The present invention uses, for example, an X-ray CT apparatus to detect local cerebral blood flow (γ-CBF> The present invention relates to a so-called local cerebral blood flow measurement device that can perform measurements, and particularly to its analysis algorithm.

(Prior Art of the Invention and its Problems) Conventionally, local cerebral blood flow measurement has generally been performed as follows.

A tomographic image is created by scanning a predetermined slice of the head several times over time in parallel with tracer inhalation, and based on the created tomographic image, the tracer concentration in a predetermined brain tissue region is determined over time. The change (hereinafter abbreviated as C1(t)) and the temporal change in the tracer concentration in the artery (hereinafter abbreviated as Ca(t)) are obtained.

Then, based on C1(t) and Ca(t), which are calculated by 1q, F
The following formula (Kety-3chmidt
The cerebral blood flow parameter, ie, λi (distribution coefficient), is determined by the relationship of

Ki (rise rate) and fi (cerebral blood flow) are calculated.

Ci(season− λi −K i fTCa(t) ・exp[−k
1(T-t)] dt...(1] or -Ci(Ding)=ki(λ i -Ca(t)-C1(
t))dt...
・(1°) However, Ci: organization i at time T
Various methods have been attempted to determine r-BF based on the tracer concentration and the above equation (1) or equation (1°). One of them was proposed earlier by the present applicant (Japanese Patent Application Laid-open No. 80-199527) and is called the "Area-Intel method," and the other is U.S. Patent No. 4,610,258.
The rLsQ-2 method disclosed in No. The following will be explained in order. In the following explanation, for convenience, the subscript "1" of each of the parameters λi, ki, and fi is omitted, and the parameters are simply expressed as λ and f.

(Area-InteCl method) Integrating both sides of the above equation (1゛) with respect to t up to OT and setting 1=0 and C(t)=O, C(t) = k (λ f, Ca( t) dt fo
C(t) dt )...(2) After solving equation (2) above for k, we find that λ is actually given in advance as Ca(t) and C(t) is given as discrete data as shown in Figure 3. It is being

In this way, the above method uses the data Ca(t) and C(
Although this method allows analysis using both the rise and fall of t), it has the following problems. ■It is not possible to accurately determine f for λ. (2) Since it is necessary to increase the number of data (number of scans) of C(t) until the falling edge becomes almost O, the number of scans increases.

<LSQ-2 method> This method faithfully applies the least squares method based on the above equation (1), and n − (j=1...n) ...(3) where For C1(Tj), λ and k are determined so as to minimize the tracer concentration in brain tissue local area 1 according to equation (3).

Specifically, from the two equations, we solve for λ and k, and this rL
The SQ-2 method has the problem that it becomes a nonlinear equation because it includes the term ek, and requires a numerical solution such as successive approximation for k, which takes a long processing time.

However, for the "Area-■nteg method", λ,
, the accuracy of f (S/N' of functional image
) is excellent.

The present invention has been made in view of the above circumstances, and provides a local cerebral blood flow measurement device equipped with an analysis device that can accurately and quickly determine various calculation parameters and blood flow volume used in local cerebral blood flow measurement. The purpose is to provide.

[Means of the invention] (Means for solving the problem) Blood flow tissue distribution coefficient λ, rise rate k, and local cerebral blood are calculated based on the arterial blood tracer concentration Ca(t) and the tissue tracer concentration C1(t). In one including an analysis device for determining the flow Mf,
This method is characterized by finding λ and k that minimize the following equation En=.

(Function) Since there is no need to solve nonlinear equations, high-speed processing is possible.

(Example) Next, an example of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of a local cerebral blood flow measuring device according to an embodiment of the present invention. The apparatus 100 of this embodiment is an X-ray C
T device 50. It includes a tracer inhaler section 60 and a data collection control section 20.

The X-ray CT device section 50 has a tail-shaped hole 3 into which the head of a subject 2 lying on the bed top 1 can be inserted, and is connected to an X-ray tube (not shown) that rotates around the head of the subject 1. a gantry 5 that forms a tomographic tail shape on the slice plane 4 by X-ray radiation; and an exposure command device 45 that commands X-ray exposure by controlling, for example, a high voltage applied to an X-ray tube in the gantry 5. , a data collection device 6 that collects data output from a detector (not shown) in the gantry 5 and compiles it into a large number of projection data;
an image reconstruction device 7 that reconstructs a tomographic image of the slice plane 4 based on projection data output from the image reconstruction device 7; , Ca(t) and C1(t
) and an image analysis device 8 (described in detail later) that analyzes the tomographic image constructed by the image reconstruction device 7 and/or λ analyzed by the image analysis device 8, numerical data of f, and funk. and an image display device 9 that displays a national image.

The tracer inhaler section 60 is a closed system in order to reuse the tracer, and includes a mixed gas tank 12 that mixes the tracer supplied from the tracer cylinder 10 and oxygen supplied from the oxygen cylinder 11; an oxygen concentration monitoring device 14 that monitors the oxygen concentration in the mixed gas tank 12 and automatically controls a control valve 13 that adjusts the oxygen supply in order to maintain the oxygen concentration in the mixed gas tank 12 constant; A mask 15 that covers the 0 mouth of the gas mixture, and a one-way check valve for supplying the mixed gas into the mask 15 and preventing the exhaled breath of the subject 2 from getting mixed in, is provided near the mask 15. An intake tube 16 connected between the gas tank 12 and the mask 15; a valve drive device 42 that drives a switching valve 46 for switching between mixed gas and oxygen; a buffer bag 17 for buffering the mixed gas flow of the mask 15; An exhalation tube 119 is disposed near the mask 15 and the mixed gas tank 12 via a carbon dioxide (CO2) adsorption device 18, which will be described later, and passes through the slice surface 4. It is configured to include a carbon dioxide gas suction device 18 that adsorbs and removes carbon dioxide gas.

The data collection control unit 20 controls the X-ray exposure timing via the exposure command device 45 and also controls the switching timing between the mixed gas and oxygen via the valve drive device 42.

Next, it is a block diagram showing an example of the configuration of the image analysis device 8. Reference numeral 21 denotes image data before tracer inhalation (hereinafter referred to as P
22 is a first memory capable of storing a plurality of image data after tracer inhalation (hereinafter referred to as P (t
)) is a second memory capable of storing . 23
is a subtraction unit that performs subtraction between P (t) stored in the second memory 22 and P (0) stored in the first memory 21 . Reference numeral 25 denotes a third storage unit capable of storing the exhaled tracer concentration Ca1r(t) output from the subtraction unit 23 or Ca1r(t) given from the outside via EXTl. This is due to mode switching by the switching means 24. Further, 26 is C1(t) outputted from the attenuator 23: (i=1.2. . .
, N>, and 28 receives the hematocrit value Ht (%) from the outside and stores this l-1
This is an α calculation unit that calculates a conversion count α of the tracer concentration in blood based on t.

The hematocrit value is a value measured from the blood of the subject 2.

30 is α or E calculated by the α calculation unit 28
Ca(t) based on α given externally via XT2
This is a Ca(t) calculation unit that calculates Ca(t). The calculation of this Ca(t) calculation unit 30 is expressed by the following equation.

Ca(t)= a −Ca1r(t) ・
(4) Regarding α, it is determined by the mode switching of the switching means 29 whether to use the calculation result of the α calculating section 28 or to use α given from the outside via EXT2. 31 is the above C1(t), Ca(t).

This is a calculation unit that calculates λ and k based on λ given from EXT3, and the details will be described later. Reference numeral 38 denotes an f calculation unit that calculates the blood flow if based on λ and k output from the calculation unit 31.

Next, the details of the calculating section 31 and its algorithm will be explained for the λ with reference to FIGS. 4 and 5. The calculation unit 31 includes a curve fitting circuit 32. Aj table 33. Bj table 34°λ, kaf calculation circuit 35. Mark Q
It has a deviation calculation circuit 36, and further calculates the Ca(t) by a car 7.
It includes a switch SW1 for selective input to the fitting circuit 32 or the A table. The calculation circuit for the above λ is the above Aj.

In addition to inputting Bj, the above-mentioned C1(t) (=Cj) and λ are input.

The calculation part for the above λ is the following formula % formula % (5) λ and k are determined so as to minimize λ and k.

For this reason, the above Aj and Bj are expressed as the following equations.

Aj = , /TiCa(t)dt (j=1~n>
−(6) Bj=fJCi(t)dt(j=1~
n> ・(71θ For the above λ, the calculation circuit 35 is connected to each circuit 3 as shown in FIG.
It has 5A, 35B, and 35C. Each numerical value AAj, ABj for solving the least squares equation.

Find Acj, ssj, and scj as follows (
(j-1 to j-n) has a circuit.

BCj=ΣBj Cj...
(8) Also, the λ calculation circuit 35B calculates λ using the following equation.

...(9) Further, 35C is a calculation circuit which switches either the λ calculated by the λ calculation circuit 35B or the λ inputted from EXT3 by the switch SW2, and calculates k using the following formula based on it. It is.

...Go) Next, an example of calculation by the calculation unit 31 for λ having the above configuration will be explained.

First, the calculations of Aj and Bj shown in equations (6) and (7) above are calculated using the trapezoidal formula when the above Ca(t) and C1(t) are discrete data, but in order to improve accuracy, Use higher-order integrals.

Regarding the handling of Ca(t), the following two methods are used.

(2) When approximating Ca(t) with an exponential function (curve fitting), the curve fitting circuit 32 is used.

■ Use Ca(t) as discrete data as is.

These selections can be made by the switch SW1.

The above method (2) is effective when the number of sampling points of Ca(t) is small or when noise is large, but it cannot be applied to an arbitrary curve.

On the other hand, the above method (2) requires a certain number of sampling points, but can be used for any form of Ca(t).

In addition to the exponential function, there is a gamma function as a fitting function in the above ■.Which one to use depends on how the tracer is injected into the arterial blood during the test.For the inhalation method, the exponential function is used for the intravenous injection method. In this case, the gamma function is used. Furthermore, the types of functions used for fitting are not limited to these. The specific form of the function is as follows.

(1) Exponential function = O: (t<0) Ca(t)=K (1-exD(-A-t)) : (
0≦℃≦Tl ＞=K (1-exp(-A-T1))
exp[-A-(t-TJanuary: (1 day 1 Kut) However, the parameter curves for A and Tt are as shown in Fig. 3 Ca(t).

(2) Gamma function Ca(t) = 0: (t<0) = K-t'exp(-t/β): (°C≧0) However,
The α and β parameter curves have a mountain shape.

Note that each of the above parameters is a value determined by fitting.

The value of λ is selected by the switch SW2 as to whether it is given externally or calculated internally. When λ is substituted with a constant, k is obtained by substituting the equation derived from =O into 8En/8 by the circuit 35G into the equation solved for k.

The standard deviation measurement circuit 36 calculates the standard deviation r using the least squares method using the following equation.

[Effect of the invention] According to the present invention described in detail above, the following effects can be obtained.

(1) Highly accurate λ for the data of arterial blood tracer concentration Ca(t) and tissue tracer concentration C1(t) measured over the rise and fall of Xe gas inhalation for a short period of time (for example, 4 to 6 minutes) , we were able to measure f.

In other words, when data obtained by adding noise to C1(t) was analyzed, the average value of the analysis parameters (λ, , f) was close to the true value, and the S/N ratio was good. In other words, accurate measurement can be performed even with a small number of samplings (scans) of C1(t). This also applies when the substitution method is used.

(2) Although it maintains a certain degree of accuracy, it does not solve nonlinear equations like in the past, so it can create functional images at high speed.

(You can select either a mode that uses 31Ca(t) as it is as the original sampling data or a mode that uses it after fitting it with an exponential function or gamma function depending on the application.Especially when using the exhalation tube scan method. The fitting mode is effective because the number of samples of Ca(t) is not increased.

Furthermore, although it is difficult to reduce the number of samples in a mode that uses only original sampling data, it has the advantage that it can be applied to any card.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 is a block diagram showing the configuration of an image analysis device, and FIG. 3 is a characteristic of Ca(t) and C1(t) sieved by the device. 4 is a block diagram showing the configuration of λ in FIG. 2,
FIG. 5 is a block diagram showing the configuration of the calculation circuit at λ in FIG. 8...Analysis device, 31...λ, Calculation department agent Patent attorney Nori Chika Ken Yudo Ogo
Norio 1st figure 4th funeral figure

Claims (3)

[Claims] (1) Based on the arterial blood tracer concentration Ca(t) and tissue tracer concentration Ci(t) of the subject, the blood flow tissue distribution coefficient λ
In a local cerebral blood flow measurement device that has an analysis device that calculates the rise rate k and calculates the local cerebral blood flow from this, the analysis device uses the following formula ▲ There are mathematical formulas, chemical formulas, tables, etc. ▼ Here, Ci (Tj) is A local cerebral blood flow measuring device characterized by comprising an analysis algorithm for determining the λ and k that minimize the value En of the tracer concentration in the tissue locality i. (2) A patent claim in which the analysis device has a function that allows selection between a mode of fitting an exponential function or a gamma function to a time density curve representing Ca(t) and a mode of using sampling data of Ca(t) as is. The local cerebral blood flow measurement device according to item 1. (3) The local cerebral blood flow measurement according to claim 1, wherein the analysis device has a function of selecting between a substitution mode in which the f is obtained by arbitrarily substituting the λ, and a mode in which the f is obtained by calculation. Device.