CN1969766A - Mono-calibrating fatty liver B-mode ultrasonic image quantitative analysis method - Google Patents

Abstract

The invention discloses a quantitative analyzing method of Ultrasonic B image of single scaling fatty liver, which comprises the following steps: obtaining each grade of typical fatty liver and Ultrasonic B image echo strength changing coefficient as reference parameter; analyzing detected Ultrasonic B image and echo strength changing coefficient; obtaining interpolation estimated value; comparing echo strength changing coefficient and interpolation estimated value; getting quantitative grading index of Ultrasonic B image of single scaling fatty liver.

Description

Mono-calibrating fatty liver B-mode ultrasonic image quantitative analysis method

Technical field

The invention belongs to the B ultrasonic image analysis technology, be specifically related to a kind of mono-calibrating fatty liver B-mode ultrasonic image quantitative analysis method.

Background technology

The B ultrasonic instrument is by to human body emission ultrasound wave, then according to the echo power of each the layer tissue boundary reflection that receives, carries out pictorial display with different gray levels.The B ultrasonic image has higher clinic value to diagnosing hepatic diseases, especially fatty liver B-mode ultrasonic image, it has with the lesion degree of fatty liver and becomes positively related ultrasonic tissue to levy surely, strengthen as the liver parenchyma front-end echo, liver deep echo is weakened gradually: slight fatty liver shows as liver rear portion luminous point echo and lowers slightly; The moderate fatty liver shows as liver deep layer 1/3 part luminous point echo and obviously lowers; Severe fatty liver shows as liver deep layer 2/3 part luminous point echo and does not show substantially.In addition, the texture of fatty liver B-mode ultrasonic image also can be different than normal liver B ultrasonic image: normal liver B ultrasonic image echo is even, texture is smooth neat, and the reflection of fatty liver B-mode ultrasonic image echo and scattering are bigger, texture more slightly is cloud and more mixed and disorderly, and the quantitative analysis that these ultrasonoscopies are characterized as fatty liver B-mode ultrasonic image provides important evidence.Current fatty liver B-mode ultrasonic image is analyzed mainly is the variation of with the naked eye judging image gray levels (being presented as the echo power on the ultrasonography), but because human eye can only be differentiated 8～16 grades of gray levels usually, can not accurately estimate the slight change of echo intensity, and manual analysis B ultrasonic image relies on doctor's subjective estimation to a great extent, this empirical, perceptual understanding does not have repeatability, can't obtain unified quantitative analysis result.Over past ten years, domestic and international many scholars have carried out the quantitative analysis that echo intensity changes to liver B ultrasonic image, comprise the histogram analysis, liver B ultrasonic image texture analyses of liver B ultrasonic image etc., also produced relevant patent for example " tissue double-frequency ultrasound decay imaging " (number of patent application is 98118262.3), " ultrasound quantity diagnosis system and echological picture quantitative method thereof " (number of patent application is 94111751.0) etc.But since ultrasound wave when in tissue, propagating because of the absorption of reflection, scattering, diffusion and tissue to ultrasonic energy, caused hyperacoustic decay, ultrasonic propagation is must echo far away more just weak more.For the characteristic of correct reflection tissue, eliminate ultrasound wave owing to the far and near factor that produces decay of distance, need on the B ultrasonic instrument, use yield value to regulate, so that the B ultrasonic brightness of image is suitable, and near, far field display brightness uniformity.The probe of different frequency has Different Effects to hyperacoustic echo attenuation in addition, thus during the B ultrasonic imaging because of B ultrasonic instrument degree of aging, tested liver individual variation, need adjust accordingly to reach best B ultrasonic image imaging effect to the B ultrasonic instrument parameter.Because the transmutability of B ultrasonic instrument parameter has caused the unstability and the nonrepeatability of B ultrasonic image imaging effect, so be difficult to the echo data of tested liver B ultrasonic image is carried out unified quantitative analysis.Therefore, for the B ultrasonic image that makes imaging under different B ultrasonic instrument parameters has repeatable analysis result, can utilize the calibration phantom parameter of current B ultrasonic instrument to be calibrated the B ultrasonic image of the tested liver of quantitative analysis on this basis.Existing correlational study achievement patent has " two calibration liver ultrasonic attenuation quantitative analysis tech " (number of patent application is 200310110170.7), " liver dual-calibrating back-scattering B ' type B ultrasonic image quantitative analysis technology " (number of patent application is 200610017335.X) etc.But selected two calibration phantoms for use in these researchs, and required the material of two calibration phantoms must be respectively identical with the acoustic attenuation coefficient of normal liver and typical severe fatty liver, such phantom and B ultrasonic image thereof be difficult to obtain in actual applications.

Summary of the invention

The object of the present invention is to provide a kind of mono-calibrating fatty liver B-mode ultrasonic image quantitative analysis method, this method can be utilized in the practical application calibration phantom that obtain easily and that material that different acoustic attenuation phenomenons is arranged is made under the super parameter of different B, for the B ultrasonic instrument parameter is calibrated, obtain repeatable quantitative analysis results.

A kind of mono-calibrating fatty liver B-mode ultrasonic image quantitative analysis method provided by the invention, its step comprises:

(1) pre-determines the B ultrasonic instrument parameter value that R organizes different range;

(2) will normal, typical slight fatty liver, moderate fatty liver and the liver B ultrasonic image of severe fatty liver and the B ultrasonic image of calibration phantom be divided into the R group, B ultrasonic instrument parameter value during image imaging in every group is identical, and is predetermined a certain group of parameter value in (1);

(3) in every width of cloth B ultrasonic image, select several convex polygons or concave polygon resample area, analyze some echo characteristics parameters of resample area in the B ultrasonic image;

(4) each parameter that step (3) is obtained is carried out normalization, obtains corresponding normalized parameter;

(5), ask the resample area echo intensity variation coefficient of every width of cloth B ultrasonic image according to the normalized parameter that obtains in the step (4);

(6) the echo intensity variation coefficient F of the resample area of the tested liver B ultrasonic image of calculating _Patient

(7) with the calibration phantom B ultrasonic image of tested liver B ultrasonic image under identical B ultrasonic instrument parameter in choose resample area, calculate the echo intensity variation coefficient F of this resample area _Pattern

(8) select and F the echo intensity variation coefficient of the calibration phantom B ultrasonic image that obtains from step (5) _PatternTwo coefficient F of absolute value minimum of difference ^R1 _Pattern, F ^R2 _Pattern, these two coefficients are the r1 in correspondence, the echo intensity variation coefficient of the calibration phantom B ultrasonic image of imaging under the r2 group B ultrasonic instrument parameter, wherein, r1, r2 ∈ 1,2 ..., R};

(9) according to the echo intensity variation coefficient F that calibrates phantom B ultrasonic image ^R1 _Pattern, F ^R2 _PatternAnd F _PatternRelativeness, and r1, the echo intensity variation coefficient of each fractionating fat liver B ultrasonic image of imaging under the r2 group B ultrasonic instrument parameter, interpolation is estimated and the echo intensity variation coefficient of tested liver B ultrasonic image with each fractionating fat liver B ultrasonic image of imaging under the B ultrasonic instrument parameter;

(10) the interpolation estimated value and the F of analysis (9) gained _PatientRelativeness, interpolation is estimated to quantize grading index score to obtain tested fatty liver B-mode ultrasonic image once more.

The inventive method utilizes the one calibration phantom that obtains easily in actual applications (to be called emulating organization ultrasonic body mould again, this calibration phantom has similar acoustic attenuation phenomenon to the human liver under the super instrument parameter of different B) the B ultrasonic image of imaging under various B ultrasonic instrument parameters, for the B ultrasonic instrument parameter is calibrated, thereby get rid of B ultrasonic instrument degree of aging, emitted energy, power amplification, degree of depth gain, the front court suppresses, the multinomial variable factors such as different frequency of back field compensation and probe are to graphical analysis result's influence, and the analysis result of the tested liver B ultrasonic image of acquisition repeatability; On this basis, the present invention observes B ultrasonic pattern accuracy deficiency and the strong weakness of subjectivity at the traditional B hypergraph as human eye in the analytic process again, before analyzing tested liver B ultrasonic image, the back court gray scale difference increases and texture overstriking and mixed and disorderly multiple tissue characterization, calculate corresponding echo intensity variation coefficient, and with itself and imaging normal under identical B ultrasonic instrument parameter with it, typical slight fatty liver, the echo intensity variation coefficient of the B ultrasonic image of the liver B ultrasonic image of moderate fatty liver and severe fatty liver and calibration phantom compares, according to the final quantitative classification index that obtains tested liver B ultrasonic image of the relativeness between them, reach the objectivity and the accuracy that improve in the fatty liver B-mode ultrasonic image analytic process.

Description of drawings

Fig. 1 is the phase I flow chart of the inventive method embodiment;

Fig. 2 is the second stage flow chart of the inventive method embodiment;

Fig. 3 is human liver's B ultrasonic image sampling area schematic of the present invention.

The specific embodiment

The present invention will be further described below in conjunction with the drawings and specific embodiments.

The present invention is divided into the implementation phase of being two: the phase I is the preanalysis stage, and the operating procedure in all these stages only need be carried out once; Second stage is the analysis phase, and the operating procedure in all these stages needs to carry out once in the process of analyzing each tested liver B ultrasonic image.

1, the phase I (stage of preanalysis just) mainly is to obtain each the classification typical case's fatty liver of imaging under the super instrument parameter of different B and the B ultrasonic image echo intensity variation coefficient of calibration phantom in advance, with its exponential reference coefficient of fatty liver quantitative classification as the tested liver B ultrasonic image of analysis, its operating procedure is specially as shown in Figure 1:

(1) pre-determine the B ultrasonic instrument parameter value of R group different range, these B ultrasonic instrument parameters comprise that degree of depth gain, front court suppress, 5 of back field compensations, rate of scanning, scan depths, and the method for specifically determining is: the adjustable extent of q item parameter is carried out Z _q(Z _qCan determine Z by the operator according to the adjustment accuracy of this parameter _qIt is high more to obtain big more then precision) five equilibrium, then the parameter value of every each Along ent of parameter to be made up, every kind of combination just is set at one group of predetermined B ultrasonic instrument parameter value, and is total $R=\underset{q=1}{\overset{5}{\mathrm{\Π}}}{Z}_q$ Group.

(2) will normal, typical slight fatty liver, moderate fatty liver and the liver B ultrasonic image of severe fatty liver and the B ultrasonic image of calibration phantom be divided into the R group, B ultrasonic instrument parameter value during image imaging in every group is identical, and is predetermined a certain group of parameter value in (1).

(3) in every width of cloth B ultrasonic image, choose resample area, analyze five kinds of echo characteristics parameters in this zone, these five kinds of echo characteristics parameters all become positive correlation with the fatty live lesions degree, the average gray that comprises front and back field resample area grey level histogram is poor, energy inverse, entropy and the contrast of rectangular histogram similarity and total resample area gray level co-occurrence matrixes are divided the square inverse, and these parameters are stored.Now the analytic process with the t width of cloth image of r group B ultrasonic image is that example illustrates its concrete steps:

(3.1) as shown in Figure 3, in the t width of cloth image of r group B ultrasonic image, choose resample area.If this image is human liver's a B ultrasonic image, then the polygon of several arbitrary shapes is respectively chosen in optional position and the optional position in the liver parenchyma back court area L in the H of the zone, liver parenchyma front court of its B ultrasonic image, and does not comprise the angiosomes V that is arranged in liver parenchyma in each polygonal region; If this image is the B ultrasonic image of calibration phantom, then in the front court of its whole B ultrasonic image and the optional position of back court respectively choose the polygon of several arbitrary shapes.The zone that this B ultrasonic image front court polygon of choosing is formed is designated as front court resample area A, the zone that the back court polygon is formed is designated as back court resample area B, the zone that all polygons are formed is designated as total resample area C, and then the zone of the correspondence in the t width of cloth image of r group B ultrasonic image is designated as A _t ^r, B _t ^r, C _t ^r

(3.2) resample area of (3.1) is analyzed, calculated and store corresponding front and back field resample area A _t ^r, B _t ^rGrey level histogram average gray difference f _{T, 1} ^r, rectangular histogram similarity f _{T, 2} ^rAnd total resample area C _t ^rThe energy f reciprocal of gray level co-occurrence matrixes _{T, 3} ^r, entropy f _{T, 4} ^rDivide square f reciprocal with contrast _{T, 5} ^r, these CALCULATION OF PARAMETERS methods are as follows:

1. zoning A _t ^r, B _t ^rGrey level histogram h _a, h _bh _a, h _bI element h _a(i) and h _b(i), be respectively regional A _t ^r, B _t ^rMiddle gray level is the ratio of pixel in each zone of i.According to rectangular histogram h _a, h _bCalculate average gray difference f _{T, 1} ^rWith rectangular histogram similarity f _{T, 2} ^r, if the B ultrasonic image gray levels is L ₀, then have:

f _{T, 1} ^r=(M ^A _t-M ^B _t)/(d _t* 3.0) (formula 1)

Wherein ${M^A}_t=\underset{i=1}{\overset{L_0}{\mathrm{\Σ}}}{h}_A\left(i\right)\×i,$ ${M^B}_t=\underset{i=1}{\overset{L_0}{\mathrm{\Σ}}}{h}_B\left(i\right)\×i,$ d _tBe A _t ^rWith B _t ^rThe distance of central point;

${f_{i,2}}^r={\left(\underset{i=1}{\overset{L_0}{\mathrm{\Σ}}}{(h_a\left(i\right)-h_b\left(i\right))}^2\right)}^{1/2}$ (formula 2)

2. zoning C _t ^rGray level co-occurrence matrixes P.Element P (m among the P ₁, m ₂) be specific pixel in zone C ^r _tThe middle probability that occurs, specific pixel is to meeting the following conditions:

A) two gray values of pixel points equal m respectively ₁, m ₂

B) two pixels are at a distance of d pixel, and d gets certain integer of 1～4 usually;

C) line of two pixels and horizontal line angle are θ, and θ gets 0 usually,

π, Certain angle among 2 π.

According to gray level co-occurrence matrixes P calculating energy f reciprocal _{T, 3} ^r, entropy f _{T, 4} ^rDivide square f reciprocal with contrast _{T, 5} ^r, if the B ultrasonic image gray levels is L ₀, then have:

${f_{t,3}}^r={\left(\underset{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">m</figure-callout>}}_1=1}{\overset{L_0}{\mathrm{\&Sigma;}}}\underset{m_2=1}{\overset{L_0}{\mathrm{\&Sigma;}}}{P}^2({\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">m</figure-callout>}}_1,m_2)\right)}^{-1}$ (formula 3)

${f_{t,4}}^r=\underset{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">m</figure-callout>}}_1=1}{\overset{L_0}{\mathrm{\&Sigma;}}}\underset{m_2=1}{\overset{L_0}{\mathrm{\&Sigma;}}}P({\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">m</figure-callout>}}_1,m_2)\log P({\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">m</figure-callout>}}_1,m_2)$ (formula 4)

${f_{t,5}}^r={(-\underset{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">m</figure-callout>}}_1=1}{\overset{L_0}{\mathrm{\&Sigma;}}}\underset{m_2=1}{\overset{L_0}{\mathrm{\&Sigma;}}}\frac{P({\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">m</figure-callout>}}_1,m_2)}{1+{(m_1-m_2)}^2})}^{-1}$ (formula 5)

Other images of r group B ultrasonic image are chosen the resample area line parameter analysis of going forward side by side according to the described method of (3.1)～(3.2), finally obtain parameter f _{T, u} ^r, t=1,2 ... 5, u=1,2 ... 5.

Arrive this, r group B ultrasonic image analysis process finishes.Carry out the analysis of other group B ultrasonic images then after the same method, finish, finally obtain the changes of echo coefficient f of each B ultrasonic image up to R being organized whole analysis of B ultrasonic image _{T, u} ^r, r=1 ... R, t=1,2 ... 5, u=l, 2 ... 5.

(4) each parameter f that (3) are obtained _{T, u} ^r, r=l, 2 ..., R, t=1,2 ..5, u=l, 2 ... 5 carry out normalization, get corresponding normalized parameter

, r=1,2 ..., R, t=1,2 ... 5, u=1,2 ..., 5, the normalization computing formula is:

${{\stackrel{-}{f}}_{t,u}}^r=\frac{{f_{t,u}}^r-{f_u}^{\min }}{{f_u}^{\max }-{f_u}^{\min }}$ (formula 6)

Wherein $f_u^{\max }=\max \left(\right\{{{\mathrm{<figure-callout\; id="1"\; label="f"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">f</figure-callout>}}_{1,u}}^1,{{\mathrm{<figure-callout\; id="1"\; label="f"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">f</figure-callout>}}_{1,u}}^2,...,{{\mathrm{<figure-callout\; id="1"\; label="f"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">f</figure-callout>}}_{1,u}}^r,{f_{2,u}}^1,{f_{2,u}}^2,...,{f_{2,u}}^r,...,{f_{t,u}}^1,{f_{t,u}}^2,...,{f_{t,u}}^r\left\}\right),$

$f_u^{\min }=\min \left(\right\{{{\mathrm{<figure-callout\; id="1"\; label="f"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">f</figure-callout>}}_{1,u}}^1,{{\mathrm{<figure-callout\; id="1"\; label="f"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">f</figure-callout>}}_{1,u}}^2,...,{{\mathrm{<figure-callout\; id="1"\; label="f"\; filenames="A20061012522700111.png"\; state="\{\{state\}\}">f</figure-callout>}}_{1,u}}^r,{f_{2,u}}^1,{f_{2,u}}^2,...,{f_{2,u}}^r,...,{f_{t,u}}^1,{f_{t,u}}^2,...,{f_{t,u}}^r\left\}\right).$

(5) according to the normalized parameter that obtains in (4)

, r=1,2 ..., R, t=1,2 ... 5, u=l, 2 ..., 5, ask the echo intensity variation coefficient of the resample area in every width of cloth B ultrasonic image, then the resample area echo intensity variation coefficient of t width of cloth image is in the r group B ultrasonic image:

$F_t^r=\frac{1}{5}\underset{u=1}{\overset{5}{\mathrm{\&Sigma;}}}{w}_u{{\stackrel{-}{f}}_{r,u}}^r,w_u>0$ (formula 7)

W wherein _uBe the weight of u item echo characteristics parameter, weight w _uSize then represent the positive correlation degree of u item parameter and fatty live lesions, and satisfy $\underset{u=1}{\overset{5}{\mathrm{\&Sigma;}}}\left|w_u\right|=1$ , for example can establish w _u=0.2, u=1,2,3,4,5.

2, second stage is the analysis phase, all need carry out once when analyzing each tested liver B ultrasonic image, the steps include:

(1) in tested liver B ultrasonic image, chooses resample area, and calculate the echo intensity variation coefficient F of resample area according to (3.2), (4) and (5) described method according to phase I (3.1) described method _Patient

(2) with the calibration phantom B ultrasonic image of tested liver B ultrasonic image imaging under identical B ultrasonic instrument parameter in according to choosing resample area with (1) same method, and calculate the echo intensity variation coefficient F of resample area _Pattern

(3) analyze the echo intensity variation coefficient F of the calibration phantom B ultrasonic image of acquisition in advance from the phase I ^r _Pattern, r=1,2 ..., select among the R and F _PatternTwo coefficient F of absolute value minimum of difference ^R1 _Pattern, F ^R2 _Pattern, these two coefficients are at r1, the echo intensity variation coefficient of the calibration phantom B ultrasonic image of imaging under the r2 group B ultrasonic instrument parameter, r1 wherein, r2 ∈ 1,2,3...R.

(4) according to the echo intensity variation coefficient F that calibrates phantom B ultrasonic image ^R1 _Pattern, F ^R2 _PatternAnd F _PatternRelativeness, and r1, the echo intensity variation coefficient of each fractionating fat liver B ultrasonic image of imaging under the r2 group B ultrasonic instrument parameter, echo intensity variation coefficient with tested liver B ultrasonic image each fractionating fat liver B ultrasonic image of imaging under identical B ultrasonic instrument parameter is carried out linear interpolation to be estimated, specific practice is: establish 0-3 level fatty liver B-mode ultrasonic image corresponding to normally, typical slight fatty liver, the liver B ultrasonic image of moderate fatty liver and severe fatty liver then with tested liver B ultrasonic image with the interpolation estimated value of the echo intensity variation coefficient of the k level fatty liver B-mode ultrasonic image of imaging under the B ultrasonic instrument parameter is:

(formula 8)

F wherein ^R1 _{K level fatty liver}, F ^R2 _{K level fatty liver}For at r1, the echo intensity variation coefficient of the k level fatty liver of imaging under the r2 group B ultrasonic instrument parameter.

(5) F of analysis gained _{K level fatty liver}, k=0,1,2,3 and F _PatientRelativeness, obtain tested liver B ultrasonic image quantization grading index score, its computational methods are:

(formula 9)

F wherein _{K1 level fatty liver}And F _{K2 level fatty liver}Be at all F _{K level fatty liver}, k=0 is in 1,2,3, with F _PatientTwo echo intensity variation coefficients of absolute value minimum of difference, and satisfy F _Patient∈ [F _{K1 level fatty liver}, F _{K2 level fatty liver}], K ₁＜K ₂

Wherein a kind of specific embodiment of only having given an example above, for persons skilled in the art, according to content disclosed by the invention, can there be different ways to realize above steps, as B ultrasonic instrument parameter value, echo characteristics parameter and interpolation algorithm the different modes of choosing can be arranged all.

Claims (3)

1, a kind of mono-calibrating fatty liver B-mode ultrasonic image quantitative analysis method, its step comprises:

(1) pre-determines the B ultrasonic instrument parameter value that R organizes different range;

(2) will normal, typical slight fatty liver, moderate fatty liver and the liver B ultrasonic image of severe fatty liver and the B ultrasonic image of calibration phantom be divided into the R group, B ultrasonic instrument parameter value during image imaging in every group is identical, and is predetermined a certain group of parameter value in (1);

(3) in every width of cloth B ultrasonic image, select several convex polygons or concave polygon resample area, analyze some echo characteristics parameters of resample area in the B ultrasonic image;

(4) each parameter that step (3) is obtained is carried out normalization, obtains corresponding normalized parameter;

(5), ask the resample area echo intensity variation coefficient of every width of cloth B ultrasonic image according to the normalized parameter that obtains in the step (4);

(6) the echo intensity variation coefficient F of the resample area of the tested liver B ultrasonic image of calculating _Patient

(7) with the calibration phantom B ultrasonic image of tested liver B ultrasonic image imaging under identical B ultrasonic instrument parameter in choose resample area, calculate the echo intensity variation coefficient F of this resample area _Pattern

(8) select and F the echo intensity variation coefficient of the calibration phantom B ultrasonic image that obtains from step (5) _PatternTwo coefficient F of absolute value minimum of difference ^R1 _Pattern, F ^R2 _Pattern, these two coefficients are the r1 in correspondence, the echo intensity variation coefficient of the calibration phantom B ultrasonic image of imaging under the r2 group B ultrasonic instrument parameter, and r1 wherein, r2 ∈ 1,2 ..., R};

(9) according to the echo intensity variation coefficient F that calibrates phantom B ultrasonic image ^R1 _Pattern, F ^R2 _PatternAnd F _PatternRelativeness, and r1, the echo intensity variation coefficient of each fractionating fat liver B ultrasonic image of imaging under the r2 group B ultrasonic instrument parameter, interpolation is estimated and the echo intensity variation coefficient of tested liver B ultrasonic image with the fatty liver B-mode ultrasonic images at different levels of imaging under the B ultrasonic instrument parameter;

(10) the interpolation estimated value and the F of analysis (9) gained _PatientRelativeness, interpolation is estimated to quantize grading index score to obtain tested fatty liver B-mode ultrasonic image once more.

2, quantitative analysis method according to claim 1 is characterized in that: the B ultrasonic instrument parameter comprises degree of depth gain, front court inhibition, back field compensation, rate of scanning and scan depths in the step (1).

3, quantitative analysis method according to claim 1 and 2 is characterized in that: the echo characteristics parameter of step (3) comprises that front and back field resample area grey level histogram average gray is poor, energy inverse, entropy and the contrast of rectangular histogram similarity and total resample area gray level co-occurrence matrixes are divided the square inverse.

Images