CA1218448A - Method and device for remote tissue identification by statistical modeling and hypothesis testing of echo ultrasound signals - Google Patents
Method and device for remote tissue identification by statistical modeling and hypothesis testing of echo ultrasound signalsInfo
- Publication number
- CA1218448A CA1218448A CA000450264A CA450264A CA1218448A CA 1218448 A CA1218448 A CA 1218448A CA 000450264 A CA000450264 A CA 000450264A CA 450264 A CA450264 A CA 450264A CA 1218448 A CA1218448 A CA 1218448A
- Authority
- CA
- Canada
- Prior art keywords
- tissue
- sample
- tissue type
- values
- autoregressive
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S15/00—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems
- G01S15/88—Sonar systems specially adapted for specific applications
- G01S15/89—Sonar systems specially adapted for specific applications for mapping or imaging
- G01S15/8906—Short-range imaging systems; Acoustic microscope systems using pulse-echo techniques
- G01S15/8977—Short-range imaging systems; Acoustic microscope systems using pulse-echo techniques using special techniques for image reconstruction, e.g. FFT, geometrical transformations, spatial deconvolution, time deconvolution
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/52—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S15/00
- G01S7/52017—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S15/00 particularly adapted to short-range imaging
- G01S7/52023—Details of receivers
- G01S7/52036—Details of receivers using analysis of echo signal for target characterisation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/02—Indexing codes associated with the analysed material
- G01N2291/024—Mixtures
- G01N2291/02475—Tissue characterisation
Abstract
Abstract:
"Method and device for remote tissue identification by statistical modeling and hypothesis testing of echo ultrasound signals".
For remote identification of tissue types the scattering of ultrasound energy from living tissue is modeled as an autoregressive or autoregressive moving average random process. Autoregressive or autoregressive moving average models of candidate tissue types are generated from pulse-echo data that is known to come from that particular tissue type. Kalman prediction error filters (220) are used for each candidate tissue type to generate estimates of the probability lnp(Zn/Hi) that an unknown pulse echo signal belongs to the class generated by that tissue type (i). Unknown pulse-echo signals are filtered in a specific Kalman filter (220) to test the hypothesis that the unknown signal belongs to the class associated with that particular Kalman filter.
"Method and device for remote tissue identification by statistical modeling and hypothesis testing of echo ultrasound signals".
For remote identification of tissue types the scattering of ultrasound energy from living tissue is modeled as an autoregressive or autoregressive moving average random process. Autoregressive or autoregressive moving average models of candidate tissue types are generated from pulse-echo data that is known to come from that particular tissue type. Kalman prediction error filters (220) are used for each candidate tissue type to generate estimates of the probability lnp(Zn/Hi) that an unknown pulse echo signal belongs to the class generated by that tissue type (i). Unknown pulse-echo signals are filtered in a specific Kalman filter (220) to test the hypothesis that the unknown signal belongs to the class associated with that particular Kalman filter.
Description
12~8448 PH~ 21152 Field of the invention The invention relates to a method for remote identification of a tissue using ultrasound, comprising the steps of:
directing ultrasound energy into an unknown tissue;
measuring values of a characteristic of echoes of the ultra-sound, which are scattered fr~m the ~ulknown tissue.
The invention also relates to a device for remote identi-fication of a tissue comprising:
means for directing ultrasound energy into an unkno~n tissue;
means for detecting a characteristic of echoes of the ultra-sound which are scattered from the unkncwn tissue. The invention is particularly useful in connection with apparatus and methods for the diagnostic imaging of human and animal body struct~res.
Background of the invention Methods and apparatus which utilize ultrasound energy for the diagnostic imaging of body structures are well kncwn. Typically, ultrasound energy is directed into and scattered from body tissues.
m e amplitude of the resultant echoes is detected and displayed to form an image which characterizes the scattering structures. Virtu-ally all commercial medical ultrasound imaging systems form images from the envelope of echoes which are reflected from an interface of tissue types which have different acoustic impedances. me images are, therefore, useful to delineate the outlines of various bcdy organs and lesions. ~owever, because the image makes no use of the phase infor-mation in the echo signals, it is not generally possible to identify tissue types.
The prior art teaches a number of ultrasound systems which are capable of identifying tissues having specific characteristics.
U.S. Patent 4,389,893 by OPHIR and MAKLAD is typical of a class of apparatus which attempts to characterize tissue types by measuring local ultrasonic attenuation. Likewise, U.S. Patent 4,270,546 to PERILHOU and COURSANT describes apparatus for identifving the direc-tional characteristics of tissue structures.
~Z18448 PIL~ 21152 2 12-10-1983 It is an object of the invention to provide a method and a device for tissue characterization that are capable to differentiate and identify the tissues of the various body organs as well as the patholo~y of healthy and diseased tissues to provide a remote, noninvasive biopsy-Summary of the invention The method according to the invention is characterized in that it further comprises the steps of:
accumulat.ing a statistically significant sample of the measured values of the characteristic;
filtering the accu~ulated sample to determine, for each of a plurality of possible tissue types, a signal which is a measure of the likelihood that the accumulated sample oE measured values was generated lS by a mathematical mode1. which characterizes the tissue type;
combining the filtered signals using a predetermined logicaldecision function to choose one of the models which most likely produced the sample; and assigning the tissue type characterized by the chosen model as 20 the i~entity of the unknown tissue The device according to the invention is characterized in that it further comprises:
means for accumulating a statistically significant sample of signals which represent values of the detected characteristic;
2s a plurality of filter means, each filter means being associated with a possible tissue type, which function to filter the acc~mulated sample to extract a signalwhich is ameasure of the likelihood that the accumulated ~ample was generated by a mathematical model which characterizes the associated tissue type; and means for combining. the signals extracted by the filters,which use a predetermined logical decision function to select the extracted signal which is associated with the model which most likely produced the sample and assign the tissue type associated with the filter producing the chosen signal as the identity of the unknown tissue.
We have determined that ultrasonic pulse-echo data from different tissue types or pathologies exhibit diffentiable statistical regularities.
Tissue types or pathologies are classified by utilizing known modeling and hypothesis testing statistical techniques with a dedicated or general PH~ 21152 3 purpose digital o~mputer.
Samples of a pulse-echo record from unknown tissue are treated as a vector x = [xl, x2, ..., x ]. We hypothesize that this record was generated by one of a finite set of tissue types. HR
denotes the hypothesis that the data vector was generated by tissue type "R"; R = l, 2, ..., K. We then calculate the conditional pro-bability density functions p(x¦HR) (the probabilit~ density of data vector x, given that the pulse-echo record was generated by tissue type "R") using prior knowledge of the statistical properties of the pulse-echo records associated with each tissue type. m e statistical properties associated with the tissue type are determined from statis-tical models based on sets of training data obtained from known tissue types. The values of the conditional probability for each possible tissue type represent likelihood functions which are logically oom-pared and a predetermined decision rule is applied to select the tissuetype which most probably generated the data ~ector.
The likelihood functions can be computed recursively from the data samples; that is the data samples can be processed one at a time, as they are received, and the likelihood functions updated with each new data sample. Recursive processing permits real time (on-line) tis-sue classification using dedicated microoomputer-based signal pro-cessors and allows oontinuous processing of new data until one tissue type hypothesis dominates under the decision rule.
Hypothesis testing -techniques have been widely and effec-tively used in many signal processing applications of the prior art.A general discussion of these techniques and their application to specific prior art problems is set forth in the text Uncertain Dynamic Systems by Fred C. Schweppe, Prentice Hall, Inc., 1973. The success of hypothesis testing methods is, hcwever, critically dependent llpon the validity of the statistical models of the underlying physical pro-cess which are incorporated in the hypothesis testing algorithm.
We have determined that hypothesis testing is a suitable method for tissue classification from ultrasound pulse-echo data if a sequence of pulse-echo samples is modeled as an autoregressive process;
that is if the samples satisfy the equation Xn - alXn-l + a2Xn-2 + ''' + aLXn-L + Wn where Wn is an uno~rrelated random sequence (white noise) and the ~21844~3 autogressi~e parameters [al, a2~ -., aL] characterize the particular process. Estimates of these parameters can be obtained from "training sets" of pulse-echo data which are generated by known tissue types.
Thus, each tissue type is characterized by a set of autoregressive parameters. Using the autoregressive parameters for each tissue type we construct an algorithm for recursively computing the likelihood fwlction for any given pulse-echo data sequence.
The autoregressive parameters can be obtained from the train-ing sets of data by either of tw~ kncwn methods. In the first method, t~.e data are used to generate an estimate of the autocorrelation func-tion using time averages. m e estimate of the autocorrelation func-tion is then used to generate the autoregressive parameters using the Durbin-Levinson algorithm (or some variation of the Durbin-Levinson algorithm).
In the seocnd method, the autoregressive parameters are calculated directly from the data training sets using the Burg algor-ithm (or one of the many variations of the Burg algorithm).
Both of these methods, as well as their many variations, are described in Non-linear Methods of Spectral Analysis, S. Haykin ~ (Editor), Springer-Verlag, 1979.
m e sequence of tissue pulse-echo samples may, if desired, also be modeled as an autoregressive moving average process.
Kalman filters can be utilized for the recursive computation of the likelihood functions. m is technique is described in the text Digital and Ka~man Filtering by S.M. Bozic, John Wiley and Sons, 1979.
Brief description of the drawings The invention may be best understood by reference to the accompanying drawings in which:
Figure l schematically illustrates apparatus for classifying tissue in accordance with the invention;
Figure 2 illustrates a digital filter for extracting a like-lihood function from a pulse-echo data vector;
Figure 3 is a flow chart for a preferred embodiment of a tissue classification algorith~;
Figures 4 through 6 are plots of the logarithms of likelihcod functions which were recursively computed from echo samples generated in ~2~8448 sponge samples; ~`ld Figures 7 through 9 are plots of the logarithms of likelihood functions which were recursively computed from ultrasound echo samples derived, in vivo, from human liver, pancreas and spleen tissue.
Brief description of the appendix The APPENDIX is a listing of Fortran computer programs which implenent a preferred en~odinent of the invention.
10 Description of a preferred em~odin~nt Figure 1 is a preferred en~odiment of apparatus for tissue characterization in aco~rdance with the invention. A transmitter 10 activates an ultrasound transducer 20 via a TR switch 30 to direct pulses of ultrasound enera,y into a test body 40 which may, for example, be a lS human a~don~n. Ultrasound energy from the transn~itter is scattered by ~1y tissues including unknown tissues in a region 50. Portions of the scattered energy are returned as echoes to the transducer 20 and are directed via the TR switch into a receiver 60.
The receiver detects and processes the echo signals in a con-20 ventional manner to generate video signals which are utilized to aeneratean image on a display unit 70. The displayed image may, for example, be a conventional B-scan image of the internal structure of the ~ody 40.
The ampified radio frequency echo signals form the receiver 60 are also transmitted via line 80 to a sampling circuit 90. A time gate 25 100 is operatively connected to the sampling circuit 90 in a manner which causes the sampler to extract a series of periodic samples of the values of received echo signals which were scattered from tissue in the unknown region 50. The operation of the time gate may, for example, be controlled in a known manner by means of a cursor which is manually set 30 over a given region of the display 70 or may be triggered by high amplitude echoes which are scattered form the edges of the region 50.
Signals representing the amplitude of the series of saTnples derived from the sampler 90 are digitized in an analog-to-digital converter 110.
The series of digitized signals are then applied to a parallel chain of 35 digital likelihood function filters 120a through 120n (nore particularly descri~ed below). Each filter ca]culates the value of a likelihood function which represents the conditional probability that the series of san~ple signals ~ere produced in a given tissue type. The ouput signaLs ~8448 from the parallel filters are fed to decision function logic 130, which using known algorithms, assigns one of the possible tissue types to the unkr,own tissue represented by the s 9 1e vector.
If the a priori probabilities of occurrence of the various tissue types are equal (or unknown) a preferred em~odiment of the decision function logic assigns the unknown tissue as the tissue type associated with the likelihood function filter which produces the largest output signal. The decision function logic may, however, be modified in a known manner to choose the lik.elihood function having the largest aposteriori lO probability if the various tissue types have different a priori probabilities of occurrence or to minimize the expected penalty if there is a large penatly attached to misidentification errors of a specific type.
The ouput signal from the decision function logic 130 is routed to the display 70 where it is used to identify the tissue type of the 5 unkno~Jn region; for ex 9 1e by display of a message associated with a cursor or by assigning a particlar color or brightness level to pixels in an image of the region where the echo sam~les originated.
In an alternate em~odiment of the invention the output signals of the likelihood function filters may be utilized (either directly or 20 after processing in decision function logic) to modulate the color or intensity of the associated region in the display. In the latter case the saturation of color in the associated region could provide an inclication of the confidence level of the tissue characterization for that region.
Figure 2 illustrates the typical filter for calculating the logarithm of the likelihood function for a given hypothesis ~. The series of periodic data s 9 1es Z(nT) is serially fed to the input of a delay stage 200 having a delay time T and to the positive input of a subtractor 230.
The output of the delay stage 200 is fed through a one-step Ka~1lan 30 predictor error filter which characterizes the hypothesis HR. The output of the Kalman filter 220 is fed to the negative input of the subtractor 230.
The output of the subtractor thus represents the difference between the actual value and the predicted value for each component of the data vector:
that is the prediction error of the Kalman filter. The prediction error 35 is processed in a weighted squarer 240 and a bias signal is added to the output of the squarer in an adder stage 250. The signals from the output of the adder 250 are summed in an accumulator 260 whose output represents a likelihood function for the hypothesis which is modeled by the Kalman 121~34~8 ~A 21152 7 12-10-1983 filter.
The a-ttached Appendix comprises aFortran computer program ~hich represents the preferred embodiment for generating likelihood functions from data sample vectors. The scattering in the various tissue types are modeled as time invariant, autoregressive processes. The prediction error, ~A~ich is called the innovations representations of the process, is assumed to be white and Gaussian. The program further functions to calculate the coefficients which characterize the autoregressive model for each tissue tvpe from data vector samples which are obtained from lO echoes of known tissue regions.
In the appendix: the program TCP generates an autoregressive model for each of three classes of training sample data vectors (A-lines) and tests the models on other samples from tissue in the same classes;
the subroutine PRK obtains the lattice predictive m~del from 15 data in program TCP;
the subroutine GETTCD reads data from a disc and normalizes it to zero mean and unit variance;
the subroutine PRERK, using the lattice model, obtains the Innovations process of the input data;
the subroutine APSPRB calculates a posteriori probabilities of one of three equally probably hy~otheses using relative log likelihood functions;
Figure 3 is a flow chart of the processor used to calculate to likelihood functions by the program of the Appendix.
Figure 3 represents a typical calculation of a log likelihood function ln p(Yn/Hi) for a hypothesis Hi. In the first block 300 the innovations model subroutine PRERK is used to calculate the prediction error for each data point l...n The second block 340 corresponds to the weighted squarer 240 of Figure 2 and the addition circle 350 corresponds 30 to the bias addition in block 250 of Figure 2. The ouput of the addition circle 350 is accumulated over all data points in the s~ner block 360.
The latter three operations are implemented in the main body of the progrc~m following the comments at the top of the third page of the Appendix. The log likelihood functions are the compared using a decision rule 330 to 35 obtain an identification of the tissue type.
Example 1 Pulse echo data was collected from samples of a natural sponge lZ18448 Pl~ 21152 8 12-10-1983 and from t~ different types of synthetic sponge using a 19mm diameter, 3.5 ~lz transducer in a water tank and a Philips 5580 s-Scanner. The output of the scanner TGC amplifier was fed to a Biomation 8100 Transient ~ecorder (20 ~lHz sample rate, 5 bits/sample) which was interfaced to a 5 D~C 11/~0 nuni-computer. The pulse-echo records from each sponge sample were divided into two sets; a training set and a test set. Parameters for the autoregressive model were generated, using the method set forth in the ~ppendix, for each of the test samples.In this way autoregressive mc~els ~.~ere generated for each sponge type. Data vectors were then obtained from echoes which originated in the three test samples. Figure 4 illustrates the values of the likelihood functions ln p(Z~ Hi) computed for a data vectorderived from sponge sample 1 using the computed hypotheses based on mxlels from the teaching samples of sponge types 1, 2 and 3. As illustrated, after a reasonable number of samples (n), the likelihood lS function associated with hypothesis 1 has the hio,hest value thereby correctly iclentifying the sponge sample as type 1. Figures 5 and 6 illustrate similar computations for test samples of sponge types 2 and 3. In all cases the apparatus correctly identified the test sponge sample.
For comparison the power spectra of the pulse-echo records of the 20 three sponge types were computed and were found to be essentially indistinguishable.
Example 2 Teaching data and test data vectors ~iere obtained from in vivo 25pulse-echo data from a human liver (L), spleen (S) and pacreas ~P) and the procedures utilized with the sponge data were repeated. The computed likelihood functions are illustrated for liver test data vectors in Figure 7, for pancreas test data vectors in Figure 8 and for spleen test data vectors in Figure 9. The filters correctly identified liver and pancreas 30pulse-echoes with a high degree of reliability. The algorithm correctly differentiat~d spleen data from liver but with substantially more difficul-ty and lower confidence than the other samples.
The preferred em~odiment of the present invention characterizes tissue types on a basis of the statistical characteristics of samples of the 35amplitude of ultarsonic pulse-echoes. The statistical characteristics of other measured characteristics of pulse-echo data may however also be utilized in the computational algorithm. For example, the instantaneous frec~uency or phase of -the reflected signal might be sampled in place of or in conjuction with the amplitude of the signal.
- APPENDIX lZ1844~3 TPE T~ FORTRAN
'~****t.~***********************~**********************~*****~************
C PROGRAM TCP ~, C PR06RAM GENERATES AR MODEL FOR EACH THREE CLASSES OF TRAINI~5 C SAMPLE A-LINES (REAL OR SYNTHETICl AN~' TEST THE MODEI_S ON OTHER
r SAMPLE A-LINES FROM THE SAME CLASSES
C
C
C************************~********************************************
C ..
- IMPLICIT REAL*~ ~A-H,O~Z) REAL*~ ~S~6('JD4$),R~<(51,3J,RY~T~5lJ,ERROR(51,3j,ERRT~51~,RINN~D4~) ~ ,RKS~51,3),AS~51,3) - REAL XLL(~n4~,3),XPLT~10'4),YPLT~lD"4),RES(~D4 INTEGER IOR(3),IOR5~3),ISTPT~3 .
~EFINE FILE 1'~36,~D4~U~lV4),13(36,~D4~,U,IV5),14~36,~D4~,U~IV~
~ ,15~36,~04~,U,IV7),16~36,~D43,U,IV~1,17~36,'04~,U,IV9) C SPECIFICATIONS
C .
5D ~JRlTE(6,51) 51 FORMAT~' TYPE: FIRST RECOR~) ANb # OF RECORDS FOR TRAINING'~
' THE SAME FOR THE RECOR~1S TO BE TESTE~'/
' START PTS. FOR EAC~ CLASS'J - -' # OF SAMPLESJRE50R~'f & ~ MAX ORC)ER OF THE AR ~O[)EL'J
~X~'OPTlOtl~ ISR bATA. ~=SYN DATA') REA~)~5,*) NFTR,NTR,NFTS,NTS,~I5TPT~I),I=i,3),NSAM,MOR,ISOPT
IF~MlND~NFTR~NTR.NFTS,NTS).LT.l .OR. MAXOiNFTR.NFTS).GT.l .OR. MAXO~NFTR+NTR,NFTS+NTS).GT.13 .OR. MOR.GT.50 .Ofi.
ISOPT.LT.l .OR. ISOPT.GT.~ .OR. NSAM.LT.~ .OR. NSAM.GT.'~04~) 50 ~0 59 I=1,3 IF~ISTPT(I).LT.l .OR. ISTPT~I)+NSAM.GT.~D4q) GOTO 5D
C
C IF THE OPTION IS SYN. ~ATA
IF~ISOPT oEQ~ 1) GOTO 95 ~0 WRITE~6,Bl) ~1 FORMAT~J' TYPE:'J
' OR~ER OF AR MODEL ~M~ FOR CLASS 1 & ~I) I=l,M'J
& ' THE SAMc FOR CLASS ~ & 3, PLEASE ONE CLASSJLIN') DO ~5 I=1,3 READ~5,*) M,~R~S~J,I),J=l,M) IORS~I) = M
~5 CO~ U-C CO JERT A-CQEFS. TO ~-COEFS. eY sue. ~TOA
~0 8~ I=1,3 -~
CALL I~TOA~R~S(l,II,AS(1,I),IORS~
aa CONTINUE
C
C OeTAIN K-COEFS. FROM TRAl~IN5 ~ATA
C
q5 ~0 lOq ISP=l,~
CALL PRt~ISP,NFTh,NTR,ISTPT(ISP),NSAM,MOR,R~l,ISP)~
& ERROR(l,ISP),~,ISOPT,AStl,ISP3,IORS(ISP),bSIG~
CALL PRli(ISP,NFTS,NTS,ISTPT(ISPJ,NSAM,MOR,RKT,ERRT,., ~ ISOPT,AS~l,ISP),IORS(ISP),~SI6) lD9 CONTINUE - -C
C HYPOTHESIS TESTING
C ' :' 150 WRITE(~,151) 151 FORMAT(/ TYPE: THE ORDER OF THE AR MO~EL FOR CLASS 1 A-LINES i & ~ THE SAME FOR CLASS 2 JaY., THE SAME FOR CLASS 3 J
~X, THE REbUCTION FACTOR & MAX ~ OF POINTS FOR PLOTS f & ~X, 1 = E~UAL ~ARIANC S 2 = NOT 1 ) ~ - -REA~(~,*) (IOR(I~,I=l,~),IR~C,MNPT,IEQVAR
~0 153 I=1,3 IF(IOR(I).LT.D .OR. IOR(I~.GT.50) GOTO 15D
IORMAX = MAXO(IOR(l),IOR(,~),IOR(3)~
IF~iR~C.LT.1 .OR. MNPT.LT.2~ 60T0 150 NPT = NSAMJIROC
IF(NPT.GT.MNPT~ NPT = MNPT
C
~0 299 I=l,l, C
IR = I
DO 2~9 ISP=1,3 ; IF(ISOPT.EQ.1) CALL GETTC[)(r)SIG,1,204~,IR,ISP) ISEE~ = IR*3 + ISP - 1 IF(ISOPT.EQ~2~ CALL GET[)R~AS(l,ISP),IORS(ISP),l.,[)SIG,2n4~, & IScE~) C ' .. "
lF(I.GT.1) GOTO 175 DO 170 J=l,NSAM
YPLT(~) = DSIG(J~ -Y.PLT(J~ = J
i70 CONTINU~
C WRITE(b,172~ ISP -t 172 FORMAT(JJT3D, NORMALIZE~) A-LINE FRO~ CLASS ,I4 C CALL PLOT(XPLT,YPLT,300~
., ~o 175- ` ~)0 ,_~ IH=1,3 lZ18448 C ( ERATE THE INNO~ATIOt`lS PROCES5 e.Y SUB. PRER~
C - .
CALL PRERI-~DSJG, 048,RI~l,IH:~,lO~IH-J,RI~`JN:~
C CALCULATE THE L06-LIKELIHOOD FLItlCTIONS SEQUEtlTIALY
C
~AR = EhROF(lOR(JH),IH) IF~IE~AR~Eo.lJ VAR = 1.
PIAS = -.5*t)LOG(VAR) Y.T = O.
. DO 211 J=l, D4 IND = J
lF(IN[) .LE. IORMAXI GOTO 20q . XT = Y.T - .5*RIGII`l IN~J** JVAP.
XLL(J,IHJ = XT + J*BlAS
RES~J) = RINN~IN[;) 211 COtlTlNUE
I~JR = (lSP - 1)*1~ + I
IFILE = 11 + Il~
WRITE(lFILE IWR) RES
I~R = ~I-1)*3 + IH
IFILE = 14 ~ lSP
WRITE~I~ILE I~lRJ ~XLL~J,I~),J=1,204~J
~2q COG!TINUE
C
C
25q CONTINUE
29q CONTINUE
STOP
END
C
C***********************************************************************
C
SUBROUTINE PR~ISP,NF,Gl,IST,NS,MOR,Rh,ERROR,ICl,lC ,AS,IORS,DSIG) C
C SUe.ROUTINE OBTAINS THE LATTICE PREDICTIVE MODEL FROM DATA
C
IMPLICIT REAL*~ (A-H,O-Z~
REAL*~ PHI(51,51),PHIO(51,51),R~(51:1,TK~51,1,J,SCR~51),DSIG(204a) & ,A(51),ERROR(51J.AS(IORS) REAL SPECT(1.~9j,5PAYIS(1 9~ -C
O ~ J=1,1.9 .~
SPECT(I) = O. - _ -3 CONTINUE -. -~0 9 J=1,51 . ..
~0 Y ~ ,51 . - ~ONT I iNUE ~218448 O 3~ I=1,N
IR = NF + I -IF(IC..EQ.1~ CALL GETTC[)(D~I6,JST,NS,IR,ISP~
ISEE~) = IR*3 + ISP - 1 IF(IC~.E~. -.) CALL GETDR AS,IOR5,1.,[~5I6,NS,ISEE[~J
C
CALL PR[~TMP~SPECT,DSI~,NSJ
C
C COMPUTE THE COVARIANCE MATRIX
C
CALL DCOVR1(DSIG,NS,MOR,PHI~51) MORP1 = MOR + I
DO 19 J=-1,MORP1 b O i 9 K=i~MORP1 PHID(K,J) = PHIO(K,J~ + PHI(K,J) 1 q CONTINUE
C
DO ~9 M=1,MOR
CALL DCLHRM(PHI,51,M,A,TK(M,I),TERR,SCR) 2q CONTINUE
3q CONTINUE
C
D 0 4 ~ J=1,1 9 SPECT(J) = SPECT(JJJFLOAT(N) SPAY.IS f Ji = FLOAT(J-i3Jl a.
4~ CONTINUE
C
DO 4q M=1,MOR
CALL DCLHRM(FHID,51,M,A,RK(M),ERROR~M),SC~J
ERROR(M) = ERROR~M)/((tlS - M)*N*~) -C
C PRINT THE K CHART
C
DO i~ M=1,MOR
AVK = O.bO
- DO ~q I=1,N
AVK = AVK + TK(M,I) 8q CONTINUE
AVK = AVKJN
WRITE(6,~1) M~(T~(M,I~,I=1,N),AV~,R~(M~,ERROR~M) ~1 FORMhT~I3,15F~.6) 1 ~q CONTINUE ~ -C CALL PLOt(SPAXIS,SPECT,1 ~) RETURN -EN~
C
/~2 .
.,......................... - - 18448 . - sua~ouT~ ETTC~?fDSIG,IsT,~ls,I~ sP~ ~2 C ROUTINE READS THE [)ATA FRO~l [)ISC, THE NORMALIZE THEM TO
C ZERO-ME~N ANC` UNTT-VARIANCE.
C
REAL*B ~5IG~S~,SQ,SM,[TMP
INTEGER* ID~2D4~i r ~EFINE FILE7tl-,40~L,I~ 10(1 ~40~L~IV~ ,40~6~L~IV3 C
J = ISP ~ ~
READ(J IR~ ID
SM = 0.[)0 SO = O.DO
MSZ = 55 MSZ2 = MSZi2 00 lq J=l~NS
I = J + IST - 1 IF(J.NE.13 GOTO $5 DO 11 L~ MSZ~
L = L2 + IST - 1 SM = SM + ID(L) SQ = SQ + IDtL~*I~)(L) IA~[) = I + MSZ.~
I5BT = I - M5Z - 1 IF(IAD~.GT.NS+IST-l) GOTO 16 SM = SM + IC)~IADD) SQ = SQ + I~(IADC))*I[)(lADD~
lb IF(IS~T .LT. IST) GOTO 17 SM = SM - I[)(ISeT) S~ = SO - lD~ISBT)*I~tISBT) 17 FMZ = MSZ
IF(J.LE.MSZ~I FMZ = MSZ2 + J
IF(NS-J .LT. MSZ2) FMZ = MSZ ~ NS - J + 1 bSIG(J) = ID(I) - SMZFMZ
DTMP = ~S~RT(S~JFMZ - (SMZFMZ~**2) IF(~TMP .NE. O.D~) DSlGtJj = bSI5tJ:lXDTMP
1~ CONTI~JUE
RETURN
END
C
C**********************~***************************************
C
SUeROUTINE PRERKtX,NS,RK,IOR,Y) , C SUeROUTINE OETAINS (Yi THE INNOVATION PROCESS OF (X~ USING
C THE LATTICE MODEL WITH COErFS. ~RK) lMPLICIT REAL*~ (A-H,O~Z) /~
bO ~ l=l IOR lZ18448 Z~I) = o.[~ -q CONTINUE
C
bO qq 1=1,NS
E = X~l~
W(l~ = E
ljO ~q J=~,IOR
JMl = J ~ -1 ~J(J~ = ZCJMl) + Rh~J~ll*E
E = E + RK(JMlj*~(JMl~
3q CONTINUE
Y~I~ = E + R~(IOR~*Z(IOR~
~O 4~ J=l,IOR
= W(J
4~ CONTINUE
9~ CONTINUE
RETURN
ENC
C
C***********~****************************************7-*****************
C
SUBROUTINE KTOA(Rh,A,IOR:) REAL*~ A~IOR~RK(IORI~e(51J
A(lJ = RK(l~
IF~IOR .LE, 1) RETURN
C
DO 19 M=~,IOR
A~Mi = R~(MJ
MMl = ~
~O 1~ I=l,MMl B(I) = A(I~ + R~ j*A~rl-I~
1~ CONTINUE
no 15 I=l~MMl A~I) = B(I) CONTINUE
1~ CONTINUE
RETURN
EN[) C*************** I _ ___ C * * * ~.~ * * * * * * * * * *
SUBROUTINc APSPfiB(P,TltT2) C SUeROUTINE CALCULATES APOSTERIORI PRORARILITIES ~P~ OF ONE OF
C THE THREE E~UALY PRORA~LE HYPOTHESES USINÇ .~ELATI~E
C LOG-LIKELIHOOD FUNCTIONS (Tl~T i C
REAL*~ Tl~T2 C
P = O.
IF(Tl,GT,5D. .OR. T2,GT.50.j RETURN
P = 1.
IF(Tl.GT.-50.j P = P + C)EXP~Tl Ir~T-.6T.-50,j P = P + C~EXP~T~
P = 1 . /P ' T
RETURN
EN[~
-C*************************************)~*******************************
.
directing ultrasound energy into an unknown tissue;
measuring values of a characteristic of echoes of the ultra-sound, which are scattered fr~m the ~ulknown tissue.
The invention also relates to a device for remote identi-fication of a tissue comprising:
means for directing ultrasound energy into an unkno~n tissue;
means for detecting a characteristic of echoes of the ultra-sound which are scattered from the unkncwn tissue. The invention is particularly useful in connection with apparatus and methods for the diagnostic imaging of human and animal body struct~res.
Background of the invention Methods and apparatus which utilize ultrasound energy for the diagnostic imaging of body structures are well kncwn. Typically, ultrasound energy is directed into and scattered from body tissues.
m e amplitude of the resultant echoes is detected and displayed to form an image which characterizes the scattering structures. Virtu-ally all commercial medical ultrasound imaging systems form images from the envelope of echoes which are reflected from an interface of tissue types which have different acoustic impedances. me images are, therefore, useful to delineate the outlines of various bcdy organs and lesions. ~owever, because the image makes no use of the phase infor-mation in the echo signals, it is not generally possible to identify tissue types.
The prior art teaches a number of ultrasound systems which are capable of identifying tissues having specific characteristics.
U.S. Patent 4,389,893 by OPHIR and MAKLAD is typical of a class of apparatus which attempts to characterize tissue types by measuring local ultrasonic attenuation. Likewise, U.S. Patent 4,270,546 to PERILHOU and COURSANT describes apparatus for identifving the direc-tional characteristics of tissue structures.
~Z18448 PIL~ 21152 2 12-10-1983 It is an object of the invention to provide a method and a device for tissue characterization that are capable to differentiate and identify the tissues of the various body organs as well as the patholo~y of healthy and diseased tissues to provide a remote, noninvasive biopsy-Summary of the invention The method according to the invention is characterized in that it further comprises the steps of:
accumulat.ing a statistically significant sample of the measured values of the characteristic;
filtering the accu~ulated sample to determine, for each of a plurality of possible tissue types, a signal which is a measure of the likelihood that the accumulated sample oE measured values was generated lS by a mathematical mode1. which characterizes the tissue type;
combining the filtered signals using a predetermined logicaldecision function to choose one of the models which most likely produced the sample; and assigning the tissue type characterized by the chosen model as 20 the i~entity of the unknown tissue The device according to the invention is characterized in that it further comprises:
means for accumulating a statistically significant sample of signals which represent values of the detected characteristic;
2s a plurality of filter means, each filter means being associated with a possible tissue type, which function to filter the acc~mulated sample to extract a signalwhich is ameasure of the likelihood that the accumulated ~ample was generated by a mathematical model which characterizes the associated tissue type; and means for combining. the signals extracted by the filters,which use a predetermined logical decision function to select the extracted signal which is associated with the model which most likely produced the sample and assign the tissue type associated with the filter producing the chosen signal as the identity of the unknown tissue.
We have determined that ultrasonic pulse-echo data from different tissue types or pathologies exhibit diffentiable statistical regularities.
Tissue types or pathologies are classified by utilizing known modeling and hypothesis testing statistical techniques with a dedicated or general PH~ 21152 3 purpose digital o~mputer.
Samples of a pulse-echo record from unknown tissue are treated as a vector x = [xl, x2, ..., x ]. We hypothesize that this record was generated by one of a finite set of tissue types. HR
denotes the hypothesis that the data vector was generated by tissue type "R"; R = l, 2, ..., K. We then calculate the conditional pro-bability density functions p(x¦HR) (the probabilit~ density of data vector x, given that the pulse-echo record was generated by tissue type "R") using prior knowledge of the statistical properties of the pulse-echo records associated with each tissue type. m e statistical properties associated with the tissue type are determined from statis-tical models based on sets of training data obtained from known tissue types. The values of the conditional probability for each possible tissue type represent likelihood functions which are logically oom-pared and a predetermined decision rule is applied to select the tissuetype which most probably generated the data ~ector.
The likelihood functions can be computed recursively from the data samples; that is the data samples can be processed one at a time, as they are received, and the likelihood functions updated with each new data sample. Recursive processing permits real time (on-line) tis-sue classification using dedicated microoomputer-based signal pro-cessors and allows oontinuous processing of new data until one tissue type hypothesis dominates under the decision rule.
Hypothesis testing -techniques have been widely and effec-tively used in many signal processing applications of the prior art.A general discussion of these techniques and their application to specific prior art problems is set forth in the text Uncertain Dynamic Systems by Fred C. Schweppe, Prentice Hall, Inc., 1973. The success of hypothesis testing methods is, hcwever, critically dependent llpon the validity of the statistical models of the underlying physical pro-cess which are incorporated in the hypothesis testing algorithm.
We have determined that hypothesis testing is a suitable method for tissue classification from ultrasound pulse-echo data if a sequence of pulse-echo samples is modeled as an autoregressive process;
that is if the samples satisfy the equation Xn - alXn-l + a2Xn-2 + ''' + aLXn-L + Wn where Wn is an uno~rrelated random sequence (white noise) and the ~21844~3 autogressi~e parameters [al, a2~ -., aL] characterize the particular process. Estimates of these parameters can be obtained from "training sets" of pulse-echo data which are generated by known tissue types.
Thus, each tissue type is characterized by a set of autoregressive parameters. Using the autoregressive parameters for each tissue type we construct an algorithm for recursively computing the likelihood fwlction for any given pulse-echo data sequence.
The autoregressive parameters can be obtained from the train-ing sets of data by either of tw~ kncwn methods. In the first method, t~.e data are used to generate an estimate of the autocorrelation func-tion using time averages. m e estimate of the autocorrelation func-tion is then used to generate the autoregressive parameters using the Durbin-Levinson algorithm (or some variation of the Durbin-Levinson algorithm).
In the seocnd method, the autoregressive parameters are calculated directly from the data training sets using the Burg algor-ithm (or one of the many variations of the Burg algorithm).
Both of these methods, as well as their many variations, are described in Non-linear Methods of Spectral Analysis, S. Haykin ~ (Editor), Springer-Verlag, 1979.
m e sequence of tissue pulse-echo samples may, if desired, also be modeled as an autoregressive moving average process.
Kalman filters can be utilized for the recursive computation of the likelihood functions. m is technique is described in the text Digital and Ka~man Filtering by S.M. Bozic, John Wiley and Sons, 1979.
Brief description of the drawings The invention may be best understood by reference to the accompanying drawings in which:
Figure l schematically illustrates apparatus for classifying tissue in accordance with the invention;
Figure 2 illustrates a digital filter for extracting a like-lihood function from a pulse-echo data vector;
Figure 3 is a flow chart for a preferred embodiment of a tissue classification algorith~;
Figures 4 through 6 are plots of the logarithms of likelihcod functions which were recursively computed from echo samples generated in ~2~8448 sponge samples; ~`ld Figures 7 through 9 are plots of the logarithms of likelihood functions which were recursively computed from ultrasound echo samples derived, in vivo, from human liver, pancreas and spleen tissue.
Brief description of the appendix The APPENDIX is a listing of Fortran computer programs which implenent a preferred en~odinent of the invention.
10 Description of a preferred em~odin~nt Figure 1 is a preferred en~odiment of apparatus for tissue characterization in aco~rdance with the invention. A transmitter 10 activates an ultrasound transducer 20 via a TR switch 30 to direct pulses of ultrasound enera,y into a test body 40 which may, for example, be a lS human a~don~n. Ultrasound energy from the transn~itter is scattered by ~1y tissues including unknown tissues in a region 50. Portions of the scattered energy are returned as echoes to the transducer 20 and are directed via the TR switch into a receiver 60.
The receiver detects and processes the echo signals in a con-20 ventional manner to generate video signals which are utilized to aeneratean image on a display unit 70. The displayed image may, for example, be a conventional B-scan image of the internal structure of the ~ody 40.
The ampified radio frequency echo signals form the receiver 60 are also transmitted via line 80 to a sampling circuit 90. A time gate 25 100 is operatively connected to the sampling circuit 90 in a manner which causes the sampler to extract a series of periodic samples of the values of received echo signals which were scattered from tissue in the unknown region 50. The operation of the time gate may, for example, be controlled in a known manner by means of a cursor which is manually set 30 over a given region of the display 70 or may be triggered by high amplitude echoes which are scattered form the edges of the region 50.
Signals representing the amplitude of the series of saTnples derived from the sampler 90 are digitized in an analog-to-digital converter 110.
The series of digitized signals are then applied to a parallel chain of 35 digital likelihood function filters 120a through 120n (nore particularly descri~ed below). Each filter ca]culates the value of a likelihood function which represents the conditional probability that the series of san~ple signals ~ere produced in a given tissue type. The ouput signaLs ~8448 from the parallel filters are fed to decision function logic 130, which using known algorithms, assigns one of the possible tissue types to the unkr,own tissue represented by the s 9 1e vector.
If the a priori probabilities of occurrence of the various tissue types are equal (or unknown) a preferred em~odiment of the decision function logic assigns the unknown tissue as the tissue type associated with the likelihood function filter which produces the largest output signal. The decision function logic may, however, be modified in a known manner to choose the lik.elihood function having the largest aposteriori lO probability if the various tissue types have different a priori probabilities of occurrence or to minimize the expected penalty if there is a large penatly attached to misidentification errors of a specific type.
The ouput signal from the decision function logic 130 is routed to the display 70 where it is used to identify the tissue type of the 5 unkno~Jn region; for ex 9 1e by display of a message associated with a cursor or by assigning a particlar color or brightness level to pixels in an image of the region where the echo sam~les originated.
In an alternate em~odiment of the invention the output signals of the likelihood function filters may be utilized (either directly or 20 after processing in decision function logic) to modulate the color or intensity of the associated region in the display. In the latter case the saturation of color in the associated region could provide an inclication of the confidence level of the tissue characterization for that region.
Figure 2 illustrates the typical filter for calculating the logarithm of the likelihood function for a given hypothesis ~. The series of periodic data s 9 1es Z(nT) is serially fed to the input of a delay stage 200 having a delay time T and to the positive input of a subtractor 230.
The output of the delay stage 200 is fed through a one-step Ka~1lan 30 predictor error filter which characterizes the hypothesis HR. The output of the Kalman filter 220 is fed to the negative input of the subtractor 230.
The output of the subtractor thus represents the difference between the actual value and the predicted value for each component of the data vector:
that is the prediction error of the Kalman filter. The prediction error 35 is processed in a weighted squarer 240 and a bias signal is added to the output of the squarer in an adder stage 250. The signals from the output of the adder 250 are summed in an accumulator 260 whose output represents a likelihood function for the hypothesis which is modeled by the Kalman 121~34~8 ~A 21152 7 12-10-1983 filter.
The a-ttached Appendix comprises aFortran computer program ~hich represents the preferred embodiment for generating likelihood functions from data sample vectors. The scattering in the various tissue types are modeled as time invariant, autoregressive processes. The prediction error, ~A~ich is called the innovations representations of the process, is assumed to be white and Gaussian. The program further functions to calculate the coefficients which characterize the autoregressive model for each tissue tvpe from data vector samples which are obtained from lO echoes of known tissue regions.
In the appendix: the program TCP generates an autoregressive model for each of three classes of training sample data vectors (A-lines) and tests the models on other samples from tissue in the same classes;
the subroutine PRK obtains the lattice predictive m~del from 15 data in program TCP;
the subroutine GETTCD reads data from a disc and normalizes it to zero mean and unit variance;
the subroutine PRERK, using the lattice model, obtains the Innovations process of the input data;
the subroutine APSPRB calculates a posteriori probabilities of one of three equally probably hy~otheses using relative log likelihood functions;
Figure 3 is a flow chart of the processor used to calculate to likelihood functions by the program of the Appendix.
Figure 3 represents a typical calculation of a log likelihood function ln p(Yn/Hi) for a hypothesis Hi. In the first block 300 the innovations model subroutine PRERK is used to calculate the prediction error for each data point l...n The second block 340 corresponds to the weighted squarer 240 of Figure 2 and the addition circle 350 corresponds 30 to the bias addition in block 250 of Figure 2. The ouput of the addition circle 350 is accumulated over all data points in the s~ner block 360.
The latter three operations are implemented in the main body of the progrc~m following the comments at the top of the third page of the Appendix. The log likelihood functions are the compared using a decision rule 330 to 35 obtain an identification of the tissue type.
Example 1 Pulse echo data was collected from samples of a natural sponge lZ18448 Pl~ 21152 8 12-10-1983 and from t~ different types of synthetic sponge using a 19mm diameter, 3.5 ~lz transducer in a water tank and a Philips 5580 s-Scanner. The output of the scanner TGC amplifier was fed to a Biomation 8100 Transient ~ecorder (20 ~lHz sample rate, 5 bits/sample) which was interfaced to a 5 D~C 11/~0 nuni-computer. The pulse-echo records from each sponge sample were divided into two sets; a training set and a test set. Parameters for the autoregressive model were generated, using the method set forth in the ~ppendix, for each of the test samples.In this way autoregressive mc~els ~.~ere generated for each sponge type. Data vectors were then obtained from echoes which originated in the three test samples. Figure 4 illustrates the values of the likelihood functions ln p(Z~ Hi) computed for a data vectorderived from sponge sample 1 using the computed hypotheses based on mxlels from the teaching samples of sponge types 1, 2 and 3. As illustrated, after a reasonable number of samples (n), the likelihood lS function associated with hypothesis 1 has the hio,hest value thereby correctly iclentifying the sponge sample as type 1. Figures 5 and 6 illustrate similar computations for test samples of sponge types 2 and 3. In all cases the apparatus correctly identified the test sponge sample.
For comparison the power spectra of the pulse-echo records of the 20 three sponge types were computed and were found to be essentially indistinguishable.
Example 2 Teaching data and test data vectors ~iere obtained from in vivo 25pulse-echo data from a human liver (L), spleen (S) and pacreas ~P) and the procedures utilized with the sponge data were repeated. The computed likelihood functions are illustrated for liver test data vectors in Figure 7, for pancreas test data vectors in Figure 8 and for spleen test data vectors in Figure 9. The filters correctly identified liver and pancreas 30pulse-echoes with a high degree of reliability. The algorithm correctly differentiat~d spleen data from liver but with substantially more difficul-ty and lower confidence than the other samples.
The preferred em~odiment of the present invention characterizes tissue types on a basis of the statistical characteristics of samples of the 35amplitude of ultarsonic pulse-echoes. The statistical characteristics of other measured characteristics of pulse-echo data may however also be utilized in the computational algorithm. For example, the instantaneous frec~uency or phase of -the reflected signal might be sampled in place of or in conjuction with the amplitude of the signal.
- APPENDIX lZ1844~3 TPE T~ FORTRAN
'~****t.~***********************~**********************~*****~************
C PROGRAM TCP ~, C PR06RAM GENERATES AR MODEL FOR EACH THREE CLASSES OF TRAINI~5 C SAMPLE A-LINES (REAL OR SYNTHETICl AN~' TEST THE MODEI_S ON OTHER
r SAMPLE A-LINES FROM THE SAME CLASSES
C
C
C************************~********************************************
C ..
- IMPLICIT REAL*~ ~A-H,O~Z) REAL*~ ~S~6('JD4$),R~<(51,3J,RY~T~5lJ,ERROR(51,3j,ERRT~51~,RINN~D4~) ~ ,RKS~51,3),AS~51,3) - REAL XLL(~n4~,3),XPLT~10'4),YPLT~lD"4),RES(~D4 INTEGER IOR(3),IOR5~3),ISTPT~3 .
~EFINE FILE 1'~36,~D4~U~lV4),13(36,~D4~,U,IV5),14~36,~D4~,U~IV~
~ ,15~36,~04~,U,IV7),16~36,~D43,U,IV~1,17~36,'04~,U,IV9) C SPECIFICATIONS
C .
5D ~JRlTE(6,51) 51 FORMAT~' TYPE: FIRST RECOR~) ANb # OF RECORDS FOR TRAINING'~
' THE SAME FOR THE RECOR~1S TO BE TESTE~'/
' START PTS. FOR EAC~ CLASS'J - -' # OF SAMPLESJRE50R~'f & ~ MAX ORC)ER OF THE AR ~O[)EL'J
~X~'OPTlOtl~ ISR bATA. ~=SYN DATA') REA~)~5,*) NFTR,NTR,NFTS,NTS,~I5TPT~I),I=i,3),NSAM,MOR,ISOPT
IF~MlND~NFTR~NTR.NFTS,NTS).LT.l .OR. MAXOiNFTR.NFTS).GT.l .OR. MAXO~NFTR+NTR,NFTS+NTS).GT.13 .OR. MOR.GT.50 .Ofi.
ISOPT.LT.l .OR. ISOPT.GT.~ .OR. NSAM.LT.~ .OR. NSAM.GT.'~04~) 50 ~0 59 I=1,3 IF~ISTPT(I).LT.l .OR. ISTPT~I)+NSAM.GT.~D4q) GOTO 5D
C
C IF THE OPTION IS SYN. ~ATA
IF~ISOPT oEQ~ 1) GOTO 95 ~0 WRITE~6,Bl) ~1 FORMAT~J' TYPE:'J
' OR~ER OF AR MODEL ~M~ FOR CLASS 1 & ~I) I=l,M'J
& ' THE SAMc FOR CLASS ~ & 3, PLEASE ONE CLASSJLIN') DO ~5 I=1,3 READ~5,*) M,~R~S~J,I),J=l,M) IORS~I) = M
~5 CO~ U-C CO JERT A-CQEFS. TO ~-COEFS. eY sue. ~TOA
~0 8~ I=1,3 -~
CALL I~TOA~R~S(l,II,AS(1,I),IORS~
aa CONTINUE
C
C OeTAIN K-COEFS. FROM TRAl~IN5 ~ATA
C
q5 ~0 lOq ISP=l,~
CALL PRt~ISP,NFTh,NTR,ISTPT(ISP),NSAM,MOR,R~l,ISP)~
& ERROR(l,ISP),~,ISOPT,AStl,ISP3,IORS(ISP),bSIG~
CALL PRli(ISP,NFTS,NTS,ISTPT(ISPJ,NSAM,MOR,RKT,ERRT,., ~ ISOPT,AS~l,ISP),IORS(ISP),~SI6) lD9 CONTINUE - -C
C HYPOTHESIS TESTING
C ' :' 150 WRITE(~,151) 151 FORMAT(/ TYPE: THE ORDER OF THE AR MO~EL FOR CLASS 1 A-LINES i & ~ THE SAME FOR CLASS 2 JaY., THE SAME FOR CLASS 3 J
~X, THE REbUCTION FACTOR & MAX ~ OF POINTS FOR PLOTS f & ~X, 1 = E~UAL ~ARIANC S 2 = NOT 1 ) ~ - -REA~(~,*) (IOR(I~,I=l,~),IR~C,MNPT,IEQVAR
~0 153 I=1,3 IF(IOR(I).LT.D .OR. IOR(I~.GT.50) GOTO 15D
IORMAX = MAXO(IOR(l),IOR(,~),IOR(3)~
IF~iR~C.LT.1 .OR. MNPT.LT.2~ 60T0 150 NPT = NSAMJIROC
IF(NPT.GT.MNPT~ NPT = MNPT
C
~0 299 I=l,l, C
IR = I
DO 2~9 ISP=1,3 ; IF(ISOPT.EQ.1) CALL GETTC[)(r)SIG,1,204~,IR,ISP) ISEE~ = IR*3 + ISP - 1 IF(ISOPT.EQ~2~ CALL GET[)R~AS(l,ISP),IORS(ISP),l.,[)SIG,2n4~, & IScE~) C ' .. "
lF(I.GT.1) GOTO 175 DO 170 J=l,NSAM
YPLT(~) = DSIG(J~ -Y.PLT(J~ = J
i70 CONTINU~
C WRITE(b,172~ ISP -t 172 FORMAT(JJT3D, NORMALIZE~) A-LINE FRO~ CLASS ,I4 C CALL PLOT(XPLT,YPLT,300~
., ~o 175- ` ~)0 ,_~ IH=1,3 lZ18448 C ( ERATE THE INNO~ATIOt`lS PROCES5 e.Y SUB. PRER~
C - .
CALL PRERI-~DSJG, 048,RI~l,IH:~,lO~IH-J,RI~`JN:~
C CALCULATE THE L06-LIKELIHOOD FLItlCTIONS SEQUEtlTIALY
C
~AR = EhROF(lOR(JH),IH) IF~IE~AR~Eo.lJ VAR = 1.
PIAS = -.5*t)LOG(VAR) Y.T = O.
. DO 211 J=l, D4 IND = J
lF(IN[) .LE. IORMAXI GOTO 20q . XT = Y.T - .5*RIGII`l IN~J** JVAP.
XLL(J,IHJ = XT + J*BlAS
RES~J) = RINN~IN[;) 211 COtlTlNUE
I~JR = (lSP - 1)*1~ + I
IFILE = 11 + Il~
WRITE(lFILE IWR) RES
I~R = ~I-1)*3 + IH
IFILE = 14 ~ lSP
WRITE~I~ILE I~lRJ ~XLL~J,I~),J=1,204~J
~2q COG!TINUE
C
C
25q CONTINUE
29q CONTINUE
STOP
END
C
C***********************************************************************
C
SUBROUTINE PR~ISP,NF,Gl,IST,NS,MOR,Rh,ERROR,ICl,lC ,AS,IORS,DSIG) C
C SUe.ROUTINE OBTAINS THE LATTICE PREDICTIVE MODEL FROM DATA
C
IMPLICIT REAL*~ (A-H,O-Z~
REAL*~ PHI(51,51),PHIO(51,51),R~(51:1,TK~51,1,J,SCR~51),DSIG(204a) & ,A(51),ERROR(51J.AS(IORS) REAL SPECT(1.~9j,5PAYIS(1 9~ -C
O ~ J=1,1.9 .~
SPECT(I) = O. - _ -3 CONTINUE -. -~0 9 J=1,51 . ..
~0 Y ~ ,51 . - ~ONT I iNUE ~218448 O 3~ I=1,N
IR = NF + I -IF(IC..EQ.1~ CALL GETTC[)(D~I6,JST,NS,IR,ISP~
ISEE~) = IR*3 + ISP - 1 IF(IC~.E~. -.) CALL GETDR AS,IOR5,1.,[~5I6,NS,ISEE[~J
C
CALL PR[~TMP~SPECT,DSI~,NSJ
C
C COMPUTE THE COVARIANCE MATRIX
C
CALL DCOVR1(DSIG,NS,MOR,PHI~51) MORP1 = MOR + I
DO 19 J=-1,MORP1 b O i 9 K=i~MORP1 PHID(K,J) = PHIO(K,J~ + PHI(K,J) 1 q CONTINUE
C
DO ~9 M=1,MOR
CALL DCLHRM(PHI,51,M,A,TK(M,I),TERR,SCR) 2q CONTINUE
3q CONTINUE
C
D 0 4 ~ J=1,1 9 SPECT(J) = SPECT(JJJFLOAT(N) SPAY.IS f Ji = FLOAT(J-i3Jl a.
4~ CONTINUE
C
DO 4q M=1,MOR
CALL DCLHRM(FHID,51,M,A,RK(M),ERROR~M),SC~J
ERROR(M) = ERROR~M)/((tlS - M)*N*~) -C
C PRINT THE K CHART
C
DO i~ M=1,MOR
AVK = O.bO
- DO ~q I=1,N
AVK = AVK + TK(M,I) 8q CONTINUE
AVK = AVKJN
WRITE(6,~1) M~(T~(M,I~,I=1,N),AV~,R~(M~,ERROR~M) ~1 FORMhT~I3,15F~.6) 1 ~q CONTINUE ~ -C CALL PLOt(SPAXIS,SPECT,1 ~) RETURN -EN~
C
/~2 .
.,......................... - - 18448 . - sua~ouT~ ETTC~?fDSIG,IsT,~ls,I~ sP~ ~2 C ROUTINE READS THE [)ATA FRO~l [)ISC, THE NORMALIZE THEM TO
C ZERO-ME~N ANC` UNTT-VARIANCE.
C
REAL*B ~5IG~S~,SQ,SM,[TMP
INTEGER* ID~2D4~i r ~EFINE FILE7tl-,40~L,I~ 10(1 ~40~L~IV~ ,40~6~L~IV3 C
J = ISP ~ ~
READ(J IR~ ID
SM = 0.[)0 SO = O.DO
MSZ = 55 MSZ2 = MSZi2 00 lq J=l~NS
I = J + IST - 1 IF(J.NE.13 GOTO $5 DO 11 L~ MSZ~
L = L2 + IST - 1 SM = SM + ID(L) SQ = SQ + IDtL~*I~)(L) IA~[) = I + MSZ.~
I5BT = I - M5Z - 1 IF(IAD~.GT.NS+IST-l) GOTO 16 SM = SM + IC)~IADD) SQ = SQ + I~(IADC))*I[)(lADD~
lb IF(IS~T .LT. IST) GOTO 17 SM = SM - I[)(ISeT) S~ = SO - lD~ISBT)*I~tISBT) 17 FMZ = MSZ
IF(J.LE.MSZ~I FMZ = MSZ2 + J
IF(NS-J .LT. MSZ2) FMZ = MSZ ~ NS - J + 1 bSIG(J) = ID(I) - SMZFMZ
DTMP = ~S~RT(S~JFMZ - (SMZFMZ~**2) IF(~TMP .NE. O.D~) DSlGtJj = bSI5tJ:lXDTMP
1~ CONTI~JUE
RETURN
END
C
C**********************~***************************************
C
SUeROUTINE PRERKtX,NS,RK,IOR,Y) , C SUeROUTINE OETAINS (Yi THE INNOVATION PROCESS OF (X~ USING
C THE LATTICE MODEL WITH COErFS. ~RK) lMPLICIT REAL*~ (A-H,O~Z) /~
bO ~ l=l IOR lZ18448 Z~I) = o.[~ -q CONTINUE
C
bO qq 1=1,NS
E = X~l~
W(l~ = E
ljO ~q J=~,IOR
JMl = J ~ -1 ~J(J~ = ZCJMl) + Rh~J~ll*E
E = E + RK(JMlj*~(JMl~
3q CONTINUE
Y~I~ = E + R~(IOR~*Z(IOR~
~O 4~ J=l,IOR
= W(J
4~ CONTINUE
9~ CONTINUE
RETURN
ENC
C
C***********~****************************************7-*****************
C
SUBROUTINE KTOA(Rh,A,IOR:) REAL*~ A~IOR~RK(IORI~e(51J
A(lJ = RK(l~
IF~IOR .LE, 1) RETURN
C
DO 19 M=~,IOR
A~Mi = R~(MJ
MMl = ~
~O 1~ I=l,MMl B(I) = A(I~ + R~ j*A~rl-I~
1~ CONTINUE
no 15 I=l~MMl A~I) = B(I) CONTINUE
1~ CONTINUE
RETURN
EN[) C*************** I _ ___ C * * * ~.~ * * * * * * * * * *
SUBROUTINc APSPfiB(P,TltT2) C SUeROUTINE CALCULATES APOSTERIORI PRORARILITIES ~P~ OF ONE OF
C THE THREE E~UALY PRORA~LE HYPOTHESES USINÇ .~ELATI~E
C LOG-LIKELIHOOD FUNCTIONS (Tl~T i C
REAL*~ Tl~T2 C
P = O.
IF(Tl,GT,5D. .OR. T2,GT.50.j RETURN
P = 1.
IF(Tl.GT.-50.j P = P + C)EXP~Tl Ir~T-.6T.-50,j P = P + C~EXP~T~
P = 1 . /P ' T
RETURN
EN[~
-C*************************************)~*******************************
.
Claims (12)
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:
1. A method for remote identification of a tissue using ultrasound, comprising the steps of:
directing ultrasound energy into an unknown tissue;
measuring values of a characteristic of echoes of the ultrasound which are scattered from the unknown tissue;
characterized in that the method further com-prises the steps of:
accumulating a statistically significant sample of the measured values of the characteristic;
filtering the accumulated sample to determine, for each of a plurality of predetermined mathematical models, each of which is associated with a possible tissue type, a signal which is a measure of the likelihood that the accumulated sample of measured values was generated by said mathematical model;
applying the filtered signals as inputs of a predetermined logical decision function which selects one of the models which most likely produced the sample; and assigning the tissue type which is associated with the chosen model as the identity of the unknown tissue.
directing ultrasound energy into an unknown tissue;
measuring values of a characteristic of echoes of the ultrasound which are scattered from the unknown tissue;
characterized in that the method further com-prises the steps of:
accumulating a statistically significant sample of the measured values of the characteristic;
filtering the accumulated sample to determine, for each of a plurality of predetermined mathematical models, each of which is associated with a possible tissue type, a signal which is a measure of the likelihood that the accumulated sample of measured values was generated by said mathematical model;
applying the filtered signals as inputs of a predetermined logical decision function which selects one of the models which most likely produced the sample; and assigning the tissue type which is associated with the chosen model as the identity of the unknown tissue.
2. A method as claimed in Claim 1, characterized in that the step of directing energy comprises directing pulses of ultrasound energy into the tissue and wherein the measured values are values of the instantaneous ampli-tude of the echoes.
3. A method as claimed in Claim 2, characterized in that the step of filtering comprises calculating the log-arithm of the likelihood function for each possible tissue type.
4. A method as claimed in Claim 2, characterized in that the step of filtering determines a signal which is a measure of the likelihood that the accumulated sample of the echo amplitudes was generated by autoregressive mathe-matical models.
5. A method as claimed in Claim 4, characterized in that the step of filtering comprises filtering the accum-ulated sample with Kalman one-step prediction error filters.
6. A method as claimed in Claim 2, characterized in that the step of filtering comprises recursively filtering values of the data sample as successive amplitude values are accumulated.
7. A method as claimed in Claim 1, characterized in that the step of filtering comprises filtering the accum-ulated sample in one or more filters which are character-ized by numerical parameters and further comprising the step of determining numerical parameters which character-ize each possible tissue type from samples of echoes which are scattered from a known sample of the particular tissue type.
8. A device for remote identification of a tissue comprising:
means for directing ultrasound energy into an unknown tissue;
means for detecting a characteristic of echoes of the ultrasound which are scattered from the unknown tissue;
characterized in that the device further com-prises:
means for accumulating a statistically signifi-cant sample of signals which represent values of the detected characteristic;
a plurality of filter means, each of which function to filter the accumulated sample to extract a signal which is a measure of the likelihood that the accumulated sample was generated by an associated pre-determined mathematical model wherein each of said mathe-matical models is associated with a possible tissue type;
and means which apply the signals extracted by the filters as inputs to a predetermined logical decision func-tion to select the extracted signal which is associated with the model which most likely produced the sample and assign the tissue type associated with the model producing the chosen signal as the identity of the unknown tissue.
means for directing ultrasound energy into an unknown tissue;
means for detecting a characteristic of echoes of the ultrasound which are scattered from the unknown tissue;
characterized in that the device further com-prises:
means for accumulating a statistically signifi-cant sample of signals which represent values of the detected characteristic;
a plurality of filter means, each of which function to filter the accumulated sample to extract a signal which is a measure of the likelihood that the accumulated sample was generated by an associated pre-determined mathematical model wherein each of said mathe-matical models is associated with a possible tissue type;
and means which apply the signals extracted by the filters as inputs to a predetermined logical decision func-tion to select the extracted signal which is associated with the model which most likely produced the sample and assign the tissue type associated with the model producing the chosen signal as the identity of the unknown tissue.
9. A device as claimed in Claim 8, characterized in that the means for directing energy are adapted to direct pulses of ultrasound energy into the unknown tissue and in that the means which detect the characteristic of the echoes are adapted to extract a time sequence of samples of the value of the instantaneous amplitude of the echoes.
10. A device as claimed in Claim 9, characterized in that the mathematical models associated with the filter means are autoregressive models.
11. A device as claimed in Claim 10, characterized in that the filter means comprise Kalman one step error pre-diction filters.
12. A device as claimed in Claim 8, characterized in that the filter means comprise one or more filters which are characterized by a series of autoregressive parameters which are produced by the process of:
directing pulses of ultrasound energy into a sample of a known tissue type;
detecting a sequence of values of the instantane-ous amplitude of echoes of the ultrasound which are scattered from the known tissue type;
accumulating a statistically significant sample of the detected amplitude; and calculating values of the autoregressive para-meters which model the sequence of accumulated samples as an autoregressive process and incorporating the calculated values in the filter.
directing pulses of ultrasound energy into a sample of a known tissue type;
detecting a sequence of values of the instantane-ous amplitude of echoes of the ultrasound which are scattered from the known tissue type;
accumulating a statistically significant sample of the detected amplitude; and calculating values of the autoregressive para-meters which model the sequence of accumulated samples as an autoregressive process and incorporating the calculated values in the filter.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US06/478,172 US4542744A (en) | 1983-03-23 | 1983-03-23 | Method and apparatus for remote tissue identification by statistical modeling and hypothesis testing of echo ultrasound signals |
| US478,172 | 1983-03-23 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA1218448A true CA1218448A (en) | 1987-02-24 |
Family
ID=23898826
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA000450264A Expired CA1218448A (en) | 1983-03-23 | 1984-03-22 | Method and device for remote tissue identification by statistical modeling and hypothesis testing of echo ultrasound signals |
Country Status (9)
| Country | Link |
|---|---|
| US (1) | US4542744A (en) |
| EP (1) | EP0120537B1 (en) |
| JP (1) | JPS59177038A (en) |
| AU (1) | AU566624B2 (en) |
| BR (1) | BR8401249A (en) |
| CA (1) | CA1218448A (en) |
| DE (1) | DE3479447D1 (en) |
| ES (2) | ES530761A0 (en) |
| IL (1) | IL71290A0 (en) |
Families Citing this family (46)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4512196A (en) * | 1983-09-30 | 1985-04-23 | North American Philips Corporation | Ultrasound imaging with FM detection |
| US4703462A (en) * | 1985-04-15 | 1987-10-27 | Sanders Associates, Inc. | Virtually steerable parametric acoustic array |
| US4719923A (en) * | 1985-05-06 | 1988-01-19 | American Telephone And Telegraph Company, At&T Bell Laboratories | Non-invasive blood flow measurements utilizing autoregressive analysis with averaged reflection coefficients |
| US4817015A (en) * | 1985-11-18 | 1989-03-28 | The United States Government As Represented By The Secretary Of The Health And Human Services | High speed texture discriminator for ultrasonic imaging |
| US4855911A (en) * | 1987-11-16 | 1989-08-08 | Massachusetts Institute Of Technology | Ultrasonic tissue characterization |
| JP2001519674A (en) * | 1991-05-10 | 2001-10-23 | ボード、オブ、リージェンツ、ザ、ユニバーシティー、オブ、テキサス、システム | Elastography measurement and imaging method and apparatus for implementing the method |
| US5417215A (en) * | 1994-02-04 | 1995-05-23 | Long Island Jewish Medical Center | Method of tissue characterization by ultrasound |
| US5462059A (en) * | 1994-05-25 | 1995-10-31 | The Regents Of The University Of California | Method for assessing and displaying vascular architecture using ultrasound |
| US5390677A (en) * | 1994-05-31 | 1995-02-21 | The Regents Of The University Of California | Method for assessing and displaying the true three dimensional magnitude of blood velocity |
| US7764231B1 (en) | 1996-09-09 | 2010-07-27 | Tracbeam Llc | Wireless location using multiple mobile station location techniques |
| US7903029B2 (en) | 1996-09-09 | 2011-03-08 | Tracbeam Llc | Wireless location routing applications and architecture therefor |
| US7274332B1 (en) | 1996-09-09 | 2007-09-25 | Tracbeam Llc | Multiple evaluators for evaluation of a purality of conditions |
| US7714778B2 (en) | 1997-08-20 | 2010-05-11 | Tracbeam Llc | Wireless location gateway and applications therefor |
| US9134398B2 (en) | 1996-09-09 | 2015-09-15 | Tracbeam Llc | Wireless location using network centric location estimators |
| US7789841B2 (en) | 1997-02-06 | 2010-09-07 | Exogen, Inc. | Method and apparatus for connective tissue treatment |
| US7108663B2 (en) | 1997-02-06 | 2006-09-19 | Exogen, Inc. | Method and apparatus for cartilage growth stimulation |
| US5904659A (en) | 1997-02-14 | 1999-05-18 | Exogen, Inc. | Ultrasonic treatment for wounds |
| US6385474B1 (en) | 1999-03-19 | 2002-05-07 | Barbara Ann Karmanos Cancer Institute | Method and apparatus for high-resolution detection and characterization of medical pathologies |
| EP1076586B1 (en) | 1998-05-06 | 2006-10-11 | Exogen, Inc. | Ultrasound bandages |
| CA2338735A1 (en) | 1998-07-21 | 2000-02-03 | Acoustic Sciences Associates | Synthetic structural imaging and volume estimation of biological tissue organs |
| US8135413B2 (en) | 1998-11-24 | 2012-03-13 | Tracbeam Llc | Platform and applications for wireless location and other complex services |
| US6304833B1 (en) * | 1999-04-27 | 2001-10-16 | The United States Of America As Represented By The Secretary Of The Navy | Hypothesis selection for evidential reasoning systems |
| CA2374568C (en) | 1999-05-21 | 2011-11-22 | Exogen, Inc. | Apparatus and method for ultrasonically and electromagnetically treating tissue |
| CA2377866A1 (en) | 1999-06-14 | 2000-12-21 | Roger J. Talish | Method and kit for cavitation-induced tissue healing with low intensity ultrasound |
| US6264609B1 (en) | 1999-09-15 | 2001-07-24 | Wake Forest University | Ultrasound apparatus and method for tissue characterization |
| AU1367101A (en) | 1999-09-24 | 2002-01-08 | Dennis J. Dupray | Geographically constrained network services |
| US10684350B2 (en) | 2000-06-02 | 2020-06-16 | Tracbeam Llc | Services and applications for a communications network |
| US9875492B2 (en) | 2001-05-22 | 2018-01-23 | Dennis J. Dupray | Real estate transaction system |
| US10641861B2 (en) | 2000-06-02 | 2020-05-05 | Dennis J. Dupray | Services and applications for a communications network |
| CA2426903C (en) | 2000-10-25 | 2011-12-06 | Exogen, Inc. | Transducer mounting assembly |
| US8082096B2 (en) | 2001-05-22 | 2011-12-20 | Tracbeam Llc | Wireless location routing applications and architecture therefor |
| US7429248B1 (en) | 2001-08-09 | 2008-09-30 | Exogen, Inc. | Method and apparatus for controlling acoustic modes in tissue healing applications |
| EP1551303A4 (en) * | 2002-05-16 | 2009-03-18 | Karmanos B A Cancer Inst | Method and system for combined diagnostic and therapeutic ultrasound system incorporating noninvasive thermometry, ablation control and automation |
| US7285092B2 (en) | 2002-12-18 | 2007-10-23 | Barbara Ann Karmanos Cancer Institute | Computerized ultrasound risk evaluation system |
| US6926672B2 (en) * | 2002-12-18 | 2005-08-09 | Barbara Ann Karmanos Cancer Institute | Electret acoustic transducer array for computerized ultrasound risk evaluation system |
| US6837854B2 (en) * | 2002-12-18 | 2005-01-04 | Barbara Ann Karmanos Cancer Institute | Methods and systems for using reference images in acoustic image processing |
| US10201324B2 (en) | 2007-05-04 | 2019-02-12 | Delphinus Medical Technologies, Inc. | Patient interface system |
| US8870771B2 (en) * | 2007-05-04 | 2014-10-28 | Barbara Ann Karmanos Cancer Institute | Method and apparatus for categorizing breast density and assessing cancer risk utilizing acoustic parameters |
| CN102843959B (en) | 2010-02-12 | 2014-11-12 | 戴尔菲纳斯医疗科技公司 | Method of characterizing the pathological response of tissue to a treatmant plan |
| CN102869301B (en) | 2010-02-12 | 2016-06-29 | 戴尔菲纳斯医疗科技公司 | The method characterizing the tissue of patient |
| US9763641B2 (en) | 2012-08-30 | 2017-09-19 | Delphinus Medical Technologies, Inc. | Method and system for imaging a volume of tissue with tissue boundary detection |
| US10123770B2 (en) | 2013-03-13 | 2018-11-13 | Delphinus Medical Technologies, Inc. | Patient support system |
| US10143443B2 (en) | 2014-05-05 | 2018-12-04 | Delphinus Medical Technologies, Inc. | Method for representing tissue stiffness |
| US10743837B2 (en) | 2014-08-04 | 2020-08-18 | Delphinus Medical Technologies, Inc. | Ultrasound waveform tomography method and system |
| US10285667B2 (en) | 2014-08-05 | 2019-05-14 | Delphinus Medical Technologies, Inc. | Method for generating an enhanced image of a volume of tissue |
| EP3688485A1 (en) * | 2017-09-25 | 2020-08-05 | Koninklijke Philips N.V. | Methods and systems for filtering of acoustic clutter and random noise |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US3614719A (en) * | 1968-11-25 | 1971-10-19 | United Aircraft Corp | Adaptive pulse echo target identification system |
| US4207624A (en) * | 1978-10-02 | 1980-06-10 | Rockwell International Corporation | Frequency domain adaptive filter for detection of sonar signals |
| JPS57550A (en) * | 1980-06-03 | 1982-01-05 | Fujitsu Ltd | Identification systm for specimen |
| US4428235A (en) * | 1980-06-20 | 1984-01-31 | Hitachi, Ltd. | Non-destructive inspection by frequency spectrum resolution |
| JPS5789858A (en) * | 1980-11-25 | 1982-06-04 | Shimadzu Corp | Ultrasonic diagnostic apparatus |
| US4389893A (en) * | 1981-06-01 | 1983-06-28 | North American Philips Corporation | Precision ultrasound attenuation measurement |
-
1983
- 1983-03-23 US US06/478,172 patent/US4542744A/en not_active Expired - Fee Related
-
1984
- 1984-03-20 IL IL71290A patent/IL71290A0/en unknown
- 1984-03-20 DE DE8484200394T patent/DE3479447D1/en not_active Expired
- 1984-03-20 EP EP84200394A patent/EP0120537B1/en not_active Expired
- 1984-03-20 BR BR8401249A patent/BR8401249A/en unknown
- 1984-03-20 ES ES530761A patent/ES530761A0/en active Granted
- 1984-03-21 AU AU25963/84A patent/AU566624B2/en not_active Ceased
- 1984-03-22 CA CA000450264A patent/CA1218448A/en not_active Expired
- 1984-03-23 JP JP59054571A patent/JPS59177038A/en active Pending
- 1984-07-27 ES ES534690A patent/ES8601674A1/en not_active Expired
Also Published As
| Publication number | Publication date |
|---|---|
| DE3479447D1 (en) | 1989-09-21 |
| ES8501223A1 (en) | 1984-11-16 |
| BR8401249A (en) | 1984-10-30 |
| ES530761A0 (en) | 1984-11-16 |
| ES534690A0 (en) | 1985-11-01 |
| ES8601674A1 (en) | 1985-11-01 |
| US4542744A (en) | 1985-09-24 |
| EP0120537A2 (en) | 1984-10-03 |
| EP0120537A3 (en) | 1986-04-16 |
| JPS59177038A (en) | 1984-10-06 |
| EP0120537B1 (en) | 1989-08-16 |
| AU2596384A (en) | 1984-09-27 |
| IL71290A0 (en) | 1984-06-29 |
| AU566624B2 (en) | 1987-10-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA1218448A (en) | Method and device for remote tissue identification by statistical modeling and hypothesis testing of echo ultrasound signals | |
| US5651363A (en) | Ultrasonic bone assessment method and apparatus | |
| Zhang et al. | The removal of wall components in Doppler ultrasound signals by using the empirical mode decomposition algorithm | |
| US7929375B2 (en) | Method and apparatus for improved active sonar using singular value decomposition filtering | |
| US5245589A (en) | Method and apparatus for processing signals to extract narrow bandwidth features | |
| US5327893A (en) | Detection of cholesterol deposits in arteries | |
| US6512854B1 (en) | Adaptive control and signal enhancement of an ultrasound display | |
| CN106526585B (en) | Tracking before target detection based on the filtering of Gaussian particle gesture probability hypothesis density | |
| AU2008327744B2 (en) | Real-time robust method for determining the trajectory of one or more cetaceans by means of passive acoustics, using a laptop computer | |
| US5730144A (en) | Method and apparatus for predicting the efficacy of cardioversion | |
| CN101595396A (en) | The apparatus and method of high-resolution ultrasound displacement measurement | |
| Bevan et al. | Experimental quantification of noise in linear ultrasonic imaging | |
| Liguori et al. | Choosing bootstrap method for the estimation of the uncertainty of traffic noise measurements | |
| Kershenbaum et al. | An image processing based paradigm for the extraction of tonal sounds in cetacean communications | |
| KR102459053B1 (en) | Ultrasonic system for detecting fluid flow in an environment | |
| CN113570594A (en) | Method and device for monitoring target tissue in ultrasonic image and storage medium | |
| RU2694271C2 (en) | Device for classifying noisy objects | |
| CN113608190B (en) | Sea surface target detection method and system based on three characteristics of singular space | |
| Santiago et al. | A novel index of abundance of juvenile yellowfin tuna in the Indian Ocean derived from echosounder buoys | |
| CN114972567A (en) | Medical ultrasonic CT multi-parameter image reconstruction method based on wave equation | |
| Pu et al. | Automated extraction of baleen whale calls based on the pseudo-Wigner–Ville distribution | |
| CN117115164B (en) | Automatic detection method and device for structural features | |
| KR102479844B1 (en) | System and meathod for recognizing underwater target | |
| Chen et al. | A method to locate spatial distribution of scattering centers from ultrasonic backscatter signal | |
| JP2603758B2 (en) | Cavitation noise detection method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MKEX | Expiry |