Apparatus and method for acoustic imaging using inverse scattering techniques

• US 4,662,222 A
• Filed: 12/21/1984
• Issued: 05/05/1987
• Est. Priority Date: 12/21/1984
• Status: Expired due to Term

• Current Assignee: Biotex Pharma Investments, LLC
• Original Assignee: Steven A. Johnson
• Inventors: Johnson, Steven A.
• Primary Examiner(s): Ciarlante, Anthony V.
• Application Number:
  US06/684,801
• Time in Patent Office: 865 Days
• Field of Search: 73/602, 73/607, 73/626, 73/596, 128/660
• US Class Current: 73/602
• CPC Class Codes: A61B 8/0825 for diagnosis of the breast, e.g. mammography
• CPC Class Codes: A61B 8/406 using means for diagnosing suspended breasts
• CPC Class Codes: G01S 15/8913 using separate transducers for transmission and reception
• CPC Class Codes: G01S 15/895 characterised by the transmitted frequency spectrum
• CPC Class Codes: G01S 15/8977 using special techniques for image reconstruction, e.g. FFT, geometrical transformations, spatial deconvolution, time deconvolution
• CPC Class Codes: G06T 11/006 Inverse problem, transformation from projection-space into object-space, e.g. transform methods, back-projection, algebraic methods

Abstract

An apparatus and method for reconstructing an acoustic image of an object using a central processing unit programmed to process data derived from acoustic energy that has been transmitted at multiple frequencies and scattered by the object. Electronic signals are propagated at multiple frequencies and transduced into acoustic energy waves which are propagated toward an object. Transducer receivers detect the acoustic energy waves after they have been scattered by the object and the detected acoustic energy waves are then electronically processed and input into a high-speed digital computer which may comprise a central processing unit and/or a central processing unit in combination with an array processor. Data is also prepared and input to the computer representing the incident field and the computer then reconstructs a high-quality image of the scanned object so as to produce an image having high spacial resolution and including actual internal viscous and elastic properties of the object through the use of new inverse scattering techniques used in the data processing steps.

Claims

1. A method of producing an image of an object from acoustic energy that has been transmitted through and scattered by the object, said image comprising a high and acoustic absorption at all points within the object, and said method comprising the steps of:

electronically transmitting an electric signal at multiple frequencies and transducing said electric signal at each said frequency into acoustic energy propagated toward said object from a plurality of transducer transmitter positions;

electronically processing said electrical signal to determine from said transmitter positions an incident field corresponding to said propagated acoustic energy, said incident field being stored in the memory of a central processing unit (CPU) in the form of digitized electric signals;

detecting at a plurality of transducer receiver positions said acoustic energy transmitted through and scattered by said object;

electronically processing said detected acoustic energy so as to transform said detected energy into a plurality of digitized electric signals stored in said memory of said CPU and corresponding to a scattered field detected at said transducer receiver positions;

said CPU preparing an initial estimate of the scattering potential for said object and an initial estimate of the internal field of said object at each pixel of a display screen on which said acoustic image is to be displayed and storing each said estimate in said memory;

said CPU determining at each said frequency and storing in said memory a scattered field derived at said frequency from said initial estimates of the scattering potential and internal field, and thereafter comparing at said frequency said scattered field detected at said receiver positions to said scattered field determined by said CPU;

said CPU determining at each said frequency and storing in said memory a new estimate of said internal field at each said pixel derived from said incident field, the last estimate of the internal field at each said pixel, and the last estimate of said scattering potential at each said pixel, said new estimate of the internal field at said frequency comprising all orders of scattering;

said CPU determining and storing in said memory a new estimate of said scattering potential derived from said new estimate of said internal field and said scattered field detected at said receiver positions;

said CPU continuing to update the estimate for said internal field and scattering potential until said scattered field determined by said CPU approximates said scattered field detected at said receiver positions within a selected range of tolerance; and

said CPU thereafter using the determined scattering potential to reconstruct and store said image in said CPU memory.

2. A method as defined in claim 1 wherein said step of electronically transmitting said electric signal at multiple frequencies comprises the steps of:

positioning a transducer array adjacent said object, said array comprising a plurality of acoustic transmitters and acoustic receivers;

sending said electric signal at a first frequency to each said transmitter so that each said transmitter will in turn propagate acoustic energy at said first frequency; and

thereafter changing the frequency of said signal and sending said electrical signal at said changed frequency to each said transmitter so as to sequentially propagate acoustic energy from said said transmitter at said changed frequency.

3. A method as defined in claim 1 wherein said step of electronically transmitting an electric signal at multiple frequencies comprises the steps of:

positioning a transducer array adjacent to said object, said array comprising a plurality of acoustic transmitters and a plurality of acoustic receivers;

generating said electric signal in the form of a waveform which is characterized by a plurality of different frequencies; and

sending said generated waveform in turn to each said transmitter so as to propagate acoustic energy at said multiple frequencies from each said transmitter.

4. A method as defined in claims 2 or 3 wherein said transducer array is configured to encircle said object.

5. A method as defined in claim 1 wherein said step of detecting at a plurality of transducer-receiver positions said acoustic energy transmitted through and scattered by said object comprises the steps of:

positioning a transducer array adjacent said object, said array comprising a plurality of transmitters and a plurality of receivers; and

after acoustic energy is transmitted from one of said transmitters, sequencing each said receiver so as to detect said scattered acoustic energy at each said receiver in turn.

6. A method as defined in claim 5 wherein said step of electronically processing said detected acoustic energy comprises the steps of:

transducing the acoustic energy detected by each said receiver transducer into a corresponding electric signal;

amplifying said corresponding electric signal; and

thereafter processing each said amplified signal so as to generate two signals which correspond to mathematical real and imaginary representations of each said amplified signal.

7. A method as defined in claim 6 wherein the step of processing said amplified signal from each said receiver transducer so as to generate said signals corresponding to said mathematical real and imaginary representations of said amplified signal comprises the steps of:

inputting the amplified signal detected at said receiver transducer to first and second multiplier circuits and multiplying the amplified signal input to said first multiplier circuit by the electric signal sent to said transmitter transducers;

shifting the phase of said electric signal sent to said transmitter transducers by 90° and

thereafter multiplying the amplified signal input to said second multiplier circuit by said signal that is shifted by 90°

; and

filtering the output of each said multiplier circuit with a low-pass filter circuit and thereafter integrating and digitizing the output of each said low-pass filter circuit.

8. A method as defined in claim 6 wherein said step of processing said amplified signals from each said receiver transducer so as to generate said signals corresponding to said mathematical real and imaginary representations of each said amplified signal comprises the steps of:

inputting each said amplified signal to a high speed analog-to-digital converter so as to digitize each said amplified signal; and

inputting each said digitized signal from said high speed analog-to-digital converter into a parallel processor programmed to take the complex fast Fourier transform of each said digitized signal.

9. A method as defined in claim 5 wherein said transducer array is configured to encircle said object.

10. A method as defined in claim 1 wherein said initial estimate of said scattering potential is zero.

11. A method as defined in claim 1 wherein said initial estimate of said scattering potential is an average value determined by an estimated average of density, acoustic speed and acoustic absorption of said object.

12. A method as defined in claim 1 wherein said incident field is used as said initial estimate of the internal field of said object at each said pixel.

13. A method as defined in claim 1 wherein said estimate of the internal field of each said object at each said pixel is determined by said CPU from said initial estimate of said scattering potential.

14. A method as defined in claim 1 wherein said step of determining said new estimate of said scattering potential comprises the steps of:

backprojecting said scattered field detected at said transducer-receiver positions to obtain a blurred image of said scattering potential of said object; and

deblurring said blurred image of said scattering potential to obtain said new estimate of the scattering potential.

15. A method as defined in claim 14 wherein said incident field, said scattered field detected at said transducer-receiver positions, each said initial and new estimate of said internal field and each said initial and new estimate of said scattering potential are obtained separately by said CPU at each said frequency.

16. A method as defined in claim 15 wherein said scattering potential is formulated using a plurality of frequency-independent components represented as a vector Γ

multiplied by a frequency-dependent matrix M, and wherein a least-squares method is applied by said CPU to the product MΓ

so as to derive therefrom a scattering potential comprised of frequency-dependent terms whose sum represents a best fit to said deblurred image of said scattering potential.

17. A method as defined in claim 16 wherein said least-squares method is applied by said CPU using a conjugate gradient method.

18. A method as defined in claim 1 wherein said incident field, said scattered field detected at said transducer-receiver positions, and each said initial and new estimate of said internal field are obtained separately by said CPU at each said frequency.

19. A method as defined in claim 18 wherein said new estimate of said internal field and said new estimate of said scattering potential are obtained by said CPU according to a conjugate gradient method.

20. A method as defined in claim 19 wherein said scattering potential obtained by said CPU comprises a plurality of frequency-independent components represented as a vector Γ

.

21. A method as defined in claim 19 wherein said conjugate gradient method is used by said CPU to determine said new estimates of said internal field and said scattering potential by computing a descent direction wherein each variable from a set of variables taken from a first vector equation defining said internal field and a second vector equation defining said scattered field detected at said receiver positions is allowed to vary simultaneously.

22. A method of reconstructing an acoustic image of an object using a central processing unit (CPU) programmed to process data derived from acoustic energy that has been transmitted at multiple frequencies and scattered by said object, said method comprising the steps of:

propagating multiple frequency acoustic energy waves toward said object from a plurality of transmitters positions;

electronically storing a plurality of digitized electronic signals derived by said CPU from said propagated acoustic energy and said transmitter positions, said digitized electronic signals representing an incident field obtained at each said frequency

detecting at a plurality of receivers multiple frequency acoustic energy waves scattered by said object;

electronically storing a plurality of digitized electronic signals representing all orders of a scattered field obtained at each said frequency and derived by said CPU from the scattered acoustic energy waves detected at said receivers;

said CPU determining (a) an estimate of an internal field at each said frequency and at each scattering point within said object, and (b) an estimate of a scattering potential characterized by density, acoustic speed and acoustic absorption at each said frequency and at each scattering point within said object, and said CPU electronically storing a plurality of digitized electronic signals representing said estimates of said internal field and said scattering potential;

said CPU determining from said estimated internal field and scattering potential a predicted scattered field at all orders of scattering and at each said frequency, and said CPU comparing said predicted scattered field to said scattered field derived from the scattered acoustic energy waves detected at said receivers;

said CPU updating said estimates of said internal field and said scattering potential until said comparison results in a predicted scattered field which approximates said scattered field derived from the scattered acoustic energy waves detected at said receivers within a selected range of tolerance; and

said CPU thereafter reconstructing from said estimated scattering potential an acoustic image of said object, and outputting a visually perceptible display of said image.

23. A method as defined in claim 22 wherein said step of said CPU updating said estimate of said internal field comprises the step of said CPU deriving said updated estimate of said internal field from said incident field and from said estimate of said scattering potential.

24. A method as defined in claim 23 wherein said step of said CPU updating said estimate of said scattering potential comprises the step of said CPU deriving said updated estimate of said scattering potential from said updated estimate of said internal field and from said scattered field derived from the scattered acoustic energy waves detected at said receivers.

25. A method as defined in claim 24 wherein said step of said CPU updating said estimate of said scattering potential further comprises the steps of:

backprojecting said scattered field derived from the scattered acoustic energy waves detected at said receivers to obtain a blurred image of said updated scattering potential; and

deblurring said blurred image.

26. A method as defined in claim 25 further comprising the steps of:

said CPU formulating a frequency-independent vector Γ

comprised of a plurality of frequency-independent components;

said CPU setting said updated estimate of said scattering potential equal to the product of said vector and a frequency-dependent matrix M; and

said CPU applying a least-squares method to said product MΓ

so as to derive therefrom a product comprising a plurality of frequency-dependent terms whose sum represents a best fit to said deblurred image.

27. A method as defined in claim 23 wherein said updated estimate of said internal field is obtained by said CPU according to a conjugate gradient method.

28. A method as defined in claim 27 further comprising the steps of:

said CPU formulating a frequency-independent vector Γ

comprised of a plurality of frequency-independent components;

said CPU applying a conjugate gradient method to obtain an updated estimate of said vector from said updated internal field and from said scattered field derived from the scattered acoustic energy waves detected at said receivers; and

said CPU determining said updated estimate of said scattering potential from the product of said updated vector and a frequency-dependent matrix M.

29. A method as defined in claim 28 wherein said conjugate gradient method is used by said CPU to determine said updated estimates of said internal field and said vector by computing a descent direction wherein each variable from a set of variables defined by (a) a first set of equations defining said internal field and (b) a second set of equations defining said scattered field detected at said receivers is allowed to simultaneously vary.

30. A method as defined in claim 22 wherein said step of propagating said multiple frequency acoustic energy waves comprises the step of electronically transmitting an electric signal at multiple frequencies and transducing said electric signal at each set frequency into said acoustic energy waves.

31. A method as defined in claim 30 wherein said step of electronically transmitting said electric signal at multiple frequencies comprises the steps of:

positioning a transducer array adjacent said object, said array comprising a plurality of acoustic transmitters;

sending said electric signal at a first frequency to each said transmitter so that each said transmitter will in turn propagate acoustic energy at said first frequency; and

thereafter changing the frequency of said signal and sending said electrical signal at said changed frequency to each said transmitter so as to sequentially propagate acoustic energy from each said transmitter at said changed frequency.

32. A method as defined in claim 30 wherein said step of electronically transmitting an electric signal at multiple frequencies comprises the steps of:

positioning a transducer array adjacent to said object, said array comprising a plurality of acoustic transmitters;

generating said electric signal in the form of a waveform which is characterized by a plurality of different frequencies; and

sending said generated waveform in turn to each said transmitter so as to propagate acoustic energy at said multiple frequencies from each said transmitter.

33. A method as defined in claims 30 or 31 wherein said transducer array is configured to encircle said object.

34. A method as defined in claim 22 wherein said step of detecting at said plurality of receivers said multiple frequency acoustic energy waves scattered by said object comprises the steps of:

positioning a transducer array adjacent said object, said array comprising a plurality of acoustic receivers;

sequencing each said receiver so as to detect said scattered acoustic energy waves at each said receiver in turn; and

electronically processing said detected acoustic energy waves so as to transform said detected acoustic energy waves into a plurality of digitized electric signals.

35. A method as defined in claim 34 wherein said step of electronically processing said detected acoustic energy waves comprises the steps of:

transducing said acoustic energy waves detected by each said receiver into a corresponding electrical signal;

amplifying said corresponding electric signal; and

thereafter processing each said amplified signal so as to generate two signals which correspond to mathematical real and imaginary representations of each said amplified signal.

36. A method as defined in claim 35 wherein said step of processing each said amplified signal so as to generate said signals corresponding to said mathematical real and imaginary representations of said amplified signal comprises the steps of:

inputting each said amplified signal to first and second multiplier circuits and multiplying the amplified signal input to said first multiplier circuit by the electric signal sent to said transmitter transducers;

shifting the phase of said electric signal sent to said transmitter transducers by 90° and

thereafter multiplying the amplified signal input to said second multiplier circuit by said signal that is shifted by 90°

; and

filtering the output of each said multiplier circuit with a low-pass filter circuit and thereafter integrating and digitizing the output of each said low-pass filter circuit.

37. A method as defined in claim 35 wherein said step of processing each said amplified signals so as to generate said signals corresponding to said mathematical real and imaginary representations of each said amplified signal comprises the steps of:

inputting each said amplified signal to a high speed analog-to-digital converter so as to digitize each said amplified signal; and

inputting each said digitized signal from said high speed analog-to-digital converter into a parallel processor programmed to take the complex fast Fourier transform of each said digitized signal.

38. A method as defined in claim 34 wherein said transducer array is configured to encircle said object.