1 Resolution Enhancement Compression- Synthetic Aperture Focusing Techniques Student: Hans Bethe Advisor: Dr. Jose R. Sanchez Bradley University Department of Electrical Engineering

2 Motivation Ultrasonic imaging (UI) is important in medical diagnosis Figure 1: Ultrasound image of kidney stone [1]Figure 2: Ultrasound image of pancreas [2]

3 Motivation 1. One of the main concerns in UI: improving spatial resolution of ultrasonic images 2. Resolution Enhancement Compression (REC): a coded excitation (or wave- shaping) technique. Its functions: Produces a pre-enhanced chirp capable of improving axial resolution Produces a pre-enhanced chirp capable of improving axial resolution Improves SNR through compression Improves SNR through compression 3. Synthetic Aperture Focusing Techniques (SAFT): a set of beam-forming techniques, capable of improving lateral resolution and SNR by delay processing 4. Objectives: Investigate and simulate REC and SAFT independently Investigate and simulate REC and SAFT independently Combine REC and SAFT to determine the improvement in both lateral and axial directions as well as SNR Combine REC and SAFT to determine the improvement in both lateral and axial directions as well as SNR

4 Outline I. Ultrasonic imaging system II. REC III. SAFT IV. Functional requirements V. Simulation results

5 I. Ultrasonic Imaging System Figure 3: Block diagram Image reconstruction system Transducer REC (excitation) SAFT (delay processing) REC (compression)

6 Transducer An electro-mechanical device used to convert signal or energy of one form to another An electro-mechanical device used to convert signal or energy of one form to another In imaging, converts electrical signal to ultrasonic signal In imaging, converts electrical signal to ultrasonic signal Target Ultrasonic pulses Echoes Transducer Figure 4: Ultrasound emission and reflection

7 Image Reconstruction System Pre- amplifier Matched filter Echo A Delay Unit Transducer excitation A Apodization image Σ

8 Image Reconstruction System Pre- amplifier Matched filter Echo A Delay Unit Transducer excitation A Apodization image Σ

9 Image Reconstruction System Pre- amplifier Matched filter Echo A Delay Unit Transducer excitation A Apodization image Σ

10 Image Reconstruction System Pre- amplifier Matched filter Echo A Delay Unit Transducer excitation A Apodization image Σ

11 Image Reconstruction System Pre- amplifier Matched filter Echo A Delay Unit Transducer excitation A Apodization image Σ

12 Image Reconstruction System Pre- amplifier Matched filter Echo A Delay Unit Transducer excitation A Apodization image Σ

13 II. REC REC: a coded excitation and pulse compression technique REC: a coded excitation and pulse compression technique Coded excitation amounts to wave-shaping Coded excitation amounts to wave-shaping Coded excitation involves generating the pre-enhanced chirp excitation signal, capable of artificially increasing a transducer’s bandwidth => yields higher axial resolution Coded excitation involves generating the pre-enhanced chirp excitation signal, capable of artificially increasing a transducer’s bandwidth => yields higher axial resolution (axial resolution = ability of imaging system to distinguish objects closely spaced along the axis of the beam) (axial resolution = ability of imaging system to distinguish objects closely spaced along the axis of the beam) objects transducer beam Figure 5: Illustration of axial resolution beam axis

14 Pre-enhanced chirp is grounded on convolution equivalence principle

15 REC Mechanism Figure 6: Coded excitation illustration

16 REC Mechanism Figure 6: Coded excitation illustration

17 REC Mechanism Figure 6: Coded excitation illustration

18 REC Mechanism Figure 6: Coded excitation illustration

19 The 2 nd central aspect of REC is pulse-echo compression, accomplished by a Wiener compression filter γ : varies the operating point of β(f) between inverse filter and matched filter states γ : varies the operating point of β(f) between inverse filter and matched filter states V ’ lin-chirp (f): frequency spectrum of modified linear chirp. V ’ lin-chirp (f): frequency spectrum of modified linear chirp.

20 The 2 nd central aspect of REC is pulse-echo compression, accomplished by a Wiener compression filter γ : varies the operating point of β(f) between inverse filter and matched filter states γ : varies the operating point of β(f) between inverse filter and matched filter states V ’ lin-chirp (f): frequency spectrum of modified linear chirp. V ’ lin-chirp (f): frequency spectrum of modified linear chirp. The original linear chirp cannot be used in β(f) because the pre-enhanced chirp is tapered by Hanning window => convolution equivalence principle no longer holds.β(f) will yield considerable side lobes => need modified linear chirp The original linear chirp cannot be used in β(f) because the pre-enhanced chirp is tapered by Hanning window => convolution equivalence principle no longer holds. If original linear chirp is used, β(f) will yield considerable side lobes => need modified linear chirp

21 SAFT = synthetic aperture focusing techniques = different beam-forming techniques SAFT = synthetic aperture focusing techniques = different beam-forming techniques Different techniques exist: Different techniques exist:  Generic synthetic aperture (GSA)  Synthetic transmit aperture (STA)  Synthetic receive aperture (SRA)  Synthetic transmit and receive aperture (STRA) III. SAFT

22 SAFT = synthetic aperture focusing techniques = different beam-forming techniques SAFT = synthetic aperture focusing techniques = different beam-forming techniques Different techniques exist: Different techniques exist:  Generic synthetic aperture (GSA)  Synthetic transmit aperture (STA)  Synthetic receive aperture (SRA)  Synthetic transmit and receive aperture (STRA) III. SAFT

23 Σ STA emission 1emission 2 emission 3emission 4 reception 3reception 4reception 2reception 1 LRI 1LRI 2LRI 3LRI 4 Figure 7: Illustration of STA

24 The essence of STA is delay-and-sum (DAS) operation Transducer Target L1L1 L3L3 L6L6 L9L9 pulses Figure 8: Illustration of DAS

25 The essence of SAFT is delay-and-sum (DAS) operation Transducer Target L1L1 L3L3 L6L6 L9L9 pulses Figure 8: Illustration of DAS echoes

26 The essence of SAFT is delay-and-sum (DAS) operation Transducer Target L1L1 L3L3 L6L6 L9L9 pulses Figure 8: Illustration of DAS echoes

27 The essence of SAFT is delay-and-sum (DAS) operation Transducer Target L1L1 L3L3 L6L6 L9L9 Figure 8: Illustration of DAS Transducer Delay processing pulses echoes

28 The essence of SAFT is delay-and-sum (DAS) (or focusing in reception) processing Transducer Target L1L1 L3L3 L6L6 L9L9 Figure 8: Illustration of DAS Transducer Delay processing Σ pulses echoes

29 IV. Functional Requirements A/ STA Transducer shall be a linear array comprising 128 elements Transducer shall be a linear array comprising 128 elements STA shall be performed through MATLAB Field II STA shall be performed through MATLAB Field II STA mode: synthetic transmit aperture (STA) STA mode: synthetic transmit aperture (STA) Delay and sum calculations shall be performed through a GPGPU Delay and sum calculations shall be performed through a GPGPU Total synthetic aperture processing time shall be < 1 second Total synthetic aperture processing time shall be < 1 second Signal-to-noise ratio (SNR) of the images shall be at least 50 dB Signal-to-noise ratio (SNR) of the images shall be at least 50 dB

30 III. Functional Requirements B/ REC Actual impulse response h 1 (t) of transducer shall have a frequency of 2 MHz. Actual impulse response h 1 (t) of transducer shall have a frequency of 2 MHz. Transducer bandwidth shall be about 83%. Transducer bandwidth shall be about 83%. Sampling rate shall be 400 MHz. Sampling rate shall be 400 MHz. Impulse response h 2 (t) of desired transducer shall have a bandwidth about 1.5 times the bandwidth of h 1 (t). Impulse response h 2 (t) of desired transducer shall have a bandwidth about 1.5 times the bandwidth of h 1 (t). The side lobes associated with compressed pulse shall be reduced below 40 dB. The side lobes associated with compressed pulse shall be reduced below 40 dB.

31 V. Simulation results

32 Figure 9: Coded excitation simulation result (time-domain signals)

33 Figure 9: Coded excitation simulation result (time-domain signals)

34 Figure 10: Coded excitation simulation result (frequency spectra)

35 Figure 11: Original and modified linear chirp comparison

36 Figure 12: Pulse compression simulation result

37 Quality metrics used to assess REC 1/ Signal-to-noise ratio (SNR) - SNR of conventional pulsing (CP): dB - SNR of REC: dB => SNR is improved when REC is utillized 2/ Modulation transfer function: used to determine the axial resolution of the system

38 Figure 13: MTF of CP and REC k 0 = wave number at which MTF curves attains 0.1

39 => Axial improvement factor = 1.51

40 Quality metrics used to assess STA 1/ Signal-to-noise ratio (SNR) - SNR before delay processing: dB - SNR after delay processing: dB => SNR is improved after delay processing is implemented 2/ Modulation transfer function: used to determine the axial resolution of the system

41 Figure 14: Illustration of out-of-phase effect

42 Figure 15: HRI before and after delaying processing in STA

43 Figure 15: Axial resolution simulation result Pre-delay k 0 = m -1 Post-delay k 0 = m -1 Axial improvement factor = 1.11

44 QUESTIONS ?

45 References [1] Ultrasound image gallery [2] Ultrasound images gallery [3] System.htmlhttp://sell.bizrice.com/selling-leads/48391/Digital-Portable-Color-Doppler-Ultrasound- System.html [4] J. R. Sanchez et al., "A Novel Coded Excitation Scheme to Improve Spatial and Contrast Resolution of Quantitative Ultrasound Imaging" IEEE Trans Ultrasonics, Ferroelectrics, and Frequency Control, vol. 56, no. 10, pp , October [5] S. I. Nikolov, “Synthetic Aperture Tissue and Flow Ultrasound Imaging [6] [6] T. Misaridis and J. A. Jensen, “Use of Modulated Excitation Signals in Medical Ultrasound” IEEE Trans. Ultrason. Ferroelectr. Freq. Control, vol. 52, no. 2, February [7] M. L. Oelze, “Bandwidth and Resolution Enhancement Through Pulse Compression”, IEEE Trans. Ultrasonics, Ferroelectrics, and Frequency Control, vol. 54, no. 4, April 2007.

46 References [8] J. R. Sanchez and M. L. Oelze, “An Ultrasonic Imaging Speckle-Suppression and Contrast-Enhancement Technique by Means of Frequency Compounding and Coded Excitation”, IEEE Trans. Ultrasonics, Ferroelectrics, and Frequency Control, vol. 56, no. 7, Julyl [9] M. Oelze, “Improved Axial Resolution Using Pre-enhanced Chirps and Pulse Compression”, 2006 IEEE Ultrasonics Symposium [10] Tadeusz Stepinski, “An Implementation of Synthetic Aperture Focusing Technique in Frequency Domain”, IEEE transactions on Ultrasonics, Ferroelectrics, and Frequency control, vol. 54, no. 7, July 2007 [11] J. A. Zagzebski, “Essentials of Ultrasound Physics’

47 Apodization 1.Process of varying signal strengths in transmission and reception across transducer 2.Reduces side lobes 3.Signal strength will become progressively weaker with increasing distance from the center 4.Control beam width => improve or degrade lateral resolution Figure 5: Illustration of apodization Center

48 II. Theoretical Background

49 Figure 7: Effect pulse duration has on axial resolution echoes objects

50 Beam width and lateral resolution Figure 5: Illustration of the effect beam width has on lateral resolution Lateral resolution = capability of imaging system to distinguish 2 closely spaced objects positioned perpendicular to the axis of ultrasound beam Larger beam width => greater likelihood of pulses covering objects => echoes from reflectors more likely to merge => degrade lateral resolution objects transducer beam beam axis