Research ArticleCancer Imaging

Simultaneous transrectal ultrasound and photoacoustic human prostate imaging

See allHide authors and affiliations

Science Translational Medicine  28 Aug 2019:
Vol. 11, Issue 507, eaav2169
DOI: 10.1126/scitranslmed.aav2169
  • Fig. 1 Schematics and photographs of TRUSPA imaging of the human prostate.

    (A) Schematic representation of transrectal imaging of prostate (P) using the TRUSPA device. (B) Schematics of the distal end of the TRUSPA device and its cross section showing key components. PCB, printed circuit board; PDMS, polydimethylsiloxane; CMUT, capacitive micromachined ultrasonic transducer array; ASICs, application-specific integrated circuits. (C) Photograph of the TRUSPA device with a 23-mm scale bar. (D) Magnified photograph showing the distal end of the device that is inserted into the rectum of the patient. The three dark lines around three sides of the device are the output end of the optical fibers that deliver light into the prostate from three different planes [the red colored planes shown in (B)]. The device is encapsulated with a gray-color PDMS lens (yellow-dotted rectangular box) above the CMUT surface to achieve elevation focusing. (E and F) Images of the front (E) and back (F) sides of the custom-made PCB, underneath the PDMS lens, facilitating close bonding of the CMUT array with four ASICs. Figures S1 to S3 provide complete details of the TRUSPA imaging system.

  • Fig. 2 Evaluating structural imaging capabilities and co-registration accuracy using a deep-tissue phantom.

    (A) Photograph showing the side view of a prostate tissue–mimicking intralipid phantom covered with a ~25-mm thick porcine tissue. The schematic positions of all nine fishing wire targets of 0.3 mm diameter, shown as dots, are as follows: 1 (12 mm, 20°), 2 (14 mm, 10°), 3 (16 mm, 0°), 4 (18 mm, −10°), 5 (20 mm, −20°), 6 (22 mm, −20°), 7 (22 mm, −10°), 8 (22 mm, 0°), and 9 (22 mm, 10°). The unit distance on the z axis is 10 mm. Blue dots represent wires painted black as shown in (B). (B) Photograph of an empty phantom tank before it is filled with intralipid solution. US, PA, and a co-registered US + PA images of the wire targets (C to E) inside the intralipid phantom and (F to H) when imaged through the porcine tissue. Yellow arrows point to US signals generated at the phantom edges. Plots of edge spread functions of target 1 (at 37 mm) along (I) axial and (J) lateral directions, and for target 9 (at 52 mm) along (K) axial and (L) lateral directions. a.u., arbitrary units. Scale bars, 10 mm (C to H).

  • Fig. 3 TRUSPA studies on ex vivo intact human prostates after radical prostatectomy.

    (A) Photograph of a polyethylene fiducial tube (0.8 mm diameter) placed inside the urethra of an excised human prostate. Encircled regions in US (grayscale), PA (red color), and co-registered US + PA images show respective contrasts at 800-nm wavelength from (B to D) a blood-filled tube with a depth of ~2 cm inside the prostate, (E to G) a blood-filled tube placed behind the prostate covering an imaging depth of ~3.5 cm, and (H to J) a tube filled with ICG solution (1 mg/ml) placed behind the prostate. (K) Intact human prostate ex vivo showing the schematic orientation of the TRUSPA device when imaging the PIRADS 4 lesion (encircled region) in the right lateral PZ measuring 1.1 cm in diameter. (L) Co-registered US + PA image of the peripheral lesion (encircled). (M) Preoperative axial T2-weighted 3-T MRI showing low-intensity mass (encircled) in the right lateral PZ. B, bladder; P, prostate; R: rectum. Scale bar, 10 mm. (N) Histological tissue section from the peripheral lesion showing high cell proliferation (Gleason grades 3 and 4) and evidence of vasculature. Scale bar, 50 μm. (O and P) Edge spread functions along the axial and lateral directions of the blood tube in (C) demonstrating resolutions (half the distance of X10–90) of about 215 and 720 μm, respectively. (Q) Spectral plot of the mean PA intensity of the ICG tube in (I) in the optical wavelength range of 750 to 950 nm, in steps of 25 nm. Scale bars, 10 mm (B to L).

  • Fig. 4 In vivo imaging of a mouse prostate tumor using intravenous ICG.

    (A) Photograph of a mouse with a subcutaneous prostate tumor (PC3 cells expressing luciferase). The tumor is encircled in white in all images. (B) Co-registered bioluminescence (color) and optical (grayscale) images of the mouse acquired using an IVIS 200 imaging system after intraperitoneal administration of d-luciferin. (C) TRUSPA device was placed over the subcutaneous tumor using US gel during in vivo imaging. Co-registered US in grayscale and PA in color images of (D) pre-ICG and (E) 5-min post-ICG injection. Scale bars, 5 mm (D and E). Mean PA signal of five regions of interests [ROIs; defined in (O)] was calculated and plotted as a function of wavelength for (F) pre-ICG injection PA images shown in fig. S10A and (G) post-ICG injection PA images shown in fig. S10B. For comparison, the standard plots of molar extinction coefficient as a function of wavelength (taken from literature, http://omlc.org/spectra/) are plotted for (H) deoxyhemoglobin (Hb), oxyhemoglobin (HbO2), and a mixture of 30% Hb and 70% HbO2, and (I) a mixture of 29.75% Hb + 69.75% HbO2 + 0.5% ICG. Spectrally unmixed images of Hb, HbO2, and ICG (J to L) before ICG injection and (M to O) 5 min after ICG injection. Scale bars, 10 mm (J to O).

  • Fig. 5 In vivo TRUSPA imaging of human prostate.

    Each TRUSPA frame in (A) to (C) consists of US, PA, and co-registered US + PA images of human prostate. Scale bars, 10 mm. Prostate (P), rectum (R), bladder (B), urethra (U), peripheral zone (PZ), transition zone (TZ), seminal vesicle (SV), neurovascular bundle (NVB), anterior fibromuscular stroma (AFS), dorsal vascular complex (DVC), levator ani fascia (LAF), and parietal endopelvic fascia (PEF) were identified in these images. (A) PA contrast from the NVB (~20 mm depth) in the posterior PZ. (B) PA contrast from vasculature structures surrounding SV (~15 mm depth) and from DVC (~40 mm depth) that spans AFS, TZ, and PZ that is connected to the bladder neck. (C) PA contrast from a suspicious (white arrow) hypoechoic mass in the PZ in the left base of the prostate.

  • Fig. 6 In vivo multimodal PET, MRI, TRUS, and TRUSPA imaging of the prostate in a patient with PCa.

    In all images, the rectum (R), rectal wall (RW), bladder (B), anterior fibromuscular stroma (AFS), and prostate (P; green contour) are labeled. (A to C) Ultrasound (US) in grayscale, PA in red color scale, and co-registered US + PA images of human prostate obtained in vivo with the TRUSPA device. Movie S1 shows real-time TRUSPA imaging of this patient, which involved linear and rotational movements of the device in the rectum to scan different regions of the prostate. The suspicious region with distinct PA contrast [yellow contour in (A) to (C)] in the right base was repeatedly visited (around 25 to 35 s in the video) during the imaging session. (D) Axial PET imaging showing PCa (yellow contour) using 68Ga-labeled radioactive tracer targeting bombesin receptor on the PCa cells. (E) Axial MRI showing anatomical information of the prostate with yellow contour covering the extent of PCa identified using PET molecular imaging. The TRUSPA FOV shown in (A) to (C) is also marked on the MRI (blue-shaded triangular region). (F) Axial TRUS image showing targeted region (yellow contour) for biopsy using the data from both MRI and PET; targeted biopsy confirmed PCa. (G) Final histopathology from the prostatectomy showing areas of hypervascularity (arrows) within the tumor. Scale bar, 10 mm.

  • Fig. 7 Contrast-enhanced TRUSPA imaging of human prostate using intravenous ICG.

    A 53-year-old patient was diagnosed with PCa based on elevated PSA (5.33 ng/ml) and TRUS-guided biopsy showing Gleason 3 + 3 cancer in the left base of the prostate. Bladder (B), rectum (R), and prostate (P, green contour) are marked. (A) Diffusion-weighted MRI showing the malignancy (red contour) in the left PZ of the prostate. (B) Sagittal TRUS showing the MRI-based fused malignant (red) and control (yellow) regions used for the targeted biopsy. (C) Sagittal view of the 3D volume rendered patient’s prostate based on MRI. Also shown is the schematic FOV of the TRUSPA device when imaging the peripheral lesion outlined in red in the left base with intravenous ICG (25 mg; 10 ml at 2.5 mg/ml). (D) Pre-ICG, (E) 2-min post-ICG, and (F) 6-min post-ICG images showing US (grayscale), PA (red), and co-registered US + PA images of the prostate. (G) Coefficient of first PCA, discussed in detail in fig. S11, plotted as a function of time relative to ICG injection. (H) Mean PA intensity in RO1 and RO2 as a function of wavelength plotted for pre- and post-ICG imaging periods. Spectrally unmixed ICG image (I) before and (J) after injecting ICG. (K) Pathology from prostatectomy showing a Gleason 3 + 4 cancer with extraprostatic extension and no cancer in lymph nodes, staged pt3aN0. Scale bars, 10 mm (A to F, I, and J). Yellow stickers on the red contour that surrounds the tumor region inside the tissue specimen show the dimensions 15.99 mm × 9.92 mm marked by the pathologist. (L) Magnification of a region inside the tumor outlined by red contour in (K). Scale bars, 1 mm (K) and 100 μm (L).

  • Table 1 Statistical analysis on seven patients who received intravenous ICG during the TRUSPA imaging.

    Slopes for change in the mean signal of the entire prostatic region between pre- and post-ICG injection for the measurements 800 nm (peak absorption of ICG), 950 nm (very low absorption of ICG), and unmixed ICG. Total of n = 70 measurements adjusted for dose and clustering within patient were used in the analysis. The fourth column “is slope = 0?” provides the probability for no change in signal difference between respective pre- and post-measures, and the fifth column “is slope = ICG slope?” provides the probability that the slope of a certain measurement (800 or 950 nm) follows the ICG slope.

    Slope
    measure
    Slope
    estimate
    95%
    confidence
    interval for
    the estimated
    slope
    Slope = 0?
    P
    Slope = ICG
    slope?
    P
    800 nm3.5−1.7 to 8.80.1520.578
    950 nm0.9−3.5 to 5.20.6390.110
    Unmixed ICG4.91.8 to 8.10.009

Supplementary Materials

  • stm.sciencemag.org/cgi/content/full/11/507/eaav2169/DC1

    Materials and Methods

    Fig. S1. Description of the TRUSPA imaging system next to the patient bed in the urology clinic.

    Fig. S2. Schematics and images that describe the TRUSPA device, operating principle, and its data acquisition.

    Fig. S3. Images of the CMUT array and ASIC.

    Fig. S4. Simulated output pressure of a CMUT cell and experimental impedance measurements of a single CMUT element.

    Fig. S5. Design and characterization of PDMS lens on the CMUT array.

    Fig. S6. Time sequence used for simultaneous US and PAI of the TRUSPA device.

    Fig. S7. Characterizing the US field of the TRUSPA device using simulations and experiments.

    Fig. S8. Output pressure of the TRUSPA device, recorded by hydrophone in immersion, as a function of different DC and AC bias voltage settings.

    Fig. S9. Characterization of TRUSPA system SNR as a function of depth and wavelength.

    Fig. S10. Multiwavelength PA images of the mouse prostate tumor imaged with intravenous ICG.

    Fig. S11. Multi-ROI time activity of ICG for the patient case presented in Fig. 7.

    Fig. S12. Multiwavelength PA images of human prostate for the patient case presented in Fig. 7.

    Fig. S13. Analysis of ICG activity during in vivo TRUSPA imaging of a human patient with PCa intravenously administered 75 mg of ICG at a concentration of 2.5 mg/ml.

    Fig. S14. Analysis of ICG activity during in vivo TRUSPA imaging of a human patient with PCa intravenously administered 5 mg of ICG at a concentration of 2.5 mg/ml.

    Table S1. 1D (linear) CMUT array parameters.

    Table S2. Typical deep-tissue imaging parameters of the TRUSPA device.

    Table S3. Intravenous ICG dose given to 10 human subjects at a concentration of 2.5 mg/ml.

    Movie S1. In vivo TRUSPA imaging of human prostate in clinic (without administering contrast agent).

  • The PDF file includes:

    • Materials and Methods
    • Fig. S1. Description of the TRUSPA imaging system next to the patient bed in the urology clinic.
    • Fig. S2. Schematics and images that describe the TRUSPA device, operating principle, and its data acquisition.
    • Fig. S3. Images of the CMUT array and ASIC.
    • Fig. S4. Simulated output pressure of a CMUT cell and experimental impedance measurements of a single CMUT element.
    • Fig. S5. Design and characterization of PDMS lens on the CMUT array.
    • Fig. S6. Time sequence used for simultaneous US and PAI of the TRUSPA device.
    • Fig. S7. Characterizing the US field of the TRUSPA device using simulations and experiments.
    • Fig. S8. Output pressure of the TRUSPA device, recorded by hydrophone in immersion, as a function of different DC and AC bias voltage settings.
    • Fig. S9. Characterization of TRUSPA system SNR as a function of depth and wavelength.
    • Fig. S10. Multiwavelength PA images of the mouse prostate tumor imaged with intravenous ICG.
    • Fig. S11. Multi-ROI time activity of ICG for the patient case presented in Fig. 7.
    • Fig. S12. Multiwavelength PA images of human prostate for the patient case presented in Fig. 7.
    • Fig. S13. Analysis of ICG activity during in vivo TRUSPA imaging of a human patient with PCa intravenously administered 75 mg of ICG at a concentration of 2.5 mg/ml.
    • Fig. S14. Analysis of ICG activity during in vivo TRUSPA imaging of a human patient with PCa intravenously administered 5 mg of ICG at a concentration of 2.5 mg/ml.
    • Table S1. 1D (linear) CMUT array parameters.
    • Table S2. Typical deep-tissue imaging parameters of the TRUSPA device.
    • Table S3. Intravenous ICG dose given to 10 human subjects at a concentration of 2.5 mg/ml.
    • Legend for movie S1

    [Download PDF]

    Other Supplementary Material for this manuscript includes the following:

    • Movie S1 (.avi format). In vivo TRUSPA imaging of human prostate in clinic (without administering contrast agent).

Stay Connected to Science Translational Medicine

Navigate This Article