High-frequency ultrasonic imaging of the anterior segment using an annular array transducer

Abstract

Objective: Very high-frequency ultrasound (VHFU; >35 megahertz [MHz]) allows imaging of anterior segment structures of the eye with a resolution of less than 40 microm. The low focal ratio of VHFU transducers, however, results in a depth of field (DOF) of less than 1 mm. The aim was to develop a high-frequency annular array transducer for ocular imaging with improved DOF, sensitivity, and resolution compared with conventional transducers.

Design: Experimental study.

Participants: Cadaver eyes, ex vivo cow eyes, in vivo rabbit eyes.

Methods: A spherically curved annular array ultrasound transducer was fabricated. The array consisted of 5 concentric rings of equal area, had an overall aperture of 6 mm, and a geometric focus of 12 mm. The nominal center frequency of all array elements was 40 MHz. An experimental system was designed in which a single array element was pulsed and echo data were recorded from all elements. By sequentially pulsing each element, echo data were acquired for all 25 transmit-and-receive annuli combinations. The echo data then were focused synthetically and composite images were produced. Transducer operation was tested by scanning a test object consisting of a series of 25-microm diameter wires spaced at increasing range from the transducer. Imaging capabilities of the annular array were demonstrated in ex vivo bovine, in vivo rabbit, and human cadaver eyes.

Main outcome measures: Depth of field, resolution, and sensitivity.

Results: The wire scans verified the operation of the array and demonstrated a 6.0-mm DOF, compared with the 1.0-mm DOF of a conventional single-element transducer of comparable frequency, aperture, and focal length. B-mode images of ex vivo bovine, in vivo rabbit, and cadaver eyes showed that although the single-element transducer had high sensitivity and resolution within 1 to 2 mm of its focus, the array with synthetic focusing maintained this quality over a 6-mm DOF.

Conclusions: An annular array for high-resolution ocular imaging has been demonstrated. This technology offers improved DOF, sensitivity, and lateral resolution compared with single-element fixed focus transducers currently used for VHFU imaging of the eye.

Figures

Figure 1

Schematic illustrating operation of a curved annular array. The figure shows a three-element array in cross-section and en face (insert, upper left). Left: When all the array elements are excited simultaneously, the wavefronts from each element will converge at the geometric focus. Right: Here, the outermost ring is excited first and the central ring last, resulting in convergence of wavefronts at a synthetic focal point closer than the geometric focus.

Figure 2

Photograph of prototype annular array and flex circuit containing the electronic fan-out.

Figure 3

During scanning, the crosspoint switch directs excitation pulses to only one of the five transducer elements at a time (bold arrow) while echo data are collected from all five channels. To acquire all 25 possible send/receive pairs needed to form an image, the transducer is scanned five times, once for each excited element.

Figure 4

B-mode images of a series of wire targets at increasing ranges about the geometric focus with (a) a focused single-element transducer, (b) the array without synthetic focusing and (c) the array with synthetic focus processing. Where synthetic focusing is not used, the array functions as the equivalent of a conventional single-element transducer. Note the improved lateral resolution obtained outside the geometric focus when synthetic focusing is used. Also note the improvement in dynamic range (reduced noise) in the array compared to the single-element transducer.

Figure 5

Central cornea and anterior lens surface of an ex vivo bovine eye scanned with the annular array with (a) no delay corrections and (b) synthetic focusing (41 focal zones, 180 μm/zone). The interfaces seen in the image include (from left to right) the epithelial surface, Bowman’s membrane, the posterior corneal surface, and the anterior lens surface. The position of the geometric focus at 12 mm in indicated by white triangles. The signal-to-noise ratios (SNR’s) were (a) 26 dB and (b) 34 dB. Note the much improved depiction of the anterior cornea (near field) and lens surface (far field) with synthetic focus.

Figure 6

Anterior segment images of an ex vivo bovine eye showing data from a) a focused single-element transducer, b) the array with no delay corrections and c) the array with synthetic focusing (41 focal zones, 120 μm/zone). Note the improved depiction of the iris and lens surface (L) with synthetic focus. The signal-to-noise ratios were (a) 41 dB, (b) 45 dB and (c) 51 dB. The white triangles indicate the geometric focus.

Figure 7

In vivo anterior segment scan of a rabbit with (a) no delay corrections and (b) synthetic focusing (41 focal zones, 177 μm/zone). The signal-to-noise ratios were (a) 35 dB and (b) 41 dB. The white triangles indicate the geometric focus.

Figure 8

Images of the anterior segment of a cadaver eye showing data from (a) the center array element, (b) the full array with no delay corrections and, (c) the full array with synthetic focusing (41 focal zones, 177 μm/zone). The signal-to-noise ratios were (a) 43 dB, (b) 43 dB, and (c) 51 dB. The white triangles indicate the position of the geometric focus.

Figure 9

Series of anterior segment images of a cadaver eye with the geometric focus (white triangles) progressively incremented by 1-mm axial steps. The top row (a) shows the summed data with no synthetic focusing. The bottom row (b) shows the data after synthetic focusing (41 focal zones, 177 mm/zone). Note how image quality is maintained as the range between the transducer and the eye changes when synthetic focusing is used.