Answers are grounded in this paper and cite their sources.
Mechanically scanned linear arrays provide a practical route to three-dimensional photoacoustic imaging, but a finite elevation response causes the same target region to contribute to several neighboring B-scans. A smaller scan step therefore increases sampling density without, by itself, removing elevation-direction mixing. Here, fine-step scans are formulated as overlapping observations governed by an experimentally measured post-beamforming elevation response. A 0.05 mm wire was used to identify a 15-point response for a 128-element, 20 MHz array scanned at 0.1 mm increments; this response was embedded in an extended-grid forward operator, and the resulting non-negative composite-regularized inverse problem was solved using the alternating direction method of multipliers. In an independent 0.08 mm crossed-wire experiment, the median elevation full widths at half maximum for delay-and-sum (DAS), Wiener, Richardson–Lucy, and ADMM reconstructions were 0.854, 0.665, 0.630, and 0.590 mm, respectively. ADMM achieved a median paired FWHM reduction of 26.3% relative to DAS, with a Peak SNR of 23.30 dB and a PSLR of −8.66 dB. Finger-vessel imaging showed reduced apparent broadening and structural merging. These results show that explicitly modeling experimentally identified inter-scan coupling can improve elevation localization using information already present in conventional fine-step scans without changing the transducer aperture.
No takes yet. Share an insight, caveat, or question.
Li et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: