Large-Field 3-D Microwave-Induced Thermoacoustic Imaging Using Transrectal Linear-Array Combined Self-Rotating Scanning and Spatial Aperture Angle Suppression

Microwave-induced thermoacoustic imaging (MITAI) uses long-wavelength microwave excitation to achieve deeper tissue penetration and a larger illumination field than laser-based methods. Intracavitary MITAI is particularly attractive because microwave excitation can be delivered extracorporeally while reusing conventional ultrasound probes, making the three-dimensional (3-D) intracavitary MITAI conceptually simple and highly feasible. However, 3-D MITAI based on a self-rotating transrectal linear-array probe suffers from severe artifacts caused by the limited receive aperture in both the rotational and elevational directions, which can drastically degrade image quality. To address this problem, we design a 3-D MITAI framework that employs a transrectal intracavitary linear-array probe performing 360$^\circ$self-rotation, and we propose two 3-D image reconstruction strategies: (1) a fast method based on two-dimensional (2-D) slice post-processing and direct 3-D point-cloud mapping, in which, for each rotation angle, the multi-channel radio-frequency (RF) data are independently reconstructed using 2-D delay-and-sum (DAS) combined with band-pass filtering, depth window gating, dynamic aperture control, coherence-factor (CF) weighting, ring-buffer detection, and multi-threshold gating, after which the cleaned pixels in the$r$–$z$cross-sectional plane are geometrically mapped into 3-D voxels or point clouds to achieve efficient 3-D reconstruction; and (2) a full-wavefield method that treats the multi-angle data acquired during rotation as a virtual 2-D cylindrical aperture, explicitly computes the propagation delays from each voxel in 3-D space to all virtual elements, and performs 3-D DAS beamforming to obtain a volumetric dataset with effective suppression of limited-aperture artifacts. Preliminary experimental results demonstrate that the proposed system achieves lateral resolutions of 0.924 mm and 0.979 mm and axial resolutions of 0.737 mm and 0.966 mm for the two methods, respectively. Both strategies successfully realize 360$^\circ$3-D imaging, and with GPU acceleration, the full-wavefield 3-D reconstruction time is reduced from 2193 s to approximately 192 s. The proposed approach provides a practical solution for transrectal 3-D MITAI and, when combined with efficient large-area microwave illumination and potentially with conventional ultrasound readout, has strong potential for multimodal intracavitary applications such as rectum and surrounding vasculature imaging, where complementary acoustic structural and electromagnetic functional information could improve diagnostic accuracy.