Wavefield-Aware quality control and shear wave speed estimation in ocular Vibro-Elastography

To investigate the wavefield complexity in complex and confined structures such as ocular tissues, and to develop methods to correctly analyze the shear wave speed in such structures. We aim to establish a wavefield-aware framework for analyzing intraocular lesions using ultrasound vibro-elastography (UVE).

Wavefield-aware UVE was evaluated using numerical simulations, tissue-mimicking phantoms, and in vivo patient data. Three representative approaches: spectral k-ω method, phase-gradient method (DPG/TPI-type), and local phase velocity imaging with wavefield-based filtering (LPVI-WBF) were applied to analyze shear wave speed (SWS) and their results were compared. Wavefield quality control (QC) based on directional and spectral energy metrics was proposed to identify regions dominated by traveling waves. Lesion-level SWS was summarized within quality controlled and confined ROIs. Multifrequency wave generation and acquisitions (100-200 Hz) enabled analysis of shear-wave dispersion and frequency-domain mechanical heterogeneity.

Simulations and phantom experiments showed that k-ω and DPG/TPI-type estimators are highly sensitive to wave reflections, standing waves, and modal interference, resulting in unstable or non-physical SWS estimates under these complex wavefield conditions. Wavefield-based QC combined with LPVI-WBF enabled stable SWS mapping within lesion ROIs. In vivo ROI-mean SWS values were similar between melanoma and nevus lesions. However, multifrequency analysis revealed distinct biomechanical signatures: melanoma lesions exhibited significantly stronger dispersion slopes (dc/df) and higher frequency-domain heterogeneity indices (H_f) than nevus lesions (Mann-Whitney test, p<0.01).

Wavefield-aware UVE analysis provides sensitive biomarkers using dispersion curve strength and inter-frequency heterogeneity to separate melanoma lesions from nevus lesions, while absolute SWS alone does not distinguish them.