Passive cavitation imaging (PCI) is essential for real-time monitoring and control of focused ultrasound (FUS)-mediated therapies. However, conventional beamformers such as Delay-Sum-Integrate (DSI) and Robust Capon Beamforming (RCB) face trade-offs in resolution, artifact suppression, and parameter interpretability. We recently introduced PADAM (Passive Acoustic Dynamic Differentiation and Mapping), a time-domain beamformer adapted from MUSIC that decomposes spatial covariance matrices of time-delayed RF signals. PADAM achieves a 6.1× improvement in lateral resolution, 4.2× reduction in artifact power, and 3× faster computation across parameter sweeps compared to RCB. Its tunable parameter (m) governs the number of retained signal subspace components and enables real-time differentiation between inertial (broadband) and stable (narrowband) cavitation—a physically meaningful capability absent in conventional methods. In vitro phantom studies confirmed inertial cavitation localization at low m and stable cavitation emergence at higher m, supported by spectral analysis. Through-skull imaging further demonstrated PADAM’s robustness to aberration. However, PADAM is not power-based and does not directly reflect energy deposition. To address this, we developed p-PADAM—a hybrid framework combining PADAM’s frequency-sensitive classification with conventional power mapping. Validated across simulations, phantoms, and skull models, p-PADAM enhances spatial specificity and mechanistic insight for image-guided FUS interventions.
Caso et al. (2025) studied this question.