Characterization of Macrolesions Induced by Myocardial 

Intermittent high intensity ultrasound pulses with circulating contrast agent microbubbles can induce scattered cavitation caused myocardial microlesions of potential value for tissue reduction therapy. Here, computer-aided histological evaluation of the effective treated volume was implemented to optimize ultrasound pulse parameters, exposure duration, and contrast agent dose. Rats were treated with 1.5 MHz focused ultrasound bursts and Evans blue staining indicates lethal cardiomyocytic injury. Each heart was sectioned to provide samples covering the entire exposed myocardial volume. Both brightfield and fluorescence images were taken for up to 40 tissue sections. Tissue identification and microlesion detection were first done based on 2-Dimages to form microlesion masks containing the outline of the heart and the stained cell regions.

Image registration was then performed on the microlesion masks to reconstruct a volume-based model according to the morphology of the heart. The therapeutic beam path was estimated from the 3-D stacked microlesions, and finally the total microlesion volume, here termed macrolesion, was characterized along the therapeutic beam axis. Radially symmetric fractional macrolesions were characterized via stepping disks of variable radius determined by the local distribution of microlesions. Treated groups showed significant macrolesions of a median volume of 87.3 μL, 2.7 mm radius, 4.8 mm length, and 14.0% lesion density compared to zero radius, length, and lesion density for sham. The proposed radially symmetric lesion model is a robust evaluation for myocardial cavitation-enabled therapy. Future work will include validating the proposed method with varying acoustic exposures and optimizing involved parameters to provide macrolesion characterization.