In silico assessment of histotripsy-induced changes in catheter-directed thrombolytic delivery

Abstract

Introduction: For venous thrombosis patients, catheter-directed thrombolytic therapy is the standard-of-care to recanalize the occluded vessel. Limitations with thrombolytic drugs make the development of adjuvant treatments an active area of research. One potential adjuvant is histotripsy, a focused ultrasound therapy that lyses red blood cells within thrombus via the spontaneous generation of bubbles. Histotripsy has also been shown to improve the efficacy of thrombolytic drugs, though the precise mechanism of enhancement has not been elucidated. In this study, in silico calculations were performed to determine the contribution of histotripsy-induced changes in thrombus diffusivity to alter catheter-directed therapy.Methods: An established and validated Monte Carlo calculation was used to predict the extent of histotripsy bubble activity. The distribution of thrombolytic drug was computed with a finite-difference time domain (FDTD) solution of the perfusion-diffusion equation. The FDTD calculation included changes in thrombus diffusivity based on outcomes of the Monte Carlo calculation. Fibrin degradation was determined using the known reaction rate of thrombolytic drug.Results: In the absence of histotripsy, thrombolytic delivery was restricted in close proximity to the catheter. Thrombolytic perfused throughout the focal region for calculations that included the effects of histotripsy, resulting in an increased degree of fibrinolysis.Discussion: These results were consistent with the outcomes of in vitro studies, suggesting histotripsy-induced changes in the thrombus diffusivity are a primary mechanism for enhancement of thrombolytic drugs.

Development of Convolutional Neural Network to Segment Ultrasound Images of Histotripsy Ablation

“…In practice, imaging probes used for treatment guidance are positioned in a coaxial opening within the therapy source [14], and visualize the target in a different plane. The ablation zone dimensions are nominally symmetric along the lateral and elevational dimensions of the focused transducer [49], and the network could be retrained by adjusting the orientation of the images. A transfer learning approach was used with a ResNet-18 architecture [36].…”