Detection of First Generation Agent Bubbles in the B-Mode Echocardiographic Images using Difference of Gaussians

Szewczuk, Maciej; Sorysz, Danuta; Piórkowski, Adam

Detection of First Generation Agent Bubbles in the B-Mode Echocardiographic Images using Difference of Gaussians

Abstract

Bubble study is a diagnostic procedure that utilizes first-generation contrast microbubbles to assess the severity of Patent Foramen Ovale, primarily through Transesophageal Echocardiography. The key clinical interest for the physician is the quantity of microbubbles passing through the PFO from the right to the left atrium. This work evaluates the feasibility of applying an algorithm based on Difference of Gaussian kernel filtering for the detection and analysis of microbubble counts within the atrial chambers. The proposed multi-stage pipeline achieves a promising detection efficiency with a Dice coefficient up to 0.90. The article describes the algorithm optimization process and its application in simulated clinical conditions to perform an analysis based on a study recording, aimed at supporting the physician and streamlining the diagnostic process.

Keywords

bubble study patent foramen ovale PFO transesophageal echocardiography TEE difference of gaussians DoG microbubble detection

Citation

Maciej Szewczuk, Danuta Sorysz, and Adam Piórkowski, "Detection of First Generation Agent Bubbles in the B-Mode Echocardiographic Images using Difference of Gaussians," Zeszyty Naukowe WWSI, vol. 20, no. 33, pp. 7–28, 2026. [Online]. https://doi.org/10.26348/znwwsi.33.7

URI

https://open.icm.edu.pl/handle/123456789/26884

Belongs to collection

Artykuły / Articles

Full item record