Adaptive Probe Geometry Estimation for High Frame Rate Imaging Using Flexible Ultrasound Arrays

This dataset was developed and utilized for the research paper titled "Adaptive Probe Geometry Estimation for High Frame Rate Imaging Using Flexible Ultrasound Arrays," which was accepted and presented at the IEEE International Symposium on Biomedical Imaging (ISBI) 2026. This dataset comprises simulated HDF5 (.h5) files developed for flexible array ultrasound research. The core collection comprises 1,000 single-target files, evenly distributed across 500 cyst simulations and 500 vessel simulations. The flexible probe geometry is modeled as a 3rd-order polynomial. The probe can take various shapes — not only symmetric concave, but also left-dome concave and right-dome concave configurations with various peak amplitudes. Dataset Generation Pipeline The simulation and data generation process follows a structured computational pipeline: Phantom Design: A tissue phantom is designed with specific echogenicity properties, accommodating anechoic, hypoechoic, or hyperechoic cysts and vessels. Scatter Generation (genscat): The defined phantom is passed through the genscat function, which generates a scatterer map representing the acoustic properties of the tissue medium. RF Data Simulation (simus): Transmission delays are computed using the designed flexible probe geometry. The scatterer map, alongside these delays, is processed by the simus function to simulate the raw Radio Frequency (RF) channel data. Beamforming (das): Delay-and-Sum (DAS) beamforming is performed using the defined true probe geometry and the simulated RF data to reconstruct the final B-mode image. Evaluation Subset Additionally, the dataset includes a specialized evaluation subset designed to rigorously test the robustness of developed models and beamforming algorithms. This evaluation data features complex multi-target scenarios, specifically containing: Double cyst cases Double vessel cases Mixed target cases (one cyst and one vessel) Data Structure Each .h5 file is structured to provide comprehensive acoustic and geometric data, specifically including: Raw Data: Pre-beamformed Radio Frequency (RF) data and associated transmit delays. Imaging: Reconstructed B-mode images alongside ideal phantom image matrices. Probe Configuration: True array geometry details and the polynomial coefficients defining the flexible probe shape. Simulation Parameters: System sampling frequency and center frequency. Target Characteristics: Region of Interest (ROI) metadata detailing the target's center coordinates, radius, type, and echogenicity. Acknowledgments This dataset was generated using the MATLAB UltraSound Toolbox (MUST). If you utilize this dataset, please consider the following references regarding the underlying simulation framework: Garcia D. "SIMUS: an open-source simulator for ultrasound imaging. Part I: theory & examples," Computer Methods and Programs in Biomedicine, 2022;218:106726. Garcia D. "Make the most of MUST, an open-source MATLAB UltraSound Toolbox," IEEE International Ultrasonics Symposium, 2021. This research is supported by the Ministry of Education, Singapore, under the Academic Research Fund Tier 1 (GMS 1052). Cite us If you use this dataset in your research or find it helpful, please cite our paper : Sainatha, R., & Panicker, M. R. (2026). Adaptive Probe Geometry Estimation for High Frame Rate Imaging Using Flexible Ultrasound Arrays. In IEEE International Symposium on Biomedical Imaging (ISBI). London, UK.