Data from: Monitoring microvascular changes over time with a repositionable 3D ultrasonic capacitive micromachined row-column sensor
Abstract
eHealth devices, including smartwatches and smart scales, have the potential to transform healthcare by enabling continuous, real-time monitoring of vital signs over extended periods. Existing technologies, however, lack comprehensive monitoring of the microvascular network, which is linked to conditions such as diabetes, hypertension, and small vessel diseases. This study introduces a novel ultrasound approach using a Capacitive Micromachined Ultrasound Transducers (CMUT) Row-Column Array (RCA) for continuous, ultrasensitive 3D Doppler imaging of microvascular changes such as hemodynamic variations or vascular remodeling. In vitro tests and in vivo studies with healthy volunteers demonstrated the sensor's ability to image the 3D microvascular network at high resolution over different time scales with automatic registration, and to detect microvascular changes with high sensitivity. Integrating this technology into wearable devices could one day enhance understanding, monitoring, and possibly early detection of microvascular-related health conditions.
https://doi.org/10.5061/dryad.47d7wm3qg
Description of the data and file structure
These data have been collected to compare different ultrasound imaging sequences used on a capacitive (CMUT) Row-Column Array (RCA).
Files and variables
File: data.zip
Description: Beamformed volume of the acquisitions in a matlab file
SA64
- Data_SA64.mat
- SA64 : Synthetic Aperture (64 emissions on 1 array, 400 frames)
- Variable: Data_SA64, dimensions [axis Z, axis X, axis Y, Frames, Emitting Array]: [37, 64, 64, 400, 1]
- Voxel size : 0.15 mm, Frame Rate : 46.875 Hz
- SA64 : Synthetic Aperture (64 emissions on 1 array, 400 frames)
HSA64
- Data_HSA64.mat
- HSA64 : Hadamard Synthetic Aperture imaging (64 emissions on 1 array, 400 frames)
- Variable: Data_HSA64, dimensions [axis Z, axis X, axis Y, Frames, Emitting Array]: [37, 64, 64, 400, 1]
- Voxel size : 0.15 mm, Frame Rate : 46.875 Hz
- HSA64 : Hadamard Synthetic Aperture imaging (64 emissions on 1 array, 400 frames)
PW1616
- Data_PW1616.mat
- PW: Plane Waves imaging (16 emissions on both arrays, 800 frames)
- Variable: Data_PW1616, dimensions [axis Z, axis X, axis Y, Frames, Emitting Array]: [37, 64, 64, 800, 2]
- Voxel size : 0.15 mm, Frame Rate : 93.75 Hz
- PW: Plane Waves imaging (16 emissions on both arrays, 800 frames)
PW3232
- Data_PW3232.mat
- PW: Plane Waves imaging (32 emissions on both arrays, 400 frames)
- Variable: Data_PW3232, dimensions [axis Z, axis X, axis Y, Frames, Emitting Array]: [37, 64, 64, 400, 2]
- Voxel size : 0.15 mm, Frame Rate : 46.875 Hz
- PW: Plane Waves imaging (32 emissions on both arrays, 400 frames)
Code/software
main_supplementary_figure_2.m
Works on : MATLAB Version: 9.10.0.2198249 (R2021a) Update 8
The program reproduce the figure S2 of the article.
It just needs to be run and is divided in 3 parts :
- loading of the data files
- processing of the data to get power Doppler of the different imaging sequences
- display of the power Doppler for the different imaging sequences.
This dataset is provided to reproduce the supplementary figure S2 of the article "Monitoring microvascular changes over time with a repositionable 3D ultrasonic capacitive micromachined row-column sensor".
This figure shows power Doppler of the fingertip of a healthy volunteer acquired with the CMUT RCA sensor with different imaging sequences: Synthetic Aperture (SA), Hadamard Synthetic Aperture (HSA) and Plane Waves Imaging (PW) described in the methods of the article. 400 or 800 frames were acquired to keep the acquisition duration constant. Here, the Pulse Repetition Rate is set to 3 kHz for long acquisitions (8.5 s).
This dataset consists of data files, named Data_XXXX that are beamformed volumes of the different sequences. The fingertip surface has been removed to erase fingerprints.
The Matlab program main_supplementary_figure_2.m applies an SVD clutter filter on these volumes to compute power Doppler volumes that are then projected to make the figure. Two different processing are applied to the Plane Waves acquisitions (X-Doppler and coherent compounding also described in the article) to compare those methods for our RCA acquisitions.