Textile-based shape-conformable and breathable ultrasound imaging probe

The study addresses the need for continuous, noninvasive, daily health monitoring using wearable devices that are both flexible and breathable. Conventional diagnostics often detect disease after symptoms emerge, and low flexibility or breathability in wearables compromises signal quality and skin health. While textiles provide high flexibility and breathability and have been used for wearable ECG and temperature sensing, they have not been suitable substrates for ultrasound (US) imaging due to strong wave reflection/attenuation at air gaps and frictional losses within porous textile structures. US imaging can visualize internal tissue morphology and dynamics for early detection of conditions that may not affect vital signs early (e.g., valvular heart disease). Existing wearable US approaches include fixated conventional probes, thin rigid bioadhesive probes, and flexible elastomer- or film-based arrays, but they can be uncomfortable for long-term wear due to poor breathability. The research question is whether a textile-based, shape-conformable, breathable US probe can be engineered to transmit US effectively through textiles while maintaining comfort and enabling long-term monitoring of internal tissues.

Prior work on wearable health devices leveraged e-textiles for ECG and temperature sensing, but lacked capability for internal tissue imaging. Early wearable US strategies included mechanical fixators for conventional probes, which can induce excessive pressure and discomfort, and thin rigid probes with hydrogel/bio-adhesive surfaces that adhere without pressure. Flexible US arrays on polyimide or elastomer substrates demonstrated B-mode, Doppler, and elastography imaging with improved skin conformity; however, plastic/elastomer films are not breathable and can cause inflammation over long-term use. Two main barriers prevented US imaging through textiles: conversion of acoustic energy to heat from fiber friction and strong reflections at textile–air boundaries due to impedance mismatch (air ~400 Rayls vs polyester ~3.0 MRayls; polyester-to-air transmission factor ~3.0 × 10⁻⁶). Thus, a breathable textile-based probe that preserves US transmission remained unachieved.

Device fabrication: The probe sandwiches 16 ultrasound elements (1–3 PZT/epoxy piezoelectric composites; element dimensions 0.50 × 5.00 × 0.29 mm; ~5 MHz resonance) between two woven polyester e-textiles (0.09 mm thick). Gold sputter coats were applied to both composite faces. Electroless copper plating on the textiles forms signal electrodes/wirings on the top textile and a common ground on the bottom textile. Plating deposits copper within the textile volume, filling fiber air gaps at electrode regions to improve acoustic transmission; non-electrode regions retain porosity for breathability. The acoustic impedance of copper-filled textile (~6.6 MRayls) lies between that of the piezo-composite (~8.5 MRayls) and human tissue (~1.6 MRayls), improving transmission factor from ~0.10 to ~0.34. Elements were bonded to textiles with low-temperature solder (160 °C, 10 s), with solder filling gaps and enhancing US coupling. To recover flexibility lost by the two-layer structure, cut-outs were added between successive top electrodes. An adhesive hydrogel sheet (≈0.7 mm) was attached under the element array for skin coupling, with minimal acoustic mismatch (hydrogel ~1.6 MRayls). Comparison probes using polyimide (0.06 mm) and PDMS (0.10 mm) films were fabricated similarly for benchmarking.

Wearability characterization: Air permeability measured via Frazier method (DAP-360). Flexural rigidity B and flexural hysteresis 2HB were measured using a pure bending tester (KES-FB2-A) with probes centered on the element array; B derived at curvature radius 1 cm, 2HB from loading–unloading difference at 1 cm; three repeats in both directions.

Mechanical stability: Large-deformation stability assessed by wrapping probes around cylinders (diameters 20, 10, 7.5, 5, 3, 2, 1 mm) with ground side inside; electromechanical coupling coefficient (EMCC, k_t) measured post-bending. Repeated-deformation stability assessed by cycling at 5 mm curvature radius to 10,000 bends at 20 rpm with k_t measured after 1, 5, 10, 50, 100, 500, 1,000, and 10,000 cycles.

Imaging performance: Pulse–echo on 10 mm-thick 3.0 wt% agar with single-element sequential transmissions; echoes from back surface recorded by the same element; -6 dB pulse width and bandwidth computed. Resolution measured with a wire phantom (0.1 mm wires at 10 mm spacing; c=1432 m/s; attenuation 0.59 dB cm⁻¹ MHz⁻¹); B-mode reconstructed using RF acquisition and image reconstruction detailed below. Blood vessel phantom (6 mm diameter vessel at 15 mm depth; c=1450 m/s; attenuation 0.5 dB cm⁻¹ MHz⁻¹) imaged in longitudinal and transverse planes.

Human studies: Textile probe attached on the neck to image longitudinal planes of internal jugular vein (IJV) and common carotid artery (CCA) through hydrogel. Acquisitions at 20 fps for 10 s to observe CCA pulsation. IJV diameter assessed at upper-body elevations 0°, 30°, 90° without changing probe location. For feasibility of long-term use, a 24-hour study acquired 10 s at 10 fps every 30 min (probe disconnected between sessions except during sleep; supine posture) and monitored image quality, diameters, and CCA pulse rate.

Signal acquisition and reconstruction (Verasonics V1): Elements excited with two-cycle 5 MHz, 50 V pulses; received RF sampled at 20 MHz, 1536 samples. Synthetic aperture (SA) with 16 transmissions used; to boost SNR, Hadamard-encoded SA excited all elements with ±1 apodizations (H16), followed by decoding. RF band-pass filtered (3.0–7.0 MHz); time-gain compensation 0.5 dB cm⁻¹ MHz⁻¹. Delay-and-sum (DAS) beamforming used assuming linear element geometry (appropriate for flat phantom surfaces and near-linear neck longitudinal axis), then generalized coherence factor (GCF) weighting applied to suppress incoherent noise and grating lobes. Resolution quantified by FWHM of intensity profiles around known targets. Vessel diameters extracted from brightness peaks along lateral 0 mm line after Gaussian smoothing (σ=2.5 mm lateral, 0.5 mm axial). EMCC measured from impedance spectra (Agilent 4294A) using resonant and anti-resonant frequencies. Ethical approval KE21-63 and KE24-4; informed consent obtained.

• The textile-based probe achieved high breathability and flexibility while enabling effective US transmission through textiles by copper-filling electrode regions and solder bonding at interfaces. Average air permeability was 11.7 cm³ cm⁻² s⁻¹ (vs 0.0 for polyimide- and PDMS-based probes).
• Flexural rigidity: textile-based probe B ≈ 0.066 × 10⁴ N·m²·m⁻¹, significantly lower than polyimide-based (2.3 × 10⁴) and comparable to PDMS-based (0.047 × 10⁴). Flexural hysteresis 2HB: polyimide 3.4 × 10² N·mm, PDMS 0.048 × 10² N·mm, textile 0.093 × 10² N·mm.
• Mechanical stability: EMCC (k_t) showed no significant degradation after large bending (down to small diameters) and after 10,000 bending cycles for textile and polyimide probes; PDMS exhibited larger variability and degradation, likely due to adhesion issues and modulus mismatch.
• Pulse–echo performance: Average -6 dB pulse width 0.64 µs (~0.96 mm at 1500 m/s); -6 dB bandwidth 3.8 MHz. All 16 elements transmitted/received; sensitivity variation attributed to contact without pressing.
• Imaging resolution (wire phantom): Axial FWHM ~0.55 mm across depths; lateral FWHM increased with depth from ~0.30 mm at 10 mm to ~0.90 mm at 40 mm. Grating lobe artifacts observed due to 1.0 mm element pitch.
• Vessel phantom: 6 mm vessel at 15 mm depth was clearly visualized in both longitudinal and transverse planes, resolving vessel walls.
• Human neck imaging: CCA pulsation visualized with systolic peak and dicrotic notch; time-series diameter changes were tracked over 10 s at 20 fps. IJV diameter changes with posture detected: 5.9 mm (0°), 3.5 mm (30°), 2.4 mm (90°), consistent with venous volume changes.
• 24-hour feasibility: Repeated acquisitions (10 s every 30 min, 10 fps) consistently visualized IJV and CCA; mean vessel diameters and CCA pulse rates computed per acquisition. Pulse rate tended to be lower during sleep. Some acquisitions missed vessel walls due to position shifts, but overall 24-hour monitoring was successful. No skin abnormalities were observed post-study.

The study demonstrates that textiles, previously unsuitable for US probes due to air-induced impedance mismatch and attenuation, can function effectively by filling electrode regions with copper and solder to eliminate local air gaps while preserving breathability in non-electrode regions. The hydrogel interface maintained acoustic coupling while preventing direct skin contact with metals; nonetheless, long-term biocompatibility and corrosion/leaching under sweat exposure must be assessed and may require protective coatings. Compared to elastomer substrates (e.g., PDMS), e-textiles showed superior mechanical stability of metal interconnects without needing serpentine geometries, enabling higher element density. Although hydrogel may reduce breathability locally, its porous, hydrous nature allows sweat absorption and gas exchange, mitigating skin stress relative to non-breathable film substrates. Image quality lagged commercial probes, primarily due to grating lobes from the 1.0 mm pitch exceeding half-wavelength at 5 MHz; potential remedies include denser woven textiles for finer wiring to permit smaller pitch and/or lower frequency (with resolution trade-offs). Despite the absence of a backing layer to preserve flexibility and breathability, pulse widths remained comparable to commercial probes, likely due to absorption by the top e-textile where small residual air gaps attenuate backward-propagating waves. For robust, long-term monitoring where the target may shift due to posture or attachment errors, a two-dimensional element array with larger 3D FOV is needed. The demonstrated capability to monitor CCA pulsation (arterial stiffness surrogate) and IJV diameter (circulating volume surrogate) suggests potential for early detection of atherosclerosis risk and dehydration, respectively.

A textile-based ultrasound imaging probe was successfully fabricated by embedding 1–3 piezoelectric composite elements between electroless copper-plated woven polyester textiles, achieving both shape conformity and breathability. The device showed high air permeability, low flexural rigidity, and strong stability under large and repeated deformations. Imaging performance enabled clear visualization of phantoms and human neck vessels, including continuous monitoring of CCA pulsation and IJV diameter over 24 hours without skin complications. These results support the feasibility of wearable, long-term US monitoring for early disease detection in daily life. Future work should focus on enhancing image quality by reducing element pitch (e.g., via denser textiles), expanding FOV with 2D arrays for robust positioning, and ensuring long-term biocompatibility and corrosion resistance of embedded metals.

• Image quality was lower than commercial probes, particularly lateral resolution, due to grating lobes from the 1.0 mm element pitch exceeding half-wavelength; this limited imaging of complex organs and transverse planes.
• The probe lacks a backing layer to preserve flexibility/breathability; while pulse width remained acceptable, control over ring-down is reduced and relies on e-textile absorption.
• Long-term biocompatibility has not been fully assessed; copper and solder are shielded by hydrogel but may corrode or leach under prolonged sweat exposure, necessitating protective coatings and testing.
• Dependence on an adhesive hydrogel interface may locally reduce breathability and could affect long-term comfort, positioning, and contact uniformity.
• Monitoring robustness is sensitive to probe–target alignment; during 24-hour tests, posture-induced shifts sometimes moved vessel walls out of the imaging region. A 2D array and larger FOV are needed to mitigate misalignment.
• The stationary acquisition system limited fully continuous data collection during daily activities; real-world ambulatory monitoring will require portable electronics and cabling solutions.