Decoupling Transmission and Transduction for Improved Durability of Highly Stretchable, Soft Strain Sensing: Applications in Human Health Monitoring

Ali Kight, Ileana Pirozzi, Xinyi Liang, Doff B. McElhinney, Amy Kyungwon Han, Seraina A. Dual, Mark R. CutkoskySensors 2023 · DOI

TL;DR

This paper introduces a novel strain sensor design that decouples mechanical stretching from signal transduction, using a soft elastomer to transmit strain to a high-precision MEMS sensor. This approach achieves remarkable durability (over a million cycles with negligible drift) and adaptability for applications like implantable cardiac monitoring, overcoming limitations of traditional stretchable sensors.

What Problem Does This Paper Address?

Developing highly stretchable sensors for applications involving soft, dynamic structures like the human body has been a significant challenge. Conventional strain sensors are often rigid and limited to small deformations. While newer stretchable sensors use novel materials, they frequently suffer from poor durability, baseline drift, and unreliable interconnections, especially under repeated large strains. This limits their use in demanding applications such as long-term implantable medical devices, where sensors must withstand millions of cycles without degradation. The authors specifically target the need for a durable, implantable sensor to monitor cardiac strain, a critical metric for heart health that current methods struggle to measure continuously and reliably.

Methods

The core innovation is the decoupling of transmission and transduction. Instead of relying on a single conductive material to both stretch and generate a signal, this sensor uses two distinct components:

  • Elastomeric Transmission: A non-conductive, soft silicone elastomer (Ecoflex, Shore 00-10) acts as a mechanical intermediary. It is designed to stretch significantly (up to 20%) and transmit this mechanical strain to the transducer. Its softness is chosen to be comparable to native heart tissue.
  • MEMS Transducer: A high-precision, commercially available MEMS barometric pressure sensor (Bosch BMP384) is used. This sensor's capacitive diaphragm deflects under pressure, generating an electrical signal. The elastomer is mechanically coupled to this diaphragm.

This separation allows each component to be optimized independently: the elastomer for stretchability and biocompatibility, and the MEMS sensor for high sensitivity, accuracy, and integrated electronics.

To account for the viscoelastic nature of the silicone, which introduces hysteresis and frequency-dependent behavior, the authors developed a lumped-parameter model. They used a generalized Maxwell model to create transfer functions that map the sensor's pressure output to the actual strain. They compared different model orders (zero to third-order) to find the best fit for dynamic data.

Datasets

The paper does not explicitly mention pre-existing, large-scale datasets for training or evaluation. Instead, they generated their own experimental data through:

  • Quasi-static characterization: Testing sensors under controlled, slow (1 Hz) uniaxial stretching.
  • Cyclic loading tests: Subjecting sensors to millions of cycles of 30% strain at 20 Hz to assess durability.
  • Dynamic characterization: Testing sensors across physiologically relevant frequencies (0.5-2 Hz) to evaluate frequency-dependent behavior.
  • Application-specific testing: Using a realistic cardiac strain-time profile from literature to program a muscle lever for simulating heartbeats at various rates (60, 90, 120 bpm).
  • Curved surface testing: Using a 3D-printed ventricular phantom to simulate implantation on a non-flat surface.

The data generated from these experiments served as both training and validation sets for their modeling efforts.

Key Insights and Ideas

  • Decoupling is Key to Durability: By separating the mechanical stretching (transmission) from the signal generation (transduction), the sensor avoids the common failure modes of conductive materials in stretchable sensors, leading to exceptional durability.
  • MEMS Sensors as Robust Transducers: Leveraging off-the-shelf, high-precision MEMS sensors offers a path to robust, miniaturized, and electronically integrated sensing solutions for soft electronics.
  • Viscoelasticity Requires Advanced Modeling: The inherent viscoelasticity of soft elastomers necessitates more than simple linear calibration. A second-order lumped-parameter model proved crucial for accurately capturing dynamic strain behavior and compensating for hysteresis.
  • Tunable Sensitivity: The sensitivity of the sensor can be easily tuned by adjusting the stiffness (Young's modulus and cross-sectional area) of the elastomeric transmission element, allowing for application-specific optimization.
  • Biocompatibility and Conformability are Achievable: The chosen materials (silicone, MEMS) are biocompatible, and the sensor's design allows it to conform to curved surfaces, making it suitable for in-vivo applications.

Key Results

  • Durability: The sensor demonstrated negligible drift and no apparent physical degradation after one million cycles of 30% strain at 20 Hz.
  • Accuracy: A second-order lumped-parameter model reduced the root-mean-square error (RMSE) in predicted strain by 73% compared to a linear (zero-order) calibration when tested with physiologically relevant cardiac strain profiles.
  • Sensitivity: The sensor achieved a sensitivity of 136 Pa per percent strain with a signal-to-noise ratio (SNR) of over 100, even with a very soft elastomer.
  • Frequency Dependence: The sensor exhibited a frequency-dependent response, with sensitivity increasing by up to 25% at higher frequencies (2 Hz vs. 0.5 Hz), which was effectively modeled.
  • Low Cost and Fabrication: The sensor is fabricated using readily available materials and a facile process, costing less than fifty US dollars.

Limitations and Future Work

  • Friction on Curved Surfaces: While the sensor performed well on a curved phantom, friction between the sensor and the surface can induce shear, affecting the pressure-strain coupling. A hydrogel was used to mitigate this in benchtop tests, but its long-term in-vivo stability and biocompatibility need further investigation.
  • In-Vivo Mechanical Environment: The benchtop hydrogel solution may not perfectly replicate the in-vivo environment. Future work should explore durable, stretchable hydrogels and consider how the pericardium might facilitate mechanical transparency.
  • Long-Term Biocompatibility: While materials are biocompatible, extensive in-vivo studies are needed to evaluate long-term tissue responses, such as fibrosis and inflammation.
  • Wireless Integration: For practical implantable applications, integrating wireless communication protocols is a crucial next step.
  • Model Complexity: While the second-order model performed best, the authors acknowledge the potential for overfitting with higher-order models, suggesting careful model selection is important.

Summary updated

Abstract

This work presents a modular approach to the development of strain sensors for large deformations. The proposed method separates the extension and signal transduction mechanisms using a soft, elastomeric transmission and a high-sensitivity microelectromechanical system (MEMS) transducer. By separating the transmission and transduction, they can be optimized independently for application-specific mechanical and electrical performance. This work investigates the potential of this approach for human health monitoring as an implantable cardiac strain sensor for measuring global longitudinal strain (GLS). The durability of the sensor was evaluated by conducting cyclic loading tests over one million cycles, and the results showed negligible drift. To account for hysteresis and frequency-dependent effects, a lumped-parameter model was developed to represent the viscoelastic behavior of the sensor. Multiple model orders were considered and compared using validation and test data sets that mimic physiologically relevant dynamics. Results support the choice of a second-order model, which reduces error by 73% compared to a linear calibration. In addition, we evaluated the suitability of this sensor for the proposed application by demonstrating its ability to operate on compliant, curved surfaces. The effects of friction and boundary conditions are also empirically assessed and discussed.
BibTeX
@inproceedings{kight2023decoupling,
  author = {Ali Kight and Ileana Pirozzi and Xinyi Liang and Doff B. McElhinney and Amy Kyungwon Han and Seraina A. Dual and Mark R. Cutkosky},
  title = {{Decoupling Transmission and Transduction for Improved Durability of Highly Stretchable, Soft Strain Sensing: Applications in Human Health Monitoring}},
  booktitle = {Sensors},
  year = {2023},
  doi = {10.3390/s23041955},
}
Figure