ElastoSpec: Flexible Viscoelastic Measurement

• ElastoSpec is a flexible rheometry technique that uses magnetic-indentation and interferometry to measure the complex shear modulus of soft solids.
• It operates in both quasi-static and inertial regimes by converting force–displacement data through continuum mechanics models.
• ElastoSpec principles extend to imaging-based elastography and spectroscopic wave analyses, offering precise stiffness characterization.

ElastoSpec denotes, in its most specific usage, the flexible rheometer described in “A flexible rheometer design to measure the visco-elastic response of soft solids over a wide range of frequency” (Rolley et al., 2019). In that instrument, the visco-elastic response of a soft solid is inferred from the mechanical response of a magnet deposited at the surface of a slab of material and excited electromagnetically, while an interferometric measurement of the magnet displacement provides the readout. A broader usage is also suggested by later literature, which describes several optical, wave-based, and inverse-problem frameworks as “ElastoSpec-like” or places them explicitly “in the context of ElastoSpec”; in that broader sense, the term refers to methods that convert deformation, elastic-wave propagation, speckle evolution, reflectance anisotropy, or spectral shifts into quantitative elastic or viscoelastic parameters (Sherina et al., 17 Jun 2026, Sendra et al., 2023, Feng et al., 2022, Bayrak et al., 27 Dec 2025).

1. Original instrument and measurement principle

The original ElastoSpec is a flexible rheometer for measuring the visco-elastic response of soft solids over a wide frequency range by combining three elements: magnetic actuation of a small surface indenter, interferometric measurement of its displacement, and a continuum-mechanics model that converts the measured force–displacement response into the material’s complex shear modulus $\mu(\omega)=G'(\omega)+iG''(\omega)$ (Rolley et al., 2019). The sample is a soft solid slab. A small permanent magnet disk of radius $R$ is placed on its surface and acts as the indenter. A coil above the magnet applies a vertical oscillatory magnetic force, and the magnet’s vertical displacement $Z$ is measured optically using a laser-interferometric vibrometer. There is no mechanical contact between the sample and the rest of the apparatus except for the magnet itself.

The reported design uses a Nd magnet with radius $R \approx 1$–$5$ mm and thickness $d \approx 0.5$ mm. The coil has inner diameter $\sim 10$ mm, outer diameter $\sim 25$ mm, height $\sim 15$ mm, about 500 windings, and force constant $F/I \sim 0.01$ N/A. The vibrometer nominal displacement resolution is about $R$0 pm. Force calibration is performed using a precision scale and DC current. The method is intended for soft solids with moduli below about $R$1 MPa, and naturally in the kPa range for gels (Rolley et al., 2019).

The primary measured quantity is the effective modulus

$R$2

where $R$3 is the force acting on the sample. A digital lock-in amplifier measures the in-phase and out-of-phase components, yielding $R$4 and $R$5. This quantity is “effective” because it depends on the material rheology, sample thickness, inertia at high frequency, and indentation geometry, rather than on intrinsic rheology alone (Rolley et al., 2019).

2. Quasi-static and inertial regimes

ElastoSpec operates in two mechanically distinct regimes. In the overdamped or quasi-static regime, inertia of both the magnet and the soft solid is negligible, and the response is governed by static linear elasticity with the simple replacement $R$6 (Rolley et al., 2019). For a thick sample, $R$7, the paper gives

$R$8

so that for an incompressible material, $R$9, one obtains $Z$0. For a thin layer, $Z$1, the incompressible lubrication approximation yields

$Z$2

For arbitrary $Z$3, the relation is written as $Z$4, where $Z$5 is computed numerically in the overdamped limit (Rolley et al., 2019).

At higher frequency, the response is no longer quasi-static. The magnet and a portion of the sample are set into oscillatory motion, the solid supports damped elastic waves, and the effective response includes inertial contributions. The paper reports a resonance near $Z$6 kHz, a sign change in $Z$7, and high-frequency behavior dominated by inertia, with asymptotic scaling

$Z$8

The resonance is estimated by

$Z$9

Because the force on the sample is not simply the total force on the magnet, magnet inertia must be corrected through

$R \approx 1$0

with $R \approx 1$1 (Rolley et al., 2019).

The high-frequency inversion uses an incompressible viscoelastic layer model governed by

$R \approx 1$2

Incompressibility is enforced through an axisymmetric stream function $R \approx 1$3,

$R \approx 1$4

and the modal structure introduces

$R \approx 1$5

For infinite thickness, the stress–displacement relation in Fourier–Bessel space is

$R \approx 1$6

The mixed disk-indenter condition is then solved iteratively using a discrete Hankel transform (Rolley et al., 2019).

3. Validation, operating range, and performance

Validation in the original study was performed on a PDMS gel, Dow Corning CY52-276, mixed 1:1 and cured 24 h at room temperature. The material is nearly incompressible, with $R \approx 1$7, and has static modulus $R \approx 1$8 kPa (Rolley et al., 2019). In the low-frequency regime, data for different thicknesses $R \approx 1$9 collapse onto a single master curve once rescaled by $5$0, and the extracted rheology matches a conventional Anton-Paar rheometer up to about $5$1 Hz. The viscoelastic response follows

$5$2

with fitted values approximately $5$3 kPa, $5$4 s, and $5$5 (Rolley et al., 2019).

The same viscoelastic law predicts the observed resonance, the inertial $5$6 growth in $5$7, and the measured $5$8 at high frequency. The paper states that the theoretical model reproduces the data without adjustable parameters once the low-frequency rheology is fixed. The reported frequency range spans about $5$9 decades, from $d \approx 0.5$0 to $d \approx 0.5$1, and the setup extends to approximately $d \approx 0.5$2 kHz (Rolley et al., 2019).

The main experimental advantage over conventional plate–plate rheometers is that ElastoSpec does not require perfect sample matching to plate geometry, avoids the need for perfect adhesion to both plates, works with small samples, can access higher frequencies, and is well suited to soft biological tissues. The main limitation is that high-frequency inversion is model-dependent. The paper also notes that between $d \approx 0.5$3 and $d \approx 0.5$4 rad/s, a $d \approx 0.5$5 inaccuracy in $d \approx 0.5$6 can produce a $d \approx 0.5$7 error in $d \approx 0.5$8, so precise knowledge of $d \approx 0.5$9, magnet mass, sample density, and force calibration is required for percent-level rheology (Rolley et al., 2019).

4. Imaging-based ElastoSpec: displacement estimation and inverse elastography

A broader ElastoSpec usage is suggested by OCT and photoacoustic elastography papers that explicitly frame their methods as ElastoSpec-style or as operating “in the context of ElastoSpec.” In “Displacement field estimation from OCT images utilizing speckle information with applications in quantitative elastography” (Sherina et al., 2020), ElastoSpec is a two-step quantitative OCT elastography pipeline. The first step estimates the full internal displacement field from pre- and post-compression OCT images, and the second step reconstructs elastic material parameters. The distinctive feature is the use of tracked bright speckle formations, or “bubbles,” as sparse displacement constraints in a variational optical-flow model. The full objective is

$\sim 10$0

and the second-stage parameter inversion uses iterative regularization of the elasticity operator. The paper reports that from a single displacement field one can reliably reconstruct $\sim 10$1 and the Young’s modulus

$\sim 10$2

whereas $\sim 10$3 itself is poorly identifiable from one displacement measurement (Sherina et al., 2020).

“Simultaneous reconstruction and displacement estimation for spectral-domain optical coherence elastography” (Mason et al., 2021) makes the coupling between image formation and motion estimation explicit. It uses standard reconstruction with ISAM resampling implemented through the non-uniform fast Fourier transform, estimates relative displacement with multi-resolution cross-correlation block matching from PIVlab including sub-pixel estimation via 2D Gaussian regression, warps neighboring frames with a linear operator $\sim 10$4, denoises the motion-compensated data with BM4D, applies the adjoint warping $\sim 10$5, and then recalculates displacement on the denoised frames. Validation on commercial spectral-domain OCT data from a Wasatch Photonics 800 nm system showed that BM4D applied after motion compensation reduces image RMSE by $\sim 10$6 relative to the original reconstruction, while displacement accuracy improved by $\sim 10$7 in the lateral direction and $\sim 10$8 in the axial direction (Mason et al., 2021).

In photoacoustic elastography, “Texture Generation for Photoacoustic Elastography” (Glatz et al., 2014) addresses the fact that photoacoustic images are often considered speckle-poor. The paper shows that artificial speckle patterns can be introduced by reconstructing from only a band-limited part of the measurement data. For a hard band-pass

$\sim 10$9

the reconstructed image becomes $\sim 25$0, where the oscillatory Bessel-based point-spread function generates texture that improves optical-flow-based displacement estimation. The reported experiments show that Gaussian additive texture does not help much, whereas band-limited texture can substantially improve motion estimation and gives up to about a $\sim 25$1 decrease in angular error in some non-rigid cases (Glatz et al., 2014).

The multi-modal extension is “Quantitative Multi-Modal Optical Coherence Photoacoustic Elastography” (Sherina et al., 17 Jun 2026), which defines a quantitative OCPE framework combining OCT and PAT for quasi-static elastography. Motion recovery uses a data-fusing elastographic optical flow method, DEOFM, and stiffness reconstruction is formulated through DSI, NLI, and IIM. In a silicone elastomer phantom with ground-truth Young’s modulus values of $\sim 25$2 kPa for the background and $\sim 25$3 kPa for both inclusions, the paper reports that the combined OCT-PAT approach outperforms single-modality OCT elastography and PAT elastography, yielding higher strain signal-to-noise ratio and improved stiffness estimates. The brightness weight $\sim 25$4 gave the best trade-off in merged-image tests (Sherina et al., 17 Jun 2026).

5. Spectroscopic and wave-based extensions

Another ElastoSpec-like branch replaces quasi-static indentation by wave propagation. “Broadband-excitation-based mechanical spectroscopy of highly viscous tissue-mimicking phantoms” (Urbańska et al., 2021) launches broadband Rayleigh surface waves with a piezoelectric transducer and measures displacement using spectral-domain OCT. The 2D Fourier-transform analysis extracts phase velocity from the ridge of the dispersion graph through

$\sim 25$5

and attenuation from the imaginary part of the 2D spectrum through

$\sim 25$6

After Rayleigh-to-shear conversion, the method yields $\sim 25$7, $\sim 25$8, and the phase angle $\sim 25$9. In the reported PDMS results, the broadband analysis covers roughly $\sim 15$0–$\sim 15$1 kHz, already well above standard rheometer limits, and the OCT-derived high-frequency data connect smoothly to controlled-stress oscillatory rheometer measurements over $\sim 15$2–$\sim 15$3 Hz (Urbańska et al., 2021).

“Ultra-wideband optical coherence elastography from acoustic to ultrasonic frequencies” (Feng et al., 2022) extends this logic by using noise reduction, anti-aliasing demodulation, and advanced wave analysis so that wave motion can be measured even when the excitation frequency is well above the A-line rate of the OCT system. The paper reports a frequency range from $\sim 15$4 Hz to $\sim 15$5 MHz, stiffness measurement of hard materials including bones with mm-scale resolution, and depth-dependent shear modulus from $\sim 15$6 kPa to $\sim 15$7 MPa in cartilages ex vivo and human skin in vivo. Reported examples include $\sim 15$8 GPa for acrylate plastic, $\sim 15$9 GPa for polystyrene, $F/I \sim 0.01$0 GPa for borosilicate glass, and $F/I \sim 0.01$1 GPa for copper (Feng et al., 2022).

Guided-wave acoustoelasticity forms a related branch. “Guided elastic waves in a highly-stretched soft plate” (Delory et al., 2022) studies the $F/I \sim 0.01$2 and $F/I \sim 0.01$3 modes in a $F/I \sim 0.01$4 mm thick Ecoflex 00-30 plate stretched nearly uniaxially up to $F/I \sim 0.01$5. The difference between squared transverse wave speeds provides a model-independent stress estimate,

$F/I \sim 0.01$6

The paper also shows the limits of classical acoustoelastic theory and introduces a fractional Kelvin-Voigt modulus

$F/I \sim 0.01$7

which improves phase-velocity prediction up to about $F/I \sim 0.01$8 elongation (Delory et al., 2022).

A different spectroscopic transduction is used in “Scanning Reflectance Anisotropy Microscopy for Multi-Material Strain Mapping” (Sendra et al., 2023). SRAM measures the near-normal-incidence ellipsometric response

$F/I \sim 0.01$9

and connects it to strain through the elasto-optic effect. The reported microscope is broadband, hyperspectral, phase-sensitive, and diffraction-limited at about the $R$00–$R$01 nm level, with phase sensitivity better than $R$02. It is demonstrated on metasurfaces, semiconductors, and metals, including single nanoantennas and suspended germanium microbridges (Sendra et al., 2023).

At the microscale, “Elastomer-based whispering gallery mode microlasers with low Young’s modulus for biosensing applications” (Bayrak et al., 27 Dec 2025) uses mechanically compliant WGM microlasers as optical force sensors. The elastomer beads have Young’s modulus $R$03–$R$04 kPa, with an average of about $R$05 kPa. In AFM push–release experiments, the mode linewidth broadens linearly with force, with

$R$06

and the paper proposes the FWHM of the lasing modes as a direct force readout after calibration. Based on measured mode spacing of about $R$07 nm, the platform could measure forces up to about $R$08 nN (Bayrak et al., 27 Dec 2025). This suggests a spectroscopic ElastoSpec variant in which mechanical information is carried by cavity spectra rather than by displacement fields.

6. Related methodologies, interpretation, and limits

Several adjacent methods clarify the conceptual boundary of ElastoSpec. “RheoSpeckle: a new tool to investigate local flow and microscopic dynamics of soft matter under shear” (Ali et al., 2016) couples a stress-controlled MCR301 rheometer equipped with a transparent Couette plexiglass cell to an imaging light-scattering setup. The drift of the speckle pattern gives the mesoscopic displacement field, while drift-corrected intensity correlations measure microscopic dynamics. The method recovers rigid displacements to better than $R$09 micron, measures the velocity profile over a $R$10 mm gap with temporal and spatial resolution of $R$11 s and $R$12 microns, respectively, and shows that the microscopic decay under shear is controlled by the local affine deformation (Ali et al., 2016). “Space-resolved diffusing wave spectroscopy measurements of the macroscopic deformation and the microscopic dynamics in tensile strain tests” (Nagazi et al., 2016) similarly combines a universal testing machine with space-resolved DWS and demonstrates nanometer-scale sensitivity to microscopic dynamics, early detection of plasticity, and excellent agreement between DWS-derived strain maps and stereo-DIC (Nagazi et al., 2016).

Other neighboring approaches do not rely on wave or speckle spectra, but they address the same inverse-mechanical objective. “Elastometry of Deflated Capsules: Elastic Moduli from Shape and Wrinkle Analysis” (Knoche et al., 2013) extracts $R$13, $R$14, and $R$15 from optical images of a deflated capsule by fitting nonlinear membrane-shell shape equations and wrinkle wavelengths. The paper explicitly argues that fitting an elastic capsule with the Laplace–Young equation alone and then computing a Gibbs modulus is systematically wrong, because the method ignores elastic membrane stresses and the coupling between shape and stress (Knoche et al., 2013). At the computational end, “MyElas: An automatized tool-kit for high-throughput calculation, post-processing and visualization of elasticity and related properties of solids” (Wang et al., 2022) automates structural preprocessing, SOEC and TOEC extraction, post-processing, and anisotropy visualization through an energy-strain workflow. Although not an elastography system, it serves a comparable elasticity-analysis role for first-principles materials screening (Wang et al., 2022).

A recurrent misconception in this field is that speckle or texture is necessarily a nuisance. OCT-speckle “bubbles” are used as physically meaningful motion samples in quantitative elastography, and band-limited photoacoustic reconstruction deliberately introduces artificial speckle to make displacement estimation more reliable (Sherina et al., 2020, Glatz et al., 2014). A second misconception is that simple constitutive or geometric models are always adequate. The Ecoflex guided-wave study shows that purely hyperelastic acoustoelastic theory fails unless frequency-dependent rheology is added, and capsule elastometry shows that Laplace–Young-based Gibbs-modulus extraction can be systematically wrong (Delory et al., 2022, Knoche et al., 2013). A third limitation is identifiability: from a single OCT displacement field, $R$16 and Young’s modulus can be reconstructed reliably, but $R$17 itself is poorly identifiable (Sherina et al., 2020). These points indicate that ElastoSpec, whether understood narrowly as the magnetic-indentation rheometer or broadly as an elasto-spectroscopic measurement family, is fundamentally an inverse-problem framework whose performance depends on transduction physics, model adequacy, and parameter identifiability.