Trabecular bone surface is distant from the ideal non-reflecting, continuous, regular, flat surface suitable for accurate digital image correlation (DIC) measurements. We tested the feasibility of DIC to accurately measure surface displacements on trabecular microstructures, using four-extensometer technique as gold standard measurement.

Thirty cylindrical human trabecular bone specimens were obtained. Ten were used to evaluate if pattern creation, required for DIC, had any effect on the apparent elastic modulus (E_app) of trabecular bone, the remaining twenty were used to assess DIC accuracy in measuring E_app.

All specimens underwent a loading scheme including i) 10 preconditioning cycles, ii) 8 monotonic compressive ramps up to 0.5% nominal deformation, and iii) a second compressive series identical to the first.

Changes in E_app, due to pattern created between the two series, were assessed using the four-extensometer techniques. Pattern quality was also evaluated.

DIC and four-extensometer technique were then used in the remaining twenty specimens to measure local axial displacements and compute global axial deformation during the first and second series, respectively. DIC accuracy was assessed comparing E_app values calculated using axial deformation determined with the two techniques.

Achieved pattern had on average a speckle size of 2.8 pixels, and 42% coverage. Pattern creation did not alter significantly E_app values (median difference=−0.6%;​ Wilcoxon p=0.76).

DIC technique was not applicable on the most porous specimens. DIC-extensometer comparison was not possible in three low-density specimens with E_app < 0.4 GPa because of progressive trabecular damage over test repetitions.

Good agreement in E_app values was found in the remaining sixteen specimens (median difference=−1.5%;​ 10th percentile=−7.5%;​ 90th percentile=6.9%;​ max difference < 10%).

Four-extensometer and DIC are interchangeable techniques:​ the former is the most straightforward to measure the axial deformation of trabecular bone tissue under monotonic compression, the latter is a useful alternative whenever surface displacement maps are needed.