Two Mover Arm Validation 

After determining a two loader model was the optimal alternate design solution for our device, our group needed proof that this greatly simplified model parallels the current planar biaxial testing protocol. The main concern from this model was its capabilities in  achieving similar stretches to the tissues with this model. Thus, in order to validate this new protocol involving only two loaders, we performed planar biaxial tests on a given decellularized RV tissue using the Bose ElectroForce planar biaxial device available in the lab. First, the preconditioning blocks were run to standardize trials (create reproducible loops). Then, we ran three trials under the “standard” protocol where all four loaders were set to stretch the tissue to 10% of its original measured length along each axis. These trials were referenced in the proceeding data analysis (see below) in order to compare the validity/similarity between the future trials run with only two active loading arms. After conducting our reference trials, the Bose ElectroForce loader software was edited to inactivate two of the loaders, leaving two loaders acting orthogonally to one another. These included trials where both loaders stretched the tissue by 10%, 15%, and 20%. Note that before all trials, the tissue was unloaded on each side, allowing for the load cell to be recalibrated and the tissue to begin all trials preloaded to 0.8 g. 

All trials describe above generated files containing information such as time, temperature, device displacement, device strain, load cell measurements, axial commands for loads, marker position, etc. The MATLAB script looks for the files correlating to load data of a particular trial and extracts the load cell measures as well as the time elapsed. From this information, we were able to observe the equibiaxial maximum loads between all trials. Additionally, the script locates and opens the camera basefile, which provides  information on the pixel positions on the x-y plane (2 columns for all five markers) as the test progresses. Besides plotting the marker positions with respect to time, the Matlab script reads both basefiles and locates the index correlating to maximum stretch (while ignoring the effect of creep which we are not studying). Using this index, the generated figures provide a general idea of what the shape of the tissue (area) looks like before the tissue is stretched and after the tissue is fully stretched in all four cases. This is done by plotting all the marker positions at the start of trial (first index) and at the determined equibiaxial maximum index. These results may be seen in the figure below. Note that the blue area is representative of the tissue area encapsulated by the markers before any applied stretch and the magenta represents the same at the equibiaxial maximum.

The results from these trials and the subsequent MATLAB analysis suggested that unlike the “standard” trials, the tissue experiences an overall displacement. Regardless, the aim of the test was to visualize marker displacement and determine how the tissue stretched overall (and whether it does so comparably to the standard protocol). Thus, it was determined a more reliable method of doing so would be looking at overall shape as well as the changes in area from the undeformed to the deformed configurations. Thus, in order to calculate the area bounded by the magenta and blue curves for each trial, an alternate form of Heron’s formula was applied to the bounding marker coordinates. This allowed for the changes in area to be calculated for each trial. From these calculations and plots, it appeared that the two loader model performed at 20% stretch sufficiently mimicked the standard four loader model at 10% stretch in terms of retaining tissue shape (although there was slightly more shear present in these tests). Additionally, the change in area from the two loader model with a 20% stretch appeared to adequately replicate that observed between the standard model at 10% stretch. Furthermore, the MATLAB analysis of the loading files showed that the loads applied to the tissue were most comparable between the standard trials and the two loader (alternate) trial at 20% stretch. Note that all three standard trials were averaged for comparison towards the trials testing the  alternate models.

Compatibility Testing with Multiphoton Microscope at UCSD Microscopy Core

The UCSD Microscopy Core had access to a multiphoton microscope capable of second harmonic generation imaging in late March. Our mentor, Becky Hardie, tested whether our current prototype is compatible with the physical constraints of the microscope and also capable of being utilized for the intended purpose of imaging the decellularized tissue collagen fibers. While imaging, we found that the device does fit within the physical constraints of the microscope but could be improved by including screw holes in the baseplate to minimize the risk of the device falling off the viewing platform. The center of mass of the device is not at the physical center as the rack and pinion system and loading arms are not balanced. The quality of the SHG images obtained from this test helped confirm that the tissue remains stationary enough to capture similar quality images as are obtained when the tissue is affixed to a glass slide. Preliminary results show that the collagen fiber orientation, measured by the fiber angle, decreases as the tissue is stretched.