Weekly Update in lieu of Recurring Meeting of Cleveland Clinic Core Team
Date: June 6, 2017
Provided by: Snehal Chokhandre
Ongoing Action Items:
- Review drafts of data manuscripts for submission.
- Complete studies for tissue testing reproducibility.
- Continue tissue testing of Open Knee(s) specimens.
- Continue segmentation of Open Knee(s) specimens.
- Tissue testing.
- In the last repeatability test set for cartilage confined compression (48 yr old specimen), the last two tests were significantly different than the first three and not in the same direction (indicating tissue state is probably not responsible for the behavior). To assess whether the error was due to the filter tipping or catching along the walls of the confined compression chamber, Snehal has been conducting a series of tests with no filter (essentially an indentation test using confined compression chamber set up with 5 mm wide sample and 3 mm wide indenter).
- With two tests so far it appears that the error is about 4% and the preconditioning data has no indication of buckling, which was present in the previous confined compression test set. This may mean that the 1 mm thick filter may be responsible for the errors and bucking in confined compression. Check Specifications/ExperimentationTissueMechanics for results.
- Snehal has ordered a thicker stainless steel filter and will resume confined compression tests with the new filter when it arrives.
- To counter the misalignment of the sample, filter and the test set up in confined compression tests, Martin (from Biomomentum Inc.) suggested using a confined compression set up with magnet and piston that Biomomentum Inc. developed. He will send one set free of charge once it is manufactured.
- Martin installed another Mach-1 testing system for a sister project in Erdemir Laboratory and Snehal may be able to use both the testing systems to expedite both the protocol / repeatability evaluation and the actual tests, if needed.
- Martin has also updated the load cell amplification boards for all load cells. The boards have a build-in 1kHz analog filter, which may not still be useful to reduce the noise as the noise mostly occurs at 700-800hz. It is not a big concern as the data can be filtered, once collected appropriately at a high sampling frequency.
- Snehal made sure that all the load cells were working properly after the update and has not seen any issues so far.
- Martin has been working on using an xml format for Mach-1 testing sequences to replace the binary file that holds all the sequences. This will be useful to automate updating multiple sequences for each test with sample dimension dependent parameters using Python. This will reduce any user error due to manually updating numbers in the testing sequences.
Martin has sent a test xml file for assessment. Snehal is not sure whether each sequence will have a separate xml file and will confirm with Martin. Ahmet also wanted to know whether the xml file was generated from a LabView dump as the updates in the software may create compatibility problems with the existing xml file(s). Snehal will confirm.
- Data manuscripts.
- No progress.
- No progress.
- Snehal segmented patella cartilage for oks001 following segmentation specifications. The goal was to take the patella cartilage segmentation / geometry generated by Craig and using the same specifications generate the geometry of the same specimen. This will be used to first run simple indentation simulation on both the geometries to evaluate any differences in model predictions due to variations in geometry (due to segmentation attempts by two users).
- From a visual assessment, Snehal thought that the two geometries look different. She will compare the surface distances to quantify these differences. Ahmet pointed out that if the error/ variations is within the resolution of the image set used then that should be an acceptable attempt. Also, if the variations in the finite element model predictions (for the given scenario) are not significant then the variations in geometry will not matter as much.