基于数字图像相关的骨折检测中股骨皮质应变的估算方法

The overall goal of this experimental strain measurement technique is to quantitatively estimate femoral surface strain from high-speed video of bio-mechanical fracture testing. This method can help answer key questions in osteoporosis research, such as development of advanced diagnostic tools for prediction of bone strength. The main advantage of this technique is that femoral surface strain can be computed with high spatial resolution adequate for computation model validation.

Visual demonstration of this method is critical, as the specimen preparation and analysis steps are difficult to explain solely in words due to the unique methods of speckling the specimen with particles and the custom scripting tools used in the analysis. Begin by thawing the femora at room temperature for 24 hours. When the femur is ready for testing, remove any wrap that was applied prior to freezing, and wipe the femur with a dry towel to remove any remnant moisture, fatty deposits, or soft tissues.

Then put the greater trochanter into a pre-fabricated aluminum cup with bone cement. Using a box to contain the paint overspray particles as much as possible, spray the bone with white plastic primer to achieve a thin, uniform coating. Take care to cover the bone with one uniform layer of paint for optimal contrast and strong adhesion to the femur surface.

Let the paint dry for at least 5 minutes, and then wrap the bone with wet cloth to avoid tissue dryness. Next, for bone speckling, gradually add one part water to two parts black acrylic paint to achieve the proper paint viscosity. Dip a clean toothbrush in a palette of black paint to absorb the color, and flick the brush to make black speckles over the white coating of the bone.

Finally, let the paint dry for 5 minutes before proceeding with the image acquisition. Begin by mounting the prepared speckled femur in the mechanical test machine by inserting the potted distal end into the fixture and tightening two screws to secure the specimen. Next, adjust the lighting source to minimize shadows and the reflections on the femur's surface, while striving to achieve the highest illumination possible in the camera image.

Reduce the aperture of the front and back view high-speed video camera lenses so that the entire region of interest of the femur in the field of view is in focus. Then, readjust the light source to further improve illumination while minimizing glare. Set the image acquisition software to capture 6, 000 frames per second at a resolution of 1, 024 by 512 pixels and acquire 12, 288 total frames.

Then, conduct the fracture test. After testing is complete, use the image acquisition software to save the video to disk. Use video analysis software to open the appropriate high-speed video recording and note two key frame reference numbers.

One at the start of load frame actuator movement, and the other immediately after the fracture event. Then, to down sample an uncompressed TIF image sequence from the high-speed video, open and run the MOVframes. m script in the working directory for the pertinent femur side.

Finally, in the resulting dialogue box, enter the final reference number for the frame immediately after the fracture with a step size of 25 to 40. Click extract frames, and inspect the working directory to ensure the TIF files were extracted correctly. The following series of steps is required to use the finite element method to calculate 2D strains from the differential displacement vectors.

To begin, open the finite element meshing program to create the finite element mesh. Import the first key frame TIF image into the pre-processor as a template for SPLINE creation. Then, find two fiduciary points in the image that are at opposite corners of the frame and record their X-and Y-coordinates.

Open this image within an image editing software, and record the x and y direction values of the pixels associated with each fiduciary point. Next, in the sketch module of the finite element meshing program, use the SPLINE tool to outline a closed section representing the region of interest. Under the menu Seed Part Instance, seed the edges of this closed section with a global mesh size of one millimeter.

Under Assign Mesh Controls, set the element shape to quadrilateral and mesh the closed section. Next, start a new session within the matrix algebra scripting program and create two element row vectors called ab1 and ab2. Enter the vector names of the two previously identified fiduciary and pixel coordinate pairs into the command line.

Save the workspace as Points.met. Run the script convertimagesize. m to register the points from the finite element mesh with the extracted high-speed video image.

Then run the script rrimagetrackgui.m. Load the first image, and enter the number of the last TIF file that was extracted as the total number of images to process. To put the mesh over the bone image, load the mesh by selecting read from file and click apply.

Specify the tracking values based on the guidelines for tracking parameters. Select a guidepoint that has significant contrast around it, while avoiding areas with any glare or blurriness. Check this point by clicking Check guide and to verify the correlation peak is strong compared to its neighbors.

If the correlation peak is satisfactory, click accept, and perform tracking. After tracking is complete, select animate. Finally, once animate has finished, calculate the 2D strains by clicking right strains.

In the dialogue, enter exe in the file name area and select rightstrainrrsimple.exe. After calculations are complete, close the graphical user interface. The digital image correlation method requires a carefully prepared high contrast speckling pattern and sufficient illumination.

Otherwise, the results may be affected by several issues such as oversaturation of the surface, undesirable mixing of black and white paint when the white paint is still wet, and poor contrast in the speckling pattern. Here, the outline of the femur sample is used to identify the region of interest for strain field estimation, and for creation of a finite element mesh for strain calculation. The onset of fracture is detected by monitoring the degree of strain deviation during testing with peaks representing bone damage and final frame of fracture.

Finally, 2D strain fields are superimposed back onto the untested bone image for enhanced visualization. Once mastered, this technique can be completed in 2 hours if it is performed properly. While attempting this procedure, it is important to remember to uniformly cover the surface of the specimen with high-contrast speckling pattern for the best image signal-to-noise ratio.

Following this procedure, other methods like QCT/FEA mechanical fracture simulations on the femur can be performed in order to answer additional questions like patient-specific bone strength and fracture risk. After its development, this technique paved the way for researchers in the field of computational bone strength assessment to explore hip fracture risk in osteoporotic patients. After watching this video, you should have a good understanding on how to perform digital image correlation on bone specimen subjected to biomechanical casts, including the steps of specimen preparation, high-speed image acquisition, and calculations of strain through surface displacement tracking.