Introduction
The purpose of this post is to explain how to utilize medical imaging data in the development of a prosthetic implant. The two most common medical imaging technologies are CT and MRI. Both export a stack of 2D grey scale images over a 3D domain in the standard Digital Imaging and Communications in Medicine (DICOM) format. In this post I will go through the development of geometric (CAD) and mechanical (FEA) models based off anatomical imaging data. Through this workflow designs can be tuned for specific biometry based on realistic loading scenarios. As always all of the models used to develop this post are available at the end of the article.
Importing CT Scan
Inside Simpleware’s ScanIP, DICOM image data from a CT scan of a femur is imported. There are a variety of data sampling methods which are flexible enough for biomedical and even geological imaging. This post is going to focus on the major methodology steps and pass over details such as surface smoothing and filtering. Here, the raw DICOM data has been imported. The image stack has been partitioned based on the grey scale images and the solid bone geometry is shown below.
Once an appropriate solid representation of the femur is created the model is then opened with Simpleware’s +CAD module. An STL of the bone was exported so that the prosthetic implant can be appropriately located with respect to the femur within Catia.
Assembly of bone and implant
The STL of the femur is then imported into Catia using Digitized Shape Editor as a point cloud.
From here subsets of the femur cloud were created for the head and the medullary cavity (hollow center where the bone marrow is located).
After these two subsets were created Quick Surface Reconstruction was used to fit a sphere and cylinder to them respectively. These will be used in the assembly to mate the implant to the scan data. The center point of the sphere and the axis of the cylinder were used to create a plane for mating as well.
Here the implant has been located with respect to the femur. The following mates were created: spherical center points, planes created and the implant stem was mated to the distal end of the cylinder axis. A parametric implant model could be created to individually size the implant to the patient allowing for the axes to be mated. As you can see the stem of the implant is not perfectly aligned with the medullary cavity.
Merging Femur and Implant
The located implant was then exported from the Catia assembly as a STEP file for import into Simpleware’s +CAD module.
The implant was then fit with voxels (3D pixels) to be compatible with ScanIP’s workflow. The femur head was then removed with a planar cut. The remaining femur and implant were merged together retaining the boundary. A mesh was applied to the model and was exported as an INP file (Abaqus’s input file format).
Finite Element Model
Once in Abaqus a simple FEA model was created. The distal end of the femur was fixed and a pressure load was applied to the superior portion of the prosthetic head, simulating a standing load.
Simulation Results
A critical design feature of implants is the need for load sharing between the implant and remaining bone. If the implant is too stiff there is load shielding and the bone’s density will reduce. If the implant is too compliant there will be significant contact stresses. As with most critical design features a balance must be achieved and numerical simulation aides the designer in making informed engineering decisions. Several strain plots of the implant and bone are shown below.
Conclusion
The use of ScanIP, Catia and Abaqus provide the robust tools needed for orthopedic design. ScanIP enables use of the 3D image data that CT and MRIs produce. With this data implants can be appropriately sized and located with respect to the patient data in Catia. Simulating the mechanics of the system in Abaqus with FEA then drives design changes such as stem stiffness and cross sectional shape.
Realistically simulating the geometry and physics of the hip prosthesis system design studies can be made in silico instead of in vivo. This avoids costly clinical trials with respect to time and money for exploitative design studies. The end goal is to simulate with enough accuracy and confidence that the clinical trial is merely a formality to make sure nothing has been overlooked. This technology enables the ability to provide customized implants based on biometric data.
I hope you found this informative please feel free to comment, like, subscribe, contact us or whatever else can be done today! Thank you.
Rob Stupplebeen
Rob@OptimalDevice.com
I am a veterinary surgeon on the faculty at the University of TN and I would like to use this technology to create a finite element model of an animal bone and then digitally simulate various cuts and transpositions designed to alter the loading parameters acting on the bone. These cuts and transpositions have been advocated in the treatment of degenerative disease of the associated joint by shifting the load from abnormal to more normal regions of the joint. Will your programs be capable to achieve such an analysis? How much do they cost?
Hi Joseph,
This sounds like a very interesting use of the technology. On first pass, yes I believe that this workflow and solutions should work for you. Alas, I am unable to publish pricing but I will reach out to you directly. Thank you.
Rob
Hello. Thank you for your excellent article! I am using this technology for my thesis. I used MIMICS to generate a STL data. But I am confused, how can I create the medullary cavity in CATIA. Which commands should I use?
I am looking forward for your reply!
Hello Chen,
Thank you for the kind words! In this workflow I only used Catia to orient the implant by fitting a cylinder and sphere to the bone and implant. The medullary cavity was created in Simpleware’s ScanIP but you should be able to do it with Mimics as well.
If I were trying to do it with Catia check out this blog post doing similar operations for dental. Dental X-Ray to Bite Simulation
I hope this helps!
Rob