Synopsys Cloud
Cloud native EDA tools & pre-optimized hardware platforms
Request a Free Trial →
Using Simpleware Software to Automate Orthopedic Revision Planning
Posted on 23 November 2020 by Kerim Genc
Orthopedic revision surgeries are typically needed if an implanted orthopedic device fails or causes complications due to factors like general wear and tear, further injury, and/or infection. Although proportionally rare for procedures like total hip and knee replacements, revision surgeries can be needed to address specific patient problems and replace faulty implants. Planning for a revision surgery typically begins with X-rays and other testing to identify clinically relevant issues. However, due to the complexity of revision surgeries, 3D image scans (MRI and/or CT) may be needed to get a deeper understanding of the patient’s specific anatomy and the implant that needs to be removed. This approach helps the surgeon consider sizing and positioning of the new implant and create templates to inform their surgical planning.
Segmentation of a hip revision case using Simpleware AS Ortho.
When working with clinical CT and MRI data towards building a patient-specific surgical plan, one of the most significant bottlenecks involves manual segmentation and landmarking. In this context, segmentation is the identification of different regions of interest (ROIs) within the anatomy, such as bone, cartilage and the implant. At this stage, a great deal of time might be spent on carrying out the first stages of segmentation through manual and laborious processes.
A new solution to this problem, then, is provided by Simpleware AS Ortho, which frees up time by using AI-powered Machine Learning (ML) algorithms to automate initial segmentation. Rather than spend time identifying different anatomical regions manually, these are automatically inferred by the software, with a standard segmentation result including masks and landmarks as a 2D and 3D output. The automated segmentation distinguishes between femurs, pelvis and sacrum, and generates common landmarks, with additional options to edit the results, and export the final model for CAD or FEA. In the area of orthopedic revision planning, considerable savings are made in terms of time spent on each case, resulting in more efficient workflows.
Tackling Different Revision Cases
For common primary surgical cases, even on patients with arthritic hips, Simpleware AS Ortho makes it much easier to extract relevant information from a CT scan to make an assessment and build a model. Although segmentation of the hip and femur in primary surgical cases may be considered one of the easier categories of segmentation, Simpleware AS Ortho is still able to save a significant amount of time from the traditional manual or semi-automatic methods.
Case 1: Example of a common primary surgical case workflow using Simpleware AS Ortho compared with a standard segmentation approach used by a technician/engineer with an intermediate expertise level.
When tackling revision cases where an implant is already in the body, and it is necessary to segment metallic material from the rest of the subject’s anatomy, carrying out manual segmentation from scratch to separate the implant stem and acetabular ball can be a very time-consuming process. Although Simpleware AS Ortho may not be able to complete the segmentation in a single click, like the typical primary surgical (Case 1), it can still provide a much better starting point where the technician/engineer can take advantage of the Simpleware software environment to perform the necessary additional clean-up (Case 2).
Case 2: Example of Simpleware AS Ortho used for a more complex orthopedic case with an intermediate level of noise caused by the device already implanted in the patients left hip.
For even more difficult revisions, such as those where the image data contains an extreme amount of noise, or there are unique pathologies, automatic segmentation still offers benefits. More specifically, Simpleware AS Ortho can handle most of the segmentation tasks, removing unwanted artefacts or scatter and leaving the user with much fewer issues to manually tidy up. Effectively, this approach provides a significantly better starting position, potentially taking the worktime from approximately 8-10 hours of manual work to 0.5 hours using Simpleware AS Ortho.
Case 3: Example of Simpleware AS Ortho used for more complex orthopedic case with a high level of noise caused by the device already implanted in the patients left hip.
Improving Patient Outcome
The main advantage of using the AI in Simpleware AS Ortho for revision surgery planning therefore comes down to a reduction in tedious manual segmentation and landmarking. Engineers and technicians who use Simpleware AS Ortho will free up more of their time for higher value tasks, shortening the process of going from an initial patient scan through to a surgical plan or simulation. With automated segmentation, user error and inconsistencies during segmentation are removed or greatly reduced, with less technical expertise needed to handle complex revision cases.
While orthopedic revision surgeries typically focus on the hips and knees, there is also scope to use the same approach to automate working with other anatomical regions such as the shoulder, spine, ankle etc. The established performance of the ML algorithms is adaptable to new training data, meaning that specific anatomical solutions are available depending on the needs of the individual user. To this end, a Simpleware AS Cardio module is also available for the heart, while a fully personalized option is available through a Simpleware Custom Modeler module.
Learn More
Watch our 20-minute presentation and demo to see Simpleware AS Ortho in action, segmenting different hip revision surgical planning cases.
Any Questions?
If you would like to know more, please contact us. Our technical specialists will be happy to show you Simpleware AS Ortho and our other Machine Learning-based solutions for any anatomy.