Summarising statistical shape modelling in the human hip
Interview with lead author of the scoping review, 'Application of statistical shape modeling to the human hip joint'
What was your motivation for undertaking the scoping review?
Our work is motivated by the long-term outcomes of paediatric hip disorders. Many problems like hip dysplasia or Perthes’ disease can result in changes in the shape of the hip joint. Ultimately, these changes are understood to be a key factor in the development of osteoarthritis later in life.
We often rely on 2D X-rays to describe the hip’s shape in the clinic, and it isn’t clear whether they adequately describe the 3D anatomy of the hip. However, 3D methods like MRI or CT scans have their own limitations, and X-rays will remain the foundation of clinical imaging for the foreseeable future, so it is important to make the most of them. To do this, we need to better understand the relationship between 2D radiographs and 3D shape.
We see Statistical Shape Modelling (SSM) as a way to “bridge the gap” between 2D and 3D representations of anatomy. SSM is a type of statistical analysis that describes the complex shape differences that exist across a population in just a small handful of parameters. SSM has been used previously across fields from computer vision to evolutionary biology. We performed this scoping review to get a better understanding of the field before starting our own studies using SSM.
Who will find this study most useful, and what value does this study have for them?
We hope that this study will help researchers who are interested in using SSM for their own work. We aimed to provide a summary of the different methods that have been used for modelling, particularly the kinds of populations that have been modelled and how corresponding landmarks are placed on the anatomy. Many of the details presented in this review are transferable to related fields.
With that said, this scoping review was developed as a starting point. Those familiar with SSM will likely find less detail than they’d like, but hopefully the review will still provide something new: perhaps a study they haven’t heard of previously, or a question they hadn’t thought of raising. In fact, we have suggested some areas for future research based on gaps we have identified in the current literature.
What are the key points or ‘takeaways’ from the scoping review?
The main takeaway is that there is a lot out there! We found more than 100 journal articles on the subject, covering a huge variety of methods, applications, and populations. For prospective researchers, this means that there is plenty of inspiration available to help answer your research questions. There is a lot of mature software available to make the modelling process easier. Examples include ASM Toolkit, BoneFinder, and ShapeWorks. We are using the latter for our own studies.
A lot of work has been done in SSM of the hip, but there is a lot still to do to fill in gaps in the literature. For example, many populations are underrepresented in modelling work, and most models are produced in-house for each individual study and can’t be applied elsewhere. There are many opportunities to tackle some of these issues: one opportunity is for researchers to produce and open-source ‘reference models’ that can then be taken forward and applied by other researchers in their own studies. Care should be taken that the training populations for these reference models are appropriately representative of the global population of interest, and that these models are thoroughly validated.
As a final comment, this was a team effort. This was the first scoping review that I and some of my co-authors had ever undertaken, and we learned a great deal about the process. The review wouldn’t have gone anywhere without my wonderful co-authors, who put in a huge number of hours and provided plenty of guidance. Thank you all!
Resources
‘Application of statistical shape modeling to the human hip joint: a scoping review’ Johnson, Luke G. et al
JBI Evidence Synthesis 21(3):p 533-583, March 2023. | DOI: 10.11124/JBIES-22-0017 https://journals.lww.com/jbisrir/Fulltext/2023/03000/Application_of_statistical_shape_modeling_to_the.8.aspx?context=FeaturedArticles&collectionId=2