Benchmarking off-the-shelf statistical shape modeling tools in clinical applications

Anupama Goparaju,Alexandre Bone,Nan Hu,Heath B. Henninger,Andrew E. Anderson,Stanley Durrleman,Matthijs Jacxsens,Alan Morris,Ibolya Csecs,Nassir Marrouche,Shireen Y. Elhabian

DOI: https://doi.org/10.48550/arXiv.2009.02878

2020-09-07

Abstract:Statistical shape modeling (SSM) is widely used in biology and medicine as a new generation of morphometric approaches for the quantitative analysis of anatomical shapes. Technological advancements of in vivo imaging have led to the development of open-source computational tools that automate the modeling of anatomical shapes and their population-level variability. However, little work has been done on the evaluation and validation of such tools in clinical applications that rely on morphometric quantifications (e.g., implant design and lesion screening). Here, we systematically assess the outcome of widely used, state-of-the-art SSM tools, namely ShapeWorks, Deformetrica, and SPHARM-PDM. We use both quantitative and qualitative metrics to evaluate shape models from different tools. We propose validation frameworks for anatomical landmark/measurement inference and lesion screening. We also present a lesion screening method to objectively characterize subtle abnormal shape changes with respect to learned population-level statistics of controls. Results demonstrate that SSM tools display different levels of consistencies, where ShapeWorks and Deformetrica models are more consistent compared to models from SPHARM-PDM due to the groupwise approach of estimating surface correspondences. Furthermore, ShapeWorks and Deformetrica shape models are found to capture clinically relevant population-level variability compared to SPHARM-PDM models.

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to evaluate and verify the effectiveness and applicability of off - the - shelf statistical shape modeling (SSM) tools in clinical applications. Specifically, the paper focuses on the performance of these tools in clinical applications such as implant design based on morphometry and lesion screening. By systematically evaluating three widely - used SSM tools - ShapeWorks, Deformetrica and SPHARM - PDM, the paper explores their performance differences in different application scenarios. The authors use quantitative and qualitative indicators to evaluate the shape models generated by these tools and propose a validation framework for anatomical landmark/measurement inference and lesion screening. In addition, the paper also introduces a lesion screening method, aiming to objectively characterize subtle abnormal shape changes based on population - level statistical data of the control group.

The core problems of the paper can be summarized as follows:
1. **Evaluating the performance of SSM tools in clinical applications**: especially in applications such as implant design and lesion screening.
2. **Verifying the reliability and consistency of SSM tools**: by comparing the performance of different tools when processing the same data set, evaluate their consistency and reliability.
3. **Proposing a specific validation framework**: providing a systematic validation method for anatomical landmark inference and lesion screening to ensure the effective application of these tools in clinical practice.

Through the exploration of these problems, the paper aims to provide a more objective and reliable basis for tool selection in clinical decision - making, thereby improving the accuracy and efficiency of clinical applications.