An Automated Statistical Shape Model Developmental Pipeline: Application to the Human Scapula and Humerus

Abstract

This paper presents development of statistical shape models based on robust and rigid-groupwise registration followed by pointset nonrigid registration. The main advantages of the pipeline include automation in that the method does not rely on manual landmarks or a regionalization step; there is no bias in the choice of reference during the correspondence steps and the use of the probabilistic principal component analysis framework increases the domain of the shape variability. A comparison between the widely used expectation maximization-iterative closest point algorithm and a recently reported groupwise method on publicly available data (hippocampus) using the well-known criteria of generality, specificity, and compactness is also presented. The proposed method gives similar values but the curves of generality and specificity are superior to those of the other two methods. Finally, the method is applied to the human scapula, which is a known difficult structure, and the human humerus.

Stacked sparse autoencoder networks and statistical shape models for automatic staging of distal femur trochlear dysplasia

“…As already anticipated, the computation of the mean model depends on the opportune selection of a reference shape for the dataset and the accurate definition of point‐to‐point correspondence. In the literature, both conditions have been acknowledged to be critical, affecting the representative power and extrapolation property of the SSM . To note, the reference shape is only used to determine correspondences while is not used to compute the mean model.…”

“…Among different approaches, statistical shape models (SSM) have emerged as promising tools to the task for their ability to represent continuous morphologic variations of anatomical shapes, both globally and locally, including pathologic abnormalities. Statistical shape models, proposed in the early literature to model many different categories of digital objects, have been more recently adopted to represent the morphologic variability intrinsic to both soft and rigid body parts with applications in medical image processing, computer graphics, anatomic quantitative analysis, computer‐assisted planning, and surgery , …”

“…Statistical shape models were exploited to automate image segmentation of the liver, to reconstruct the human scapula and humerus for anatomic studies, and to quantify human femoral cortex morphology . In orthopedic surgery, SSM of the proximal femur were applied to reconstruct the 3D model from sparse data .…”

“…Statistical shape models, proposed in the early literature to model many different categories of digital objects, 25 have been more recently adopted to represent the morphologic variability intrinsic to both soft and rigid body parts with applications in medical image processing, computer graphics, anatomic quantitative analysis, computer-assisted planning, and surgery. 26,27,[28][29][30][31][32][33] Statistical shape models were exploited to automate image segmentation of the liver, 26 to reconstruct the human scapula and humerus for anatomic studies, 27 and to quantify human femoral cortex morphology. 28 In orthopedic surgery, SSM of the proximal femur were applied to reconstruct the 3D model from sparse data.…”

Mesh correspondence improvement using Regional Affine Registration: Application to Statistical Shape Model of the scapula

“…The authors used the Iterative Median Closest Point approach [16] to create a mean shape and the Coherent Point Drift algorithm [17] to perform the elastic deformation.…”