Exploring partial intrinsic and extrinsic symmetry in 3D medical imaging

We present a novel methodology to detect imperfect bilateral symmetry in CT of human anatomy. In this paper, the structurally symmetric nature pelvic bone is explored and used provide interventional image augmentation for treatment unilateral fractures patients with traumatic injuries. The mathematical basis our solution based on incorporation attributes characteristics that satisfy properties intrinsic extrinsic are robust outliers. first step, feature points automatically detected Möbius space defined data. These features then pruned via two-stage RANSAC attain correspondences also symmetry. Then, disparity function Tukey’s biweight estimator introduced minimized identify plane parametrization yields maximum contralateral similarity. Finally, regularization term enhance similarity between density histograms across partial plane, relying important biological observation that, even if injured, dislocated segments remain within body. Our extensive evaluations various cases common fracture types demonstrate validity concepts accuracy proposed method.