Session: Government Agency Student Posters

Paper Number: 173609

Mri-Based Estimation of the Hip Joint Center in Infants Using Two Sphere-Fitting Methods 

Accurate estimation of the hip joint center (HJC) is a fundamental requirement in biomechanical modeling and clinical orthopedics, particularly for assessing joint mechanics, planning surgeries, and building personalized musculoskeletal models. In pediatric populations, especially infants, estimating the HJC poses unique challenges due to incomplete ossification of bone structures and limitations in traditional imaging techniques like computed tomography (CT). While magnetic resonance imaging (MRI) offers enhanced visualization of soft tissue and cartilaginous structures, established HJC estimation methods, particularly those based on femoral head geometry, may not be feasible in this age group due to poor bone delineation and developmental variability.

This study proposes and evaluates an alternative approach for HJC estimation using MRI-based sphere fitting of the ossified acetabular rim, which is consistently visible in comparison to the femoral head in early infancy. The primary goal is to determine whether acetabular-based methods can provide results comparable to the traditional gold-standard method of femoral head-based sphere fitting. By validating this approach, the study aims to lay the foundation for a CT-compatible method of estimating the HJC in infants, a critical need in the absence of reliable femoral head visualization.

Three-dimensional MRI datasets were obtained from three male infants (ages 10.7 to 12.7 months) from a pediatric hospital under IRB-approved protocols. Two patients had normal hip anatomy, and one exhibited mild dysplasia. For each patient, landmarks were manually placed on both the ossified acetabular rim and the cartilaginous boundary surrounding the femoral head using Simpleware ScanIP. Best-fit spheres were generated from each landmark set using a least-squares algebraic fitting method implemented in MATLAB. The center-to-center Euclidean distance between the femoral head and acetabular spheres was computed to quantify error between the two methods. Descriptive statistics (mean and standard deviation) were used to assess agreement.

Results demonstrated strong agreement between acetabular- and femoral-based HJC estimation methods in healthy hips. The average error across all healthy hips was 4.73 ± 1.27 mm. In the dysplastic case, a significantly larger error of 11.47 mm was observed in the left hip, which was visually assessed by a pediatric orthopedic surgeon to have reduced femoral head coverage. These findings suggest that acetabular-based sphere fitting is a feasible alternative in healthy infant hips, particularly in cases where femoral head landmarks are not visible, such as in CT scans. However, the elevated error in the dysplastic case highlights the method’s sensitivity to anatomical abnormalities and emphasizes the need for caution in applying it to pathological hips.

This study contributes to the growing body of pediatric biomechanical modeling by presenting a novel approach that leverages MRI visibility of ossified structures while supporting future application to CT datasets. The methodology may improve HJC estimation workflows for infants, offering new possibilities for motion analysis, surgical planning, and computational modeling in early childhood.

Limitations include a small sample size and the exclusion of formal inter-rater reliability analyses. Given the exploratory nature of the study, future work should expand the dataset to include a wider range of hip morphologies, incorporate statistical shape modeling techniques, and perform cross-modality comparisons between MRI and CT datasets. Despite these limitations, this work demonstrates a promising step toward more reliable methods for HJC estimation in early infancy.

Presenting Author: Victoria Melendez Embry-Riddle Aeronautical University

Presenting Author Biography: Victoria Meléndez Nieves is a junior at Embry-Riddle Aeronautical University, pursuing aBachelor of Science in Mechanical Engineering with a concentration in Biomedical Systems. Herresearch focuses on pediatric anatomical image segmentation and hip joint center estimation ininfants, leveraging 3D modeling techniques to improve clinical diagnostics and biomechanicalanalyses. Since 2023, she has actively contributed as an undergraduate research assistant indeveloping protocols for femoral neck axis measurements and medical image segmentationusing Simpleware ScanIP and MATLAB. Her contributions have been presented at nationalconferences, including the Orthopaedic Research Society (ORS) and the American Society ofMechanical Engineers (ASME) IMECE, where she was awarded a Travel Grant for her researchand published in the ASME Journal of Engineering and Science in Medical Diagnostics andTherapy. Her technical expertise includes medical image segmentation, statistical shapemodeling, and computational biomechanics, with proficiency in MATLAB, CATIA, SimplewareScanIP, and SolidWorks.

Authors:

Victoria Melendez Embry-Riddle Aeronautical University
Tamara Chambers Embry-Riddle Aeronautical University
Danielle Charpentier Embry-Riddle Aeronautical University
Vidyadhar Upasani Rady Children's Hospital
Christine Farnsworth Rady Children's Hospital
Victor Huayamave Embry-Riddle Aeronautical University