Introduction: The purpose of this study was to determine if a statistical shape model (SSM) of the lower left first premolar, consisting of both crown and root data, could adequately describe the root form from a surface scan consisting of only crown data. Secondly, it tested if there were any angles or measurements on the tooth crown that correlate with any aspects of root morphology. The average orthodontist practicing today or in the near future is likely to use or own a digital intraoral scanner in their office. Yet optical scans only allow visualization of the crowns of teeth and external structures. We know that the orthodontic profession and the published literature support treatment of the teeth crowns and their roots in all three planes of space.1-7 Data acquired through CBCT imaging provides an accurate representation of the teeth and their roots, but it comes at a cost of relatively high radiation exposure.22-38 For this reason, the use of CBCT and other radiographic modalities to analyze orthodontic treatment is generally limited to the least use necessary.8 This study set out to find if statistical shape modeling could provide the practitioner with root form and/or positional data that could aid in patient care. Materials and Methods: Surface scans of 76 extracted mandibular first premolar teeth were entered into statistical software that created a statistical shape model from the population data and select landmark points. Then, using only the optical surface scans of 16 real patient crowns, the statistical model predicted a root form. Real patient roots, after being segmented from CBCT’s, were compared to the predicted roots and agreement was measured. Statistical analysis was performed using intraclass correlation tests and Euclidean Distance Matrix Analysis (EDMA), a technique used to compare biologic shapes using landmark points, to compare the 3D root shapes and dimensions. Spearman’s rho test was used to determine relationships within the 76 teeth population crown and root measurements. Results: The comparison between averaged real and predicted root forms using EDMA showed no significant differences. However, when an intraclass correlation coefficient test compared linear and angular measurements between individual real and predicted teeth forms, the agreement was weak or non-existent. For the population of 76 extracted mandibular first premolars, there were several different measurements and angles that showed moderate or weak agreement to each other. None of the tested measurements within the population showed strong, predictive correlation between crown and root measurements. Conclusions: For the mandibular first premolar, we were able to accurately predict root form from only optical crown scans when we averaged the real and predicted comparisons. On an individual level, the real and predicted teeth forms were statistically different. There were several moderate and weak agreements between measurements in the population of 76 extracted mandibular first premolars.

LLU Discipline

Orthodontics and Dentofacial Orthopedics


Orthodontics and Dentofacial Orthopedics


School of Dentistry

First Advisor

Caruso, Joseph M.

Second Advisor

Batesole, Mark K.

Third Advisor

Viecilli, Rodrigo

Degree Name

Master of Science (MS)

Degree Level


Year Degree Awarded


Date (Title Page)




Library of Congress/MESH Subject Headings

Cone-beam Computed Tomography; Tooth Radiography; Tooth Root - analysis; Orthodontics;

Subject - Local

Euclidean Distance Matrix analysis; Statistical Shape Model; Digital Imaging and Communications in Medicine;



Page Count


Digital Format


Digital Publisher

Loma Linda University Libraries

Usage Rights

This title appears here courtesy of the author, who has granted Loma Linda University a limited, non-exclusive right to make this publication available to the public. The author retains all other copyrights.


Loma Linda University Electronic Theses and Dissertations

Collection Website



Loma Linda University. Del E. Webb Memorial Library. University Archives