A robust comparison of biological shapes

Area of Science:

• Paleoanthropology
• Biomechanical analysis
• Comparative anatomy

Background:

• Skeletal analysis often relies on geometric morphometrics to quantify form.
• Least squares methods are commonly used for fitting comparative datasets.
• Identifying localized differences in complex skeletal structures presents challenges.

Purpose of the Study:

• To introduce and evaluate the repeated median resistant fitting algorithm for skeletal analysis.
• To compare the efficacy of resistant fitting versus least squares in detecting localized form differences.
• To validate the biological relevance of identified differences.

Main Methods:

• Development of the repeated median resistant fitting algorithm.
• Application of resistant fitting and least squares to compare two related animal skeletons.
• Comparative analysis of hominid skulls using both fitting techniques.

Main Results:

• Resistant fitting techniques more effectively determine localized skeletal differences compared to least squares.
• The repeated median resistant fitting algorithm successfully identified subtle variations.
• Differences detected by resistant fitting showed closer agreement with biological assessments.

Conclusions:

• Resistant fitting offers a more robust approach for analyzing localized skeletal variation.
• The repeated median algorithm provides a valuable tool for paleoanthropological and comparative anatomical studies.
• This method enhances the accuracy of identifying structurally significant differences in related skeletons.