Automatic counting of fungiform papillae by shape using cross-correlation

Area of Science:

• Biomedical Engineering
• Sensory Science
• Computational Biology

Background:

• Accurate enumeration of fungiform papillae (FP) is crucial for understanding taste sensitivity.
• Manual counting of FP is labor-intensive and subjective.
• Current methods often require specialized staining techniques.

Purpose of the Study:

• To develop and validate a novel, automated algorithm for counting human fungiform papillae (FP).
• To assess the accuracy and efficiency of the proposed algorithm compared to manual expert counts.
• To provide a user-friendly and rapid method for FP quantification.

Main Methods:

• A novel algorithm utilizing cross-correlation based on shape characteristics was developed to identify and count FP.
• The algorithm's performance was evaluated against manual counts performed by human experts on identical tongue images.
• Statistical analysis, including fitting a Poisson regression model using maximum likelihood, was employed to compare results.

Main Results:

• The automated algorithm demonstrated high accuracy, with counts closely mirroring those of human experts.
• The algorithm offers significant advantages in terms of ease of use and speed.
• The method functions effectively on unstained, plain tongue images, eliminating the need for sample preparation.

Conclusions:

• The proposed cross-correlation algorithm provides a reliable and efficient automated method for counting fungiform papillae.
• This technique offers a valuable tool for taste sensitivity studies, improving upon traditional manual counting methods.
• The algorithm's simplicity and speed make it suitable for widespread application in sensory research.