Machine learning and topological data analysis identify unique features of human papillae in 3D scans

Sep 17, 2026

Machine learning and topological data analysis are used on 3D microscopic scans of human tongue papillae (n=2092) to resolve how papillae are unique within and across individuals. The study computes geometric and topological features from discrete differential geometry and computational topology, then applies interpretable machine learning to relate persistent homology shape features to biological variables. Models with limited data achieve 85% accuracy for papillae type prediction and map spatial patterns of filiform and fungiform papillae. Individuals can be identified with 48% accuracy among 15 participants from a single sample, supporting papillae as unique identifiers for food preferences and oral diagnostics.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages26

Reading Time1 h 6 min

AccessFree to download

Abstract

Introduction

This is a repository copy of Machine learning and topological data analysis identify unique features of human papillae in 3D scans. White Rose Research Online URL for this paper: [https://eprints.whiterose.ac.uk/206735/](https://eprints.whiterose.ac.uk/206735/) [Version: Accepted Version](Version: Accepted Version) Article: Andreeva, R., Sarkar, [A. orcid.org/0000-0003-1742-2122 and Sarkar](A. orcid.org/0000-0003-1742-2122 and Sarkar), R. (2023) Machine learning and topological data analysis identify unique features of human papillae in 3Dscans. Scientific Reports, 13 (1) . 21529. [https://doi.org/10.1038/s41598-023-46535-9](https://doi.org/10.1038/s41598-023-46535-9) Reuse Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of the full text version. This is indicated by the licence information on the White Rose Research Online record for the item. Takedown If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing [eprints@whiterose.ac.uk](eprints@whiterose.ac.uk) including the URL of the record and the reason for the withdrawal request. [eprints@whiterose.ac.uk](eprints@whiterose.ac.uk)[ ](eprints@whiterose.ac.uk)[https://eprints.whiterose.ac.uk/](https://eprints.whiterose.ac.uk/) arXiv :2307 .06255v2 [ cs .LG] 27 Nov 2023 Machine learning and Topological data analysis identify unique features of human papillae in 3Dscans Rayna Andreeva 1 , Anwesha Sarkar2 , and Rik Sarkar 1,* 1 School of Informatics, University of Edinburgh, Edinburgh, United Kingdom 2 Food Colloids and Bioprocessing Group, School of Food Science and Nutrition, University of Leeds, Leeds, United Kingdom * [rsarkar@inf.ed.ac.uk](rsarkar@inf.ed.ac.uk) ABSTRACT The tongue surface houses a range of papillae that are integral to the mechanics and chemistry of taste and textural sensation. Although gustatory function of papillae is well investigated, the uniqueness of papillae within and across individuals remains elusive. Here, we present the 􀀂rst machine learning framework on 3D microscopic scans of human papillae (n = 2092), uncovering the uniqueness of geometric and topological features of papillae. The 􀀂ner differences in shapes of papillae are investigated computationally based on a number of features derived from discrete differential geometry and computational topology. Interpretable machine learning techniques show that persistent homology features of the papillae shape are the most effective in predicting the biological variables. Models trained on these features with small volumes of data samples predict the typeof papillae with an accuracy of 85% . The papillae type classi􀀂cation models can map the spatial arrangement of 􀀂liform and fungiform papillae on a surface. Remarkably, the papillae are found to be distinctive across individuals and an individual can be identi􀀂ed with an accuracy of 48% among the 15 participants from a single papillae. Collectively, this is the 􀀂rst unprecedented evidence demonstrating that tongue papillae can serve as a unique identi􀀂er inspiring new research direction for food preferences and oral diagnostics. Introduction The tongue is a highly sophisticated, heterogeneous anatomical structure and its operation is fundamental to speech, friction regulation and oral processing of food. The surface of the tongue is covered with tiny projections known as papillae which enable perception of taste, texture and oral mechanics. Of these numerous anatomical projections, fungiform papillae are considered as phenotypic markers of chemosensation of taste as they house the taste buds 1 , whereas 􀀂liform papillae that are devoid of taste buds are considered to be regulators of mec

FAQs

What problem does the study address about human papillae?

It investigates why papillae uniqueness within and across individuals is still unclear despite extensive study of their gustatory function.

Which features and methods are used to analyze the papillae shapes?

The work derives geometric and topological features from discrete differential geometry and computational topology, then uses interpretable machine learning with persistent homology features.

What predictive performance and identification results are reported?

Papillae type prediction reaches 85% accuracy, spatial mapping is possible for different papillae types, and individuals can be identified with 48% accuracy among 15 participants from a single sample.

Sep 17, 2026

Uploaded by Olivia Brown