Hierarchical Classification via Cascading Feature Elimination: Application to Human Phenotype Ontology-Aligned Facial Phenotyping (FaceMesh2HPO)

2026-07-06

AI Summary

This study addresses the challenge of fine-grained facial phenotyping in rare disease diagnosis by proposing an automated 3D facial phenotype classification method aligned with the Human Phenotype Ontology (HPO). Leveraging a 3D facial mesh represented by 478 landmarks, the approach innovatively integrates a hierarchical PointNet architecture with a cascaded feature elimination strategy to predict phenotypic terms progressively along the HPO hierarchy. Demographic metadata are incorporated to enhance model interpretability and ontological consistency. Evaluated on 107 HPO terms, the model achieves AUROC scores ranging from 0.55 to 0.89, with superior performance observed at parent nodes compared to leaf nodes. External validation further demonstrates the model’s capacity for cross-disease generalization, highlighting its potential utility in clinical phenotyping pipelines.

Abstract

FaceMesh2HPO is a framework for classifying facial phenotypic descriptors aligned with the Human Phenotype Ontology (HPO) to support clinical diagnosis. Using annotations from 124 clinicians across 10 disorders (107 HPO terms) combined with non-syndromic controls, we generated 3D facial meshes (478 landmarks) from 2D images and trained a hierarchical PointNet-based pipeline with cascading classification and feature elimination. The best models, incorporating 3D meshes, facial outline, and demographic metadata, achieved AUROCs between ~0.55 and ~0.89, with higher performance at parent nodes than leaf terms. External validation showed variable generalizability across disorders. Results demonstrate that hierarchical modeling of 3D facial geometry enables interpretable, ontology-linked phenotype classification, though performance on rare leaf terms remains limited. Improved data diversity and feature selection strategies are needed to enhance robustness and clinical utility.

Problem

• Hierarchical Classification
• Human Phenotype Ontology
• Facial Phenotyping
• 3D Facial Geometry
• Rare Phenotype Classification

Innovation

• hierarchical classification
• cascading feature elimination
• 3D facial mesh
• Human Phenotype Ontology
• PointNet