Fuzzy conceptual spaces for dental shade analysis: A perceptually grounded approach with an open-source python framework

Accurate tooth shade matching remains a challenging task in restorative dentistry due to the subjective nature of color perception and the limitations of visual and instrumental methods. This study presents a perceptually grounded and extensible artificial intelligence framework for dental shade analysis based on fuzzy conceptual spaces, implemented in the open-source Python library PyFCS (Python Fuzzy Color Spaces). The proposed methodology supports the construction of fuzzy color categories from dental shade guides and enables pixel-wise chromatic mapping and region-wise analysis of segmented dental samples. The framework was instantiated with the 16 shades of the VITA Classical Shade Guide. An initial controlled evaluation was conducted with twelve dental experts using an interface integrated into PyFCS . Expert-based annotations were collected for tooth regions and compared with system outputs using the full fuzzy Jaccard similarity metric, complemented by a secondary partial-overlap analysis. The highest agreement was observed in the middle third of the tooth, with a full fuzzy Jaccard similarity of 0 . 68 ± 0 . 30 and a partial-overlap Jaccard similarity of 0 . 76 ± 0 . 20 , whereas lower agreement and higher complete-disagreement rates were found in the cervical and incisal regions. These findings suggest that fuzzy conceptual spaces can provide an interpretable representation of perceptual ambiguity and gradual transitions between dental shades under controlled conditions. By supporting adaptation to different dental shade guides and institutional protocols, this contribution provides a flexible basis for the development of expert-informed, perceptually grounded tools for dental color analysis and future clinically oriented shade selection research.