# Fuzzy masks: boosting radiomic reliability in head and neck tumors amid delineation uncertainty

## TL;DR

A new fuzzy mask method improves the reliability of radiomic features in head and neck tumors by accounting for contour uncertainty.

## Contribution

FuzzMask introduces gradient transitions to model tumor contour uncertainty, enhancing radiomic reliability.

## Key findings

- FuzzMask yields up to 29 more reliable features than binary masks in head and neck cancers.
- FuzzMask increases model reliability with ICC values up to 0.99 for predictive outputs.

## Abstract

•Fuzzy mask models tumor contour uncertainty via gradient transitions.•The method yields up to 29 more reliable features than binary masks.•Gradient weighting produces 2% more independent feature clusters.•Intraclass correlation coefficient of the predictive outputs of model up to 0.99.•Intensity equalization mechanisms drive the observed reliability gains.

Fuzzy mask models tumor contour uncertainty via gradient transitions.

The method yields up to 29 more reliable features than binary masks.

Gradient weighting produces 2% more independent feature clusters.

Intraclass correlation coefficient of the predictive outputs of model up to 0.99.

Intensity equalization mechanisms drive the observed reliability gains.

The clinical utility of radiomics in head-and-neck (H&N) cancer is hindered by poor reliability caused by delineation uncertainties from the use of binary mask (BinMask). This study introduced a fuzzy mask (FuzzMask) approach to enhance the reliability of computed tomography (CT)-based radiomics for precision prognosis.

This retrospective study included 2,539 H&N cancer patients (855 laryngeal cancer (LC), 1,336 oropharyngeal cancer (OPC), 348 nasopharyngeal carcinoma (NPC)). Delineation uncertainty was simulated via perturbation techniques. Radiomic features (RFs) were extracted using BinMask and FuzzMask, respectively. The evaluation focused on feature reliability and relevance via the intraclass correlation coefficient (ICC) and hierarchical clustering. In addition, the predictive performance and output reliability of penalized Cox’s proportional hazard models were assessed using the concordance index (C-index) and ICC, respectively.

The FuzzMask improved feature reliability, yielding 21, 29, and 5 additional reliable features for LC, OPC, and NPC cancers, respectively, compared to BinMask. The FuzzMask also reduced feature redundancy, generating up to 70 more clusters in hierarchical clustering, particularly for smaller tumors in complex peritumoral environments although showed marginal improvements in predictive performance (C-index: +0.1% for OPC, +0.4% for NPC, p > 0.05). However, model reliability was enhanced by FuzzMask, with ICC values increasing by 0.024, 0.022, and 0.007 for LC, OPC, and NPC, respectively, compared to BinMask (p ≥ 0.05).

The proposed FuzzMask technique significantly improved feature reliability and model robustness against delineation uncertainty, offering greater trustworthiness for clinical translation, although predictive accuracy remains unaffected.