Construction of an Aggressiveness Diagnostic Model for Basal Cell Carcinoma Using Multimodal Ultrasound Features

This study aimed to develop a nomogram-based diagnostic model to assess basal cell carcinoma (BCC) aggressiveness by integrating clinical data and multimodal ultrasound imaging features, thereby establishing an evidence-based framework for clinical decision-making.

A retrospective analysis of 120 pathologically confirmed BCC lesions (January 2017-December 2024) was conducted. Lesions were classified as high-risk BCCs (n = 30) or low-risk BCCs (n = 90) based on histopathology. Clinical and multimodal ultrasound features acquired using a 5- to 18-MHz linear array transducer were compared between groups. Logistic regression identified predictors of aggressiveness, and a nomogram was subsequently developed and validated.

Univariate analysis identified morphological configuration, infiltration level (IL), intralesional echoic pattern (IEP), distribution of hyperechoic foci, maximum infiltration depth (MID), and average elastic Young's modulus (Eave) as factors significantly associated with BCC aggressiveness (p < .05). Multivariate logistic regression analysis revealed that IL (p = .047), IEP (p = .020), MID (p = .047), and Eave (p = .023) were independent predictors of high-risk BCC. Receiver operating characteristic curve analysis demonstrated that the nomogram model exhibited robust predictive performance, with a concordance index (C-index) of 0.921 (95% confidence interval: 0.860-0.981).

The multimodal ultrasound-based nomogram effectively predicts BCC aggressiveness, offering a noninvasive tool for clinical assessment and treatment planning.