Objectives Cholesteatomas (CSs) are locally aggressive and can infiltrate adjacent structures as they expand. This study aimed to establish a straightforward scoring system using dual-layer detector dual-energy CT (DL-DECT) to improve the identification of cholesteatomas in patients. Materials and methods Between August 2023 and July 2024, 871 patients with soft tissue density shadows in the ear region who underwent DL-DECT examination were retrospectively enrolled at our institute. Surgical treatments followed by pathological examinations were performed. Based on pathological findings, the lesions were classified into cholesteatoma (CS) and non-cholesteatoma (non-CS) groups. The diagnostic performance of anatomical and quantitative parameters derived from DL-DECT was evaluated using receiver operating characteristic (ROC) curve analysis. Logistic regression was applied to develop a diagnostic scoring system. Results A total of 87 patients (median age, 51 years; 45 men and 42 women) with suspected temporal bone CS were included, comprising 44 CS cases and 43 non-CS cases. The effective atomic number (Zeff) demonstrated the highest diagnostic accuracy (area under the curve [AUC] = 0.786), followed by the slope of the energy spectral curve (AUC = 0.784), scutum destruction (AUC = 0.759), and CT40 keV (AUC = 0.724). Logistic regression identified two significant predictors, which were incorporated into the scoring system. When the system score reached 2 points (Zeff ≤ 7.12 accompanied by scutum destruction), the AUC in the ROC analysis reached 0.868 (95% confidence interval: 0.778–0.931), significantly outperforming each individual parameter (all p < 0.05). Conclusion The DL-DECT-derived scoring system serves as an innovative imaging marker for detecting CSs. Key Points <Emphasis Type="BoldItalic">Question</Emphasis> Accurate differentiation between CSs and non-CSs is critical for selecting surgical approaches. However, high-resolution CT demonstrates limited discriminatory power. <Emphasis Type="BoldItalic">Findings</Emphasis> A straightforward diagnostic scoring system incorporating Zeff ≤ 7.12 and scutum destruction was developed to efficiently identify patients with CS. <Emphasis Type="BoldItalic">Clinical relevance</Emphasis> This scoring system may facilitate the early identification of CS, potentially improving patient outcomes through timely intervention. Graphical Abstract

Temporal bone dual-layer detector dual-energy computed tomography for identifying cholesteatoma

期刊论文
发表日期：

2025年8月9日

查看全文

Xiaoxue Fan,
Changwei Ding,
Sizhe Gao,
Xiaomei Lu,
Xiulan Ma,
Yang Hou

摘要

Objectives

Cholesteatomas (CSs) are locally aggressive and can infiltrate adjacent structures as they expand. This study aimed to establish a straightforward scoring system using dual-layer detector dual-energy CT (DL-DECT) to improve the identification of cholesteatomas in patients.

Materials and methods

Between August 2023 and July 2024, 871 patients with soft tissue density shadows in the ear region who underwent DL-DECT examination were retrospectively enrolled at our institute. Surgical treatments followed by pathological examinations were performed. Based on pathological findings, the lesions were classified into cholesteatoma (CS) and non-cholesteatoma (non-CS) groups. The diagnostic performance of anatomical and quantitative parameters derived from DL-DECT was evaluated using receiver operating characteristic (ROC) curve analysis. Logistic regression was applied to develop a diagnostic scoring system.

Results

A total of 87 patients (median age, 51 years; 45 men and 42 women) with suspected temporal bone CS were included, comprising 44 CS cases and 43 non-CS cases. The effective atomic number (Zeff) demonstrated the highest diagnostic accuracy (area under the curve [AUC] = 0.786), followed by the slope of the energy spectral curve (AUC = 0.784), scutum destruction (AUC = 0.759), and CT40 keV (AUC = 0.724). Logistic regression identified two significant predictors, which were incorporated into the scoring system. When the system score reached 2 points (Zeff ≤ 7.12 accompanied by scutum destruction), the AUC in the ROC analysis reached 0.868 (95% confidence interval: 0.778–0.931), significantly outperforming each individual parameter (all p < 0.05).

Conclusion

The DL-DECT-derived scoring system serves as an innovative imaging marker for detecting CSs.

Key Points

Question Accurate differentiation between CSs and non-CSs is critical for selecting surgical approaches. However, high-resolution CT demonstrates limited discriminatory power.

Findings A straightforward diagnostic scoring system incorporating Zeff ≤ 7.12 and scutum destruction was developed to efficiently identify patients with CS.

Clinical relevance This scoring system may facilitate the early identification of CS, potentially improving patient outcomes through timely intervention.

Graphical Abstract