目的

探讨高帧频超声造影（HFR-CEUS）在乳腺超声影像报告和数据系统（BI-RADS）分类为4 类小结节（最大径≤20 mm）良、恶性鉴别诊断中的优势和应用价值。

方法

选取2022 年9 月至2023 年12 月在苏州大学附属第一医院就诊的乳腺结节患者75 例（共计82 个结节病灶），以病理诊断为依据，将结节分良性组（58 个）和恶性组（24 个）。所有病灶均在取得病理诊断结果前进行了常规超声造影（C-CEUS）和HFR-CEUS 检查，分别记录2 种超声造影检查灌注模式的特征（包括增强方向、增强强度、增强后边缘特征、增强范围、滋养血管）和时间-强度曲线相关参数：峰值强度、上升时间、达峰时间、平均渡越时间、流入相曲线下面积，并采用χ² 检验、Fisher 检验或秩合检验比较2组的组间差异；分析2 种超声造影模式对乳腺小结节良、恶性鉴别的敏感度、特异度和准确性，绘制受试者操作特征（ROC）曲线并采用Delong 检验比较2 种造影模式的曲线下面积的差异。

结果

HFRCEUS 模式下，良性组病灶离心性增强方向（37.9%）、低增强（13.8%）、增强后范围无明显扩大（91.4%）及增强后边缘光滑（70.7%）的显示率均高于C-CEUS 模式（17.2%、1.7%、75.9%、46.6%），差异具有统计学意义（χ²=6.238、6.562、5.098、7.663，P=0.044、0.033、0.024、0.046）；恶性组病灶增强后边缘特征呈放射状的显示率（45.8%）较C-CEUS 模式（16.7%）显著升高，差异具有统计学意义（χ²=5.807，P=0.047）。HFR-CEUS 对乳腺小结节良、恶性鉴别的诊断效能高于C-CEUS（曲线下面积：0.937 vs 0.853），差异具有统计学意义（Z=2.488，P=0.013）；HFR-CEUS 的诊断特异度（91.4%）、准确性（90.2%），相较于C-CEUS（特异度为67.2%，准确性为74.4%）均有所提高。

结论

HFR-CEUS 通过提高帧频、改善图像对比度及分辨率，提升了乳腺小结节动脉早期灌注过程及边缘特征的显示能力，可提高对乳腺BI-RADS 4 类小结节良、恶性鉴别诊断的效能。

Objective

To assess the benefits and application potential of high-frame-rate contrastenhanced ultrasound (HFR-CEUS) in differentiating small benign from malignant breast nodules categorized as BI-RADS 4 nodules (maximum diameter≤20 mm).

Methods

The study involved 75 patients with a total of 82 nodules, treated at the First Affiliated Hospital of Suzhou University from September 2022 to December 2023.Based on pathological diagnosis, the nodules were classified into benign (58 nodules) and malignant groups (24 nodules).Prior to the pathological diagnosis, all patients underwent conventional CEUS (C-CEUS) and HFR-CEUS; characteristics of perfusion patterns for each method were recorded,including enhancement direction, intensity, edge characteristics post-enhancement, enhancement range, and nutrient vessels, alongside time-intensity curve-related parameters (peak intensity, rise time, time to peak,mean transit time, and wash-in area under the curve).These attributes were compared between the groups by χ² test, Fisher test or rank sum test to evaluate their differences.Additionally, the sensitivity, specificity,and accuracy of both ultrasound modalities in distinguishing benign from malignant nodules were analyzed.Receiver operating characteristic (ROC) curves were generated and the area under the curve (AUC) values of both modalities were compared by Delong test.

Results

In HFR-CEUS mode, benign lesions predominantly showed eccentric enhancement direction (37.9%), low enhancement intensity (13.8%), limited expansion of enhancement range (91.4%), and better-defined edges (70.7%) compared to conventional CEUS (17.2%,1.7%, 75.9%, and 46.6%, respectively), with statistically significant differences between the two groups(χ²=6.238, 6.562, 5.098, and 7.663; P=0.044, 0.033, 0.024, and 0.046, respectively).The incidence of radial edge features in the malignant group (45.8%) was significantly higher compared to that of the C-CEUS group (16.7%; χ²=5.807, P=0.047).Furthermore, HFR-CEUS demonstrated higher diagnostic efficacy than C-CEUS in differentiating benign from malignant breast nodules (AUC: 0.937 vs 0.853), with a statistically significant difference (Z=2.488, P=0.013).The diagnostic specificity (91.4%) and accuracy (90.2%) of HFRCEUS were also significantly improved compared to those of C-CEUS (specificity 67.2%, accuracy 74.4%).

Conclusion

HFR-CEUS enhances the visualization of early perfusion processes and edge features in small breast nodules by increasing the frame rate, and improving image contrast and resolution, thereby enhancing the diagnostic performance for benign and malignant small BI-RADS4 breast nodules.