| 李欣欣,陈修婷,李杰,杨丽,汤晓敏,闫静,马诚诚,高之振.乳腺影像报告和数据系统分类联合乳腺X线摄影影像组学鉴别诊断乳腺良、恶性无定形钙化灶[J].中国医学影像技术,2024,40(10):1519~1523 |
| 乳腺影像报告和数据系统分类联合乳腺X线摄影影像组学鉴别诊断乳腺良、恶性无定形钙化灶 |
| Breast imaging reporting and data system classification combined with mammography radiomics for differentiating breast benign and malignant amorphous calcification lesions |
| 投稿时间：2024-01-12 修订日期：2024-05-26 |
| DOI：10.13929/j.issn.1003-3289.2024.10.013 |
| 中文关键词: 乳腺肿瘤 乳房X线摄影 影像组学 乳腺影像报告和数据系统 |
| 英文关键词:breast neoplasms mammography radiomics breast imaging reporting and data system |
| 基金项目:蚌埠医学院自然科学重点项目(2021byzd118)。 |
| 作者 | 单位 | E-mail | | 李欣欣 | 蚌埠医科大学第一附属医院放射科, 安徽 蚌埠 233004
蚌埠医科大学研究生院, 安徽 蚌埠 233030 | | | 陈修婷 | 蚌埠医科大学第一附属医院放射科, 安徽 蚌埠 233004
蚌埠医科大学研究生院, 安徽 蚌埠 233030 | | | 李杰 | 蚌埠医科大学研究生院, 安徽 蚌埠 233030 | | | 杨丽 | 蚌埠医科大学第一附属医院放射科, 安徽 蚌埠 233004 | | | 汤晓敏 | 蚌埠医科大学第一附属医院放射科, 安徽 蚌埠 233004 | | | 闫静 | 蚌埠医科大学第一附属医院放射科, 安徽 蚌埠 233004 | | | 马诚诚 | 蚌埠医科大学第一附属医院放射科, 安徽 蚌埠 233004 | | | 高之振 | 蚌埠医科大学第一附属医院放射科, 安徽 蚌埠 233004 | gaozhizhen269@163.com |
|
| 摘要点击次数: 439 |
| 全文下载次数: 212 |
| 中文摘要: |
| 目的 观察乳腺影像报告和数据系统(BI-RADS)分类联合乳腺X线摄影(MG)影像组学鉴别诊断乳腺良、恶性无定形钙化灶的价值。方法 回顾性分析206例女性乳腺无定形钙化病变患者共217个钙化灶,包括43个恶性、174个良性;按7[DK(]∶[DK)]3比例将病灶随机分为训练集(n=151)与验证集(n=66)。基于乳腺MG图筛选影像组学特征,分别建立BI-RADS模型、影像组学模型及二者联合模型,观察其鉴别诊断良、恶性病灶的价值。结果 最终选出8个最优影像组学特征。BI-RADS模型、影像组学模型及联合模型鉴别训练集乳腺良、恶性无定形钙化灶的曲线下面积(AUC)分别为0.821、0.763及0.897,在验证集分别为0.800、0.746及0.893,联合模型的AUC均显著高于其他模型(P均＜0.05)。结论 BI-RADS分类联合MG影像组学有助于鉴别诊断乳腺良、恶性无定形钙化灶,且效能高于单一方法。 |
| 英文摘要: |
| Objective To observe the value of breast imaging reporting and data system (BI-RADS) classification combined with mammography (MG) radiomics for differentiating breast benign and malignant amorphous calcification lesions. Methods A total of 217 breast amorphous calcification lesions in 206 female patients, including 43 malignant and 174 benign ones were retrospectively enrolled and randomly divided into training set (n=151) and validation set (n=66) at the ratio of 7∶3. Then radiomics features were screened based on breast MG images. BI-RADS model, radiomics model and combined model were established, and the value of each model for distinguishing benign and malignant lesions was observed. Results Eight optimal radiomics features were selected. The area under the curve (AUC) of BI-RADS model, radiomics model and combined model for differentiating breast benign and malignant amorphous calcification lesions was 0.821, 0.763 and 0.897 in training set, 0.800, 0.746 and 0.893 in validation set, respectively, AUC of combined model was significantly higher than that of the other two (both P＜0.05). Conclusion BI-RADS classification combined with MG radiomics was helpful for differential diagnosis of breast benign and malignant amorphous calcification lesions, with efficacy higher than that of each alone. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
