| 吴林桦,杨蔚,周晓平,刘开慧,李健.临床特征、乳房X线摄影及MRI表现鉴别非肿块型乳腺癌与乳腺炎[J].中国医学影像技术,2023,39(11):1653~1658 |
| 临床特征、乳房X线摄影及MRI表现鉴别非肿块型乳腺癌与乳腺炎 |
| Clinical features, mammography and MRI manifestations for differentiating non-mass breast cancer and mastitis |
| 投稿时间：2023-07-16 修订日期：2023-09-16 |
| DOI：10.13929/j.issn.1003-3289.2023.11.013 |
| 中文关键词: 乳腺肿瘤 乳腺炎 乳房X线摄影 磁共振成像 诊断,鉴别 非肿块样强化 |
| 英文关键词:breast neoplasms mastitis mammography magnetic resonance imaging diagnosis, differential non-mass enhancement |
| 基金项目:宁夏回族自治区重点研发计划项目(2022BEG03166)、宁夏医科大学总医院新入职硕士培养项目(YKDZY2022013)。 |
|
| 摘要点击次数: 1344 |
| 全文下载次数: 639 |
| 中文摘要: |
| 目的 观察联合应用临床特征、乳腺X线摄影(MG)及MRI表现鉴别非肿块型乳腺癌与乳腺炎的价值。方法 回顾性收集经病理证实的390例非肿块样强化(NME)乳腺癌或乳腺炎患者为开发组, 另前瞻 性招募同一医院159例乳 腺NME病变患者为验证组。对开发组临床、MG及MRI资料行单因素及多因素logistic回归分析,筛选鉴别非肿块型乳腺癌与乳腺炎的独立相关因素并建立临床-MG、临床-MRI-及临床-MG-MRI-模型,于验证组进行验证;采用受试者工作特征曲线评估模型诊断效能,以SHAP分析评估临床-MG-MRI模型中各参数的贡献价值。结果 共纳入549例549处NME病变,含408处乳腺癌及141处乳腺炎病灶,开发组含305处乳腺癌及85处乳腺炎,验证组含103处乳腺癌及56处乳腺炎。年龄、绝经状态,MG所示可疑钙化,MRI所示病变大小、分布、脂肪抑制T2WI信号强度、内部强化特征和时间-信号强度曲线类型均为鉴别非肿块型乳腺癌与乳腺炎的独立相关因素(P均＜0.05)。临床-MG-MRI模型在开发组的曲线下面积(AUC)(0.91)高于临床-MG(0.79)及临床-MRI模型(0.87)(Z=2.341、5.067,P均＜0.05),在验证组的AUC为0.90。临床-MG-MRI模型中,年龄、病变大小、可疑钙化和病变内部强化特征对于鉴别非肿块型乳腺癌与乳腺炎的贡献较大。结论 联合应用临床特征、MG及MRI表现能有效鉴别非肿块型乳腺癌与乳腺炎。 |
| 英文摘要: |
| Objective To explore the value of combination of clinical features, mammography (MG) and MRI manifestations for differentiating non-mass breast cancer and mastitis. Methods Totally 390 female patients with breast cancer with non-mass enhancement (NME) or mastitis confirmed by pathology were retrospectively enrolled in development group, while 159 patients with NME breast lesions admitted in the same hospital were prospectively recruited as validation group. Univariate and multivariate logistic regression were used to analyze clinical data, MG and MRI manifestations of lesions in development group, so as to screen the independent related factors for distinguishing non-mass breast cancer and mastitis and establish modelclinical-MG, modelclinical-MRI and modelclinical-MG-MRI, respectively, then the efficacy of each model was verified in validation group. Receiver operating characteristic curve was drawn to evaluate the diagnostic efficacy of each model, while the contribution value of each parameter to modelclinical-MG-MRI was evaluated using SHAP analysis. Results Totally 549 patients with 549 NME lesions were enrolled, including 408 breast cancers and 141 mastitis. There were 305 breast cancers and 85 mastitis in development group, 103 breast cancers and 56 mastitis were in validation group. Age, menopausal status, suspicious calcification shown on MG, and lesion size, distribution, fat suppression (FS) T2WI signal intensity, internal enhancement pattern and time-signal intensity curve type shown on MRI were all independent related factors for differentiating non-mass breast cancer and mastitis (all P＜0.05). Area under the curve (AUC) of modelclinical-MG-MRI (0.91) was higher than that of modelclinical-MG (0.79) and modelclinical-MRI (0.87) in development group (Z=2.341, 5.067, both P＜0.05), while in validation group was 0.90. The contribution of age, lesion size, suspicious calcification and internal enhancement pattern of lesions to modelclinical-MG-MRI for differentiating non-mass breast cancer and mastitis were significant. Conclusion Combination of clinical features, mammography (MG) and MRI manifestations could be used to effectively differentiate non-mass breast cancer and mastitis. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
