采用影像组学的肾肿瘤组织学亚型分类

为了在术前更准确、非侵入地鉴别乏脂肪血管平滑肌脂肪瘤（fp-AML）和肾透明细胞癌（ccRCC）,提出一种基于CT图像的影像组学模型. 从CT图像中提取774个三维的影像组学特征;分三步进行特征选择：计算皮尔森相关矩阵剔除冗余特征,使用Welch’s t检验确定具有显著差异的特征,利用序列浮动前向选择算法选择具有鉴别能力的特征;使用基于稀疏学习的径向基函数神经网络进行分类. 结果表明：该模型获得的正确率、敏感度、特异性和受试者工作特征曲线下面积分别为90.00%、66.67%、100.0%和0.9173. 利用分类器的输出概率进行模型的可靠性评估,当概率阈值为0.95时,该模型获得的自信正确率、未定率和错分率分别为71.67%、25.00%和3.33%,结果表明所提出的影像组学模型能可靠地对fp-AML和ccRCC进行分类.     

影像学检查发现肾上腺血管平滑肌脂肪瘤一例并文献分析

目的：探讨肾上腺血管平滑肌脂肪瘤的影像学表现和诊断方法。方法：对一例肾上腺血管平滑肌脂肪瘤的患者的影像和随诊资料进行分析,并对相关文献进行分析。结果：本例患者有高血压病史,超声检查二维声像图表现为左侧肾上腺区实性等回声肿块,边界清晰,形态规则。彩色多普勒显像（CDFl）示肿瘤周边及内部未见明显血流信号。CT表现为圆形软...     

血供丰富的乏脂肪肾脏错构瘤与肾透明细胞癌的CT鉴别诊断

Objective: The aim of our study was to investigate the feature of minimal fat renal angiomyolipoma with sufficient blood supply using CT scans and improve the diagnosis accuracy required to differentiate it from clear cell renal carcinoma.Methods: Retrospective analysis of 24 cases of post-surgery confirmed angiomyolipoma with sufficient blood supply (total of 25 tumors) in our hospital that were used for a pathological comparison study. Results: Among the 24 patients diagnosed with angiomyolipoma, nobody had bloody urine. Of the 96 patients diagnosed with clear cell renal cancer, 14 had bloody urine (14.6%). In our studied group, the size of angiomyolipomas with sufficient blood supply was between 1.5 cm×2.0 cm to 8.0cm × 10.0 cm. During CT scan analysis, twenty tumors had similar density, and five of them had higher density. Only one tumor had a few dots of calcification (4%). Adipose tissue was not visible in 9 tumors, while 16 tumors had visible dots of adipose tissue, as visualized by CT scan. Intensive scanning indicated that all of the tumors showed a strong enhancement in the renal corticomedullary phase. Twenty tumors had significant heterogeneous enhancement in the early phase, while another set of five cases had homogenous prolonged enhancement. Nineteen patients had surgery to remove the angiomyolipomas, while six patients had single side kidney removal due to misdiagnosis for renal cancer in cases where the tumor severely compromised the renal parenchyma and sinus. All 25 cases were classified as renal angiomyolipoma by pathological analysis. Within the 96 cases of clear cell renal cancer, 64 tumors had relatively low density, 29 tumors had equal density, and 3 cases had relatively higher density. Fourteen of the tumors had calcification (14.6% ), and none of them had visualized adipose tissue.Enhanced CT scans indicated that 69 cases of renal cancer showed significant enhancement in the renal corticomedullary phase, which had the abnormal pattern of"fast-in-and-fast-out". Additionally, 27 cases had slow and prolonged enhancement.Conclusion: Similar to clear cell renal carcinomas, angiomyolipomas with sufficient blood supply also appear to exhibit abnormal enhancement with a pattem of "fast-in-and-fast-out" during the early phase, which is easily misdiagnosed as renal cancer.It is difficult to differentiate them merely through CT scans; the key to differentiating them is to identify the adipose tissue within the tumor. Therefore, it is helpful to use thin-layer CT scans to locate the adipose tissue.     

Mri of renal angiomyolipomas

The MR appearance of 13 renal angiomyolipomas in 11 patients were reviewed. The fatty tissue in angiomyolipomas which demonstrated hyperintensity on Tl-weighted imaging was considered to be a key point for diagnosis with MRI. Renal angiomyolipomas were classified into three types according to the proportion of fatty tissue in tumours. Fatty tissue comprises more than 80% of the volume of an angiomyolipoma in type I, 20–80% in type II and less than 20% in type III tumours. It was easy to diagnose types I and II renal angiomyolipomas with MRI, but it was quite difficult to differentiate type III renal angiomyolipomas from other solitary occupying lesions such as renal cell carcinoma. Other MR characteristics of renal angiomyolipomas are summarized and are thought to have diagnostic uses.