The clear cell carcinoma is a relatively rare tumor, represented approximately 85% by renal cell carcinoma (RCC), but has a characteristic of early metastatic, which sometimes spread to unusual places. Metastatic tumo...The clear cell carcinoma is a relatively rare tumor, represented approximately 85% by renal cell carcinoma (RCC), but has a characteristic of early metastatic, which sometimes spread to unusual places. Metastatic tumors in the paranasal sinuses are very rare. This case report a 65 years-old patient who presented sinusitis and epitaxis as a first sign of renal cell carcinoma with metastasis to the right maxillary sinus, which was opereted at Lagoa Hospital—RJ and strengthen the inclusion of this disease in the differential diagnosis of sinus disease. The renal cell carcinoma is a relatively rare tumor, representing approximately 3% of malignant tumors. About a third of patients with renal cell carcinoma have metastasis at diagnosis. The sinuses are a rare location of metastasis. The maxillary sinuses and the paranasal sinuses are most commonly affected by metastatic tumors to the sinuses region. The most frequent primary tumor to metastasis to paranasal sinuses is the renal cell carcinoma. This report highlights the rarity of the case and the importance in the differential diagnosis of sinus disease.展开更多
针对超声图像中肾脏、肾实质和肾窦自动分割存在的内部和边缘分割不准确问题,提出了一种用于肾脏超声图像分割的深度学习方法(GL-Unet11)。通过加深Unet网络,使网络具有更强的表达能力,并提出一个新的通道注意力网络——全局局部网络(GL...针对超声图像中肾脏、肾实质和肾窦自动分割存在的内部和边缘分割不准确问题,提出了一种用于肾脏超声图像分割的深度学习方法(GL-Unet11)。通过加深Unet网络,使网络具有更强的表达能力,并提出一个新的通道注意力网络——全局局部网络(GL-Net),该网络综合考虑了全局通道和局部通道对预测每个通道重要性的影响,使网络更好地关注重要信息。同时,将提出的全局局部网络加入到加深的Unet下采样部分的卷积块中,有效地增强了重要通道的特征。通过创建3个数据集,分别进行肾脏、肾实质和肾窦分割实验。实验结果表明,本方法在分割肾脏任务中,Dice系数达到96.25%,交并比(intersection over union,IOU)达到92.78%;在分割肾实质任务中,Dice系数达到92.90%,IOU达到86.75%;在分割肾窦任务中,Dice系数达到90.18%,IOU达到82.12%,分割结果均优于其它几种深度学习方法。该研究具有一定的创新性及医学应用价值。展开更多
文摘The clear cell carcinoma is a relatively rare tumor, represented approximately 85% by renal cell carcinoma (RCC), but has a characteristic of early metastatic, which sometimes spread to unusual places. Metastatic tumors in the paranasal sinuses are very rare. This case report a 65 years-old patient who presented sinusitis and epitaxis as a first sign of renal cell carcinoma with metastasis to the right maxillary sinus, which was opereted at Lagoa Hospital—RJ and strengthen the inclusion of this disease in the differential diagnosis of sinus disease. The renal cell carcinoma is a relatively rare tumor, representing approximately 3% of malignant tumors. About a third of patients with renal cell carcinoma have metastasis at diagnosis. The sinuses are a rare location of metastasis. The maxillary sinuses and the paranasal sinuses are most commonly affected by metastatic tumors to the sinuses region. The most frequent primary tumor to metastasis to paranasal sinuses is the renal cell carcinoma. This report highlights the rarity of the case and the importance in the differential diagnosis of sinus disease.
文摘针对超声图像中肾脏、肾实质和肾窦自动分割存在的内部和边缘分割不准确问题,提出了一种用于肾脏超声图像分割的深度学习方法(GL-Unet11)。通过加深Unet网络,使网络具有更强的表达能力,并提出一个新的通道注意力网络——全局局部网络(GL-Net),该网络综合考虑了全局通道和局部通道对预测每个通道重要性的影响,使网络更好地关注重要信息。同时,将提出的全局局部网络加入到加深的Unet下采样部分的卷积块中,有效地增强了重要通道的特征。通过创建3个数据集,分别进行肾脏、肾实质和肾窦分割实验。实验结果表明,本方法在分割肾脏任务中,Dice系数达到96.25%,交并比(intersection over union,IOU)达到92.78%;在分割肾实质任务中,Dice系数达到92.90%,IOU达到86.75%;在分割肾窦任务中,Dice系数达到90.18%,IOU达到82.12%,分割结果均优于其它几种深度学习方法。该研究具有一定的创新性及医学应用价值。
