Skin cancer is a serious and potentially life-threatening disease that affects millions of people worldwide. Early detection and accurate diagnosis are critical for successful treatment and improved patient outcomes. ...Skin cancer is a serious and potentially life-threatening disease that affects millions of people worldwide. Early detection and accurate diagnosis are critical for successful treatment and improved patient outcomes. In recent years, deep learning has emerged as a powerful tool for medical image analysis, including the diagnosis of skin cancer. The importance of using deep learning in diagnosing skin cancer lies in its ability to analyze large amounts of data quickly and accurately. This can help doctors make more informed decisions about patient care and improve overall outcomes. Additionally, deep learning models can be trained to recognize subtle patterns and features that may not be visible to the human eye, leading to earlier detection and more effective treatment. The pre-trained Visual Geometry Group 16 (VGG16) architecture has been used in this study to classification of skin cancer images, and the images have been converted into other color scales, there are named: 1) Hue Saturation Value (HSV), 2) YCbCr, 3) Grayscale for evaluation. Results show that the dataset created with RGB and YCbCr images in field condition was promising with a classification accuracy of 84.242%. The dataset has also been evaluated with other popular architectures and compared. The performance of VGG16 with images of each color scale is analyzed. In addition, feature parameters have been extracted from the different layers. The extracted layers were felt with the VGG16 to evaluate the ability of the feature parameters in classifying the disease.展开更多
The deep learning method automatically extracts advanced features from a large amount of data, avoiding cumbersome manual feature screening, and using digital pathology and artificial intelligence technology to build ...The deep learning method automatically extracts advanced features from a large amount of data, avoiding cumbersome manual feature screening, and using digital pathology and artificial intelligence technology to build a computer-aided diagnosis system to help pathologists quickly make objective and reliable diagnoses and improve work efficiency. Because pathological images are limited by factors such as sample size, manual labeling expertise, and complexity, artificial intelligence algorithms have not been extensively and in-depth researched on pathological images of lung cancer metastasis. Therefore, this paper proposes a lung cancer metastasis segmentation method based on pathological images, to further improve the computer-aided diagnosis method of lung cancer.展开更多
目的探讨酰胺质子转移加权(amide proton transfer weighted,APTw)与动态对比增强磁共振成像(dynamic contrast enhanced MRI,DCE-MRI)序列评估宫颈癌神经周围侵犯(perineural invasion,PNI)的价值。材料与方法回顾性分析36例行盆腔3.0 ...目的探讨酰胺质子转移加权(amide proton transfer weighted,APTw)与动态对比增强磁共振成像(dynamic contrast enhanced MRI,DCE-MRI)序列评估宫颈癌神经周围侵犯(perineural invasion,PNI)的价值。材料与方法回顾性分析36例行盆腔3.0 T MRI检查(包括APTw、DCE-MRI序列)且手术病理证实为宫颈癌的患者病例及影像资料,其中有PNI(PNI组)12例,无PNI(NPNI组)24例。由两位观察者分别测量病灶的APT值与DCE-MRI定量参数值,包括容积转移分数(volume transfer constant,K^(trans))、速率常数(exchange rate between EES and blood plasma,K_(ep))、血管外细胞外间隙容积分数(extravascular volume fraction,V_(e))以及血浆容积分数(capillary plasma volume,V_(p))。采用组内相关系数(intra-class correlation coefficient,ICC)检验2位观察者对各参数值测量结果的一致性;采用Kolmogorov-Smirov检验数据是否符合正态分布,通过两独立样本t检验或Mann-Whitney U检验比较两组间参数值的差异,采用受试者工作特征(receiver Operating Characteristic,ROC)曲线评估有差异参数诊断PNI效能,获得相应的曲线下面积(area under the curve,AUC)、阈值、敏感度和特异度。采用二元logistic回归计算有差异参数的联合诊断效能,DeLong检验进行各参数和联合参数AUC比较,Spearman相关分析检测APT值和有差异DCE-MRI参数间的相关性。结果两位观察者测得的APT值及K^(trans)值、K_(ep)值、V_(e)值、V_(p)值结果一致性良好,ICC均>0.75。两组间的APT值和V_(p)值差异有统计学意义(P<0.05),K^(trans)、K_(ep)、V_(e)差异无统计学意义(P>0.05)。PNI组的APT值(2.89%±0.72%)和V_(p)值[7.80×10^(-3)(6.80×10^(-3),1.14×10^(-2))]均大于NPNI组[APT值2.31%±0.71%;V_(p)值4.19×10^(-3)(2.04×10^(-3),7.35×10^(-3))]。评估宫颈癌PNI时,APT值和V_(p)值的AUC分别为0.717、0.785,阈值分别为2.7%及6.46×10^(-3),敏感度及特异度分别为66.7%及75.0%、83.3%及75.0%;APT值联合V_(p)值后的AUC为0.792,APT值、V_(p)值与两者联合后的AUC之间差异无统计学意义(P>0.05)。APT值与V_(p)值无相关性(r=0.219,P=0.198)。结论APTw序列及DCE-MRI的定量参数均能有效预测宫颈癌PNI,具有一定临床应用价值。展开更多
目的观察基于MR-T2WI的深度迁移学习(deep transfer learning,DTL)特征、影像组学特征及临床特征构建的联合模型(列线图)在术前预测宫颈癌淋巴脉管间隙浸润(lymph vascular space invasion,LVSI)的价值。材料与方法回顾性分析178例经术...目的观察基于MR-T2WI的深度迁移学习(deep transfer learning,DTL)特征、影像组学特征及临床特征构建的联合模型(列线图)在术前预测宫颈癌淋巴脉管间隙浸润(lymph vascular space invasion,LVSI)的价值。材料与方法回顾性分析178例经术后病理证实为宫颈癌的患者病例,其中70例LVSI(+)、108例LVSI(-),按照8∶2划分为训练集[142例,54例LVSI(+)、88例LVSI(-)]和测试集[36例,16例LVSI(+)、20例LVSI(-)]。对临床因素行单因素logistic分析,筛选出LVSI(+)独立预测因素。使用DTL方法和传统影像组学方法提取矢状位T2WI图像中病灶的DTL特征和影像组学特征,构建DTL特征数据集、影像组学特征数据集和DTL特征与影像组学特征融合的数据集,分别以t检验、Pearson分析和最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)回归对训练集各特征数据集进行特征降维,以其最佳者构建影像组学(radiomics,Rad)模型、DTL模型、融合模型(Rad+DTL模型),并筛选最佳影像组学模型;基于上述最佳影像组学模型评分与临床独立因子构建联合模型,并绘制列线图。以校准曲线评估模型校准度,以决策曲线分析评价模型的应用价值。结果淋巴结转移、粒细胞比率均为LVSI(+)的独立预测因子(P<0.05)。Rad+DTL模型为最佳影像组学模型。联合模型在训练集和测试集中的受试者工作特征曲线下面积(area under the curve,AUC)高于Rad+DTL模型(0.984 vs.0.966,P<0.05;0.912 vs.0.759,P=0.05)。联合模型的校准度较高,临床净收益更大。结论基于MR-T2WI的DTL特征、影像组学特征联合临床特征构建的联合模型可有效预测宫颈癌LVSI。展开更多
目的探讨酰胺质子转移加权成像(amide proton transfer weighted imaging,APTw)的影像组学术前预测宫颈癌淋巴血管间隙侵犯(lymphovascular space invasion,LVSI)的价值。材料与方法回顾性分析经手术病理证实的宫颈癌患者病例及影像资...目的探讨酰胺质子转移加权成像(amide proton transfer weighted imaging,APTw)的影像组学术前预测宫颈癌淋巴血管间隙侵犯(lymphovascular space invasion,LVSI)的价值。材料与方法回顾性分析经手术病理证实的宫颈癌患者病例及影像资料66例。所有患者均行盆腔3.0 T MRI检查,包括轴位T2WI、矢状位T2WI、动态对比增强磁共振成像(dynamic contrast enhanced magnetic resonance imaging,DCE-MRI)和3D-APTw序列扫描。在APTw-T2WI融合图像上对肿瘤实质区域进行感兴趣区(region of interest,ROI)勾画并记录APT值。在APT重建图像上进行肿瘤病灶分割并提取影像组学特征。采用组内相关系数(intra-class correlation coefficient,ICC)选取观察者内和观察者间复测信度好的影像组学特征(ICC>0.900)。采用递归特征消除法(recursive feature elimination,RFE)及最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)算法进行特征降维和筛选。基于logistic回归分类器构建临床模型、APTw影像组学模型和联合组学模型。采用受试者工作特征(receiver operating characteristic,ROC)曲线和决策曲线分析(decision curve analysis,DCA)评估模型的诊断效能和临床价值,采用DeLong检验比较不同模型的预测效能。结果在训练集中,APTw影像组学模型预测宫颈癌LVSI的效能高于临床模型(AUC=0.826 vs.0.675),差异有统计学意义(DeLong检验P<0.05)。联合组学模型在训练集和测试集中的AUC值分别为0.838和0.825。DeLong检验结果显示,联合组学模型在训练集中术前评估LVSI的效能显著高于临床模型和APTw影像组学模型(P均<0.05)。决策曲线显示APTw影像组学模型和联合组学模型在训练集和测试集中均具有较高的临床价值。结论基于APTw的影像组学模型在术前预测宫颈癌LVSI方面具有较高的潜力,联合临床因素能进一步提高预测效能,有望为宫颈癌患者的个体化治疗和预后评估提供重要的支持。展开更多
文摘Skin cancer is a serious and potentially life-threatening disease that affects millions of people worldwide. Early detection and accurate diagnosis are critical for successful treatment and improved patient outcomes. In recent years, deep learning has emerged as a powerful tool for medical image analysis, including the diagnosis of skin cancer. The importance of using deep learning in diagnosing skin cancer lies in its ability to analyze large amounts of data quickly and accurately. This can help doctors make more informed decisions about patient care and improve overall outcomes. Additionally, deep learning models can be trained to recognize subtle patterns and features that may not be visible to the human eye, leading to earlier detection and more effective treatment. The pre-trained Visual Geometry Group 16 (VGG16) architecture has been used in this study to classification of skin cancer images, and the images have been converted into other color scales, there are named: 1) Hue Saturation Value (HSV), 2) YCbCr, 3) Grayscale for evaluation. Results show that the dataset created with RGB and YCbCr images in field condition was promising with a classification accuracy of 84.242%. The dataset has also been evaluated with other popular architectures and compared. The performance of VGG16 with images of each color scale is analyzed. In addition, feature parameters have been extracted from the different layers. The extracted layers were felt with the VGG16 to evaluate the ability of the feature parameters in classifying the disease.
文摘The deep learning method automatically extracts advanced features from a large amount of data, avoiding cumbersome manual feature screening, and using digital pathology and artificial intelligence technology to build a computer-aided diagnosis system to help pathologists quickly make objective and reliable diagnoses and improve work efficiency. Because pathological images are limited by factors such as sample size, manual labeling expertise, and complexity, artificial intelligence algorithms have not been extensively and in-depth researched on pathological images of lung cancer metastasis. Therefore, this paper proposes a lung cancer metastasis segmentation method based on pathological images, to further improve the computer-aided diagnosis method of lung cancer.
文摘目的探讨酰胺质子转移加权(amide proton transfer weighted,APTw)与动态对比增强磁共振成像(dynamic contrast enhanced MRI,DCE-MRI)序列评估宫颈癌神经周围侵犯(perineural invasion,PNI)的价值。材料与方法回顾性分析36例行盆腔3.0 T MRI检查(包括APTw、DCE-MRI序列)且手术病理证实为宫颈癌的患者病例及影像资料,其中有PNI(PNI组)12例,无PNI(NPNI组)24例。由两位观察者分别测量病灶的APT值与DCE-MRI定量参数值,包括容积转移分数(volume transfer constant,K^(trans))、速率常数(exchange rate between EES and blood plasma,K_(ep))、血管外细胞外间隙容积分数(extravascular volume fraction,V_(e))以及血浆容积分数(capillary plasma volume,V_(p))。采用组内相关系数(intra-class correlation coefficient,ICC)检验2位观察者对各参数值测量结果的一致性;采用Kolmogorov-Smirov检验数据是否符合正态分布,通过两独立样本t检验或Mann-Whitney U检验比较两组间参数值的差异,采用受试者工作特征(receiver Operating Characteristic,ROC)曲线评估有差异参数诊断PNI效能,获得相应的曲线下面积(area under the curve,AUC)、阈值、敏感度和特异度。采用二元logistic回归计算有差异参数的联合诊断效能,DeLong检验进行各参数和联合参数AUC比较,Spearman相关分析检测APT值和有差异DCE-MRI参数间的相关性。结果两位观察者测得的APT值及K^(trans)值、K_(ep)值、V_(e)值、V_(p)值结果一致性良好,ICC均>0.75。两组间的APT值和V_(p)值差异有统计学意义(P<0.05),K^(trans)、K_(ep)、V_(e)差异无统计学意义(P>0.05)。PNI组的APT值(2.89%±0.72%)和V_(p)值[7.80×10^(-3)(6.80×10^(-3),1.14×10^(-2))]均大于NPNI组[APT值2.31%±0.71%;V_(p)值4.19×10^(-3)(2.04×10^(-3),7.35×10^(-3))]。评估宫颈癌PNI时,APT值和V_(p)值的AUC分别为0.717、0.785,阈值分别为2.7%及6.46×10^(-3),敏感度及特异度分别为66.7%及75.0%、83.3%及75.0%;APT值联合V_(p)值后的AUC为0.792,APT值、V_(p)值与两者联合后的AUC之间差异无统计学意义(P>0.05)。APT值与V_(p)值无相关性(r=0.219,P=0.198)。结论APTw序列及DCE-MRI的定量参数均能有效预测宫颈癌PNI,具有一定临床应用价值。
文摘目的观察基于MR-T2WI的深度迁移学习(deep transfer learning,DTL)特征、影像组学特征及临床特征构建的联合模型(列线图)在术前预测宫颈癌淋巴脉管间隙浸润(lymph vascular space invasion,LVSI)的价值。材料与方法回顾性分析178例经术后病理证实为宫颈癌的患者病例,其中70例LVSI(+)、108例LVSI(-),按照8∶2划分为训练集[142例,54例LVSI(+)、88例LVSI(-)]和测试集[36例,16例LVSI(+)、20例LVSI(-)]。对临床因素行单因素logistic分析,筛选出LVSI(+)独立预测因素。使用DTL方法和传统影像组学方法提取矢状位T2WI图像中病灶的DTL特征和影像组学特征,构建DTL特征数据集、影像组学特征数据集和DTL特征与影像组学特征融合的数据集,分别以t检验、Pearson分析和最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)回归对训练集各特征数据集进行特征降维,以其最佳者构建影像组学(radiomics,Rad)模型、DTL模型、融合模型(Rad+DTL模型),并筛选最佳影像组学模型;基于上述最佳影像组学模型评分与临床独立因子构建联合模型,并绘制列线图。以校准曲线评估模型校准度,以决策曲线分析评价模型的应用价值。结果淋巴结转移、粒细胞比率均为LVSI(+)的独立预测因子(P<0.05)。Rad+DTL模型为最佳影像组学模型。联合模型在训练集和测试集中的受试者工作特征曲线下面积(area under the curve,AUC)高于Rad+DTL模型(0.984 vs.0.966,P<0.05;0.912 vs.0.759,P=0.05)。联合模型的校准度较高,临床净收益更大。结论基于MR-T2WI的DTL特征、影像组学特征联合临床特征构建的联合模型可有效预测宫颈癌LVSI。
文摘目的探讨酰胺质子转移加权成像(amide proton transfer weighted imaging,APTw)的影像组学术前预测宫颈癌淋巴血管间隙侵犯(lymphovascular space invasion,LVSI)的价值。材料与方法回顾性分析经手术病理证实的宫颈癌患者病例及影像资料66例。所有患者均行盆腔3.0 T MRI检查,包括轴位T2WI、矢状位T2WI、动态对比增强磁共振成像(dynamic contrast enhanced magnetic resonance imaging,DCE-MRI)和3D-APTw序列扫描。在APTw-T2WI融合图像上对肿瘤实质区域进行感兴趣区(region of interest,ROI)勾画并记录APT值。在APT重建图像上进行肿瘤病灶分割并提取影像组学特征。采用组内相关系数(intra-class correlation coefficient,ICC)选取观察者内和观察者间复测信度好的影像组学特征(ICC>0.900)。采用递归特征消除法(recursive feature elimination,RFE)及最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)算法进行特征降维和筛选。基于logistic回归分类器构建临床模型、APTw影像组学模型和联合组学模型。采用受试者工作特征(receiver operating characteristic,ROC)曲线和决策曲线分析(decision curve analysis,DCA)评估模型的诊断效能和临床价值,采用DeLong检验比较不同模型的预测效能。结果在训练集中,APTw影像组学模型预测宫颈癌LVSI的效能高于临床模型(AUC=0.826 vs.0.675),差异有统计学意义(DeLong检验P<0.05)。联合组学模型在训练集和测试集中的AUC值分别为0.838和0.825。DeLong检验结果显示,联合组学模型在训练集中术前评估LVSI的效能显著高于临床模型和APTw影像组学模型(P均<0.05)。决策曲线显示APTw影像组学模型和联合组学模型在训练集和测试集中均具有较高的临床价值。结论基于APTw的影像组学模型在术前预测宫颈癌LVSI方面具有较高的潜力,联合临床因素能进一步提高预测效能,有望为宫颈癌患者的个体化治疗和预后评估提供重要的支持。
