多方位MRI影像组学预测局部进展期直肠癌新辅助放化疗疗效

目的:采用MRI影像组学方法,提取局部进展期直肠癌(LARC)病灶影像组学特征,并联合临床及常规影像特征构建预测模型,探讨模型对新辅助放化疗(nCRT)疗效的预测效能。方法:回顾性分析209例LARC患者nCRT前的临床及影像资料。LARC患者在nCRT后6～8周行全系膜切除术(TME),并评估肿瘤病理退缩分级(TRG)。按疗效分为nCRT反应良好组(TRG0～1级)和反应不良组(TRG2～3级),按照1:1随机分为训练组和验证组对模型进行内部验证。手动勾画T₂WI序列横轴面(TRA)、矢状面(SAG)及冠状面(COR)图像提取影像组学特征,采用LASSO回归筛选特征并构建影像组学标签。通过多因素logistics分析筛选nCRT疗效的独立预测因子并构建联合预测模型,采用ROC曲线及校正曲线对模型进行评估,并使用临床决策曲线评价模型的临床价值。结果:209例患者中nCRT反应良好组61例,反应不良组148例。T₂WI序列横轴面、矢状面、冠状面图像各提取379个影像组学特征,ICC为0.9的特征中TRA 96个,SAG 88个,COR 91个。...     

CT影像组学预测局部进展期直肠癌新辅助治疗的效果

目的探讨CT影像组学在治疗前预测局部进展期直肠癌新辅助治疗效果的价值。资料与方法回顾性分析168例新辅助治疗后行根治术的局部进展期直肠癌患者,收集治疗前临床及CT资料,根据术后病理肿瘤退缩分级分组。采用A.K.软件提取CT影像组学特征并构建影像组学标签。通过多变量Logistic回归筛选疗效预测因子并构建诺莫图模型。利用ROC曲线评价模型诊断效能,并对模型进行内部验证、校准度评价及临床应用价值分析。结果每例患者各提取了396个CT影像组学特征,降维后筛选出6个与局部进展期直肠癌新辅助治疗效果高度相关的特征。联合独立预测因子影像组学标签、癌胚抗原≥3.4 ng/ml和临床T分期(cT4)构建的诺莫图模型ROC曲线下面积(0.881)高于影像组学标签(0.791),且具有较高的校准度、内部验证一致性及临床应用价值(P>0.05)。结论基于治疗前CT及临床资料构建的模型对局部进展期直肠癌新辅助治疗效果预测具有较高的预测效能,且联合预测模型的预测效能优于影像组学标签。     

基于高分辨率T2WI影像组学联合临床特征预测食管癌新辅助放化疗疗效的价值

【摘要】目的：探讨基于高分辨率（HR）T2WI影像组学联合临床特征预测食管癌新辅助放化疗后疗效的价值。方法：回顾性分析本院2016年1月-2021年12月新辅助放化疗前接受HRT2WI成像检查并经病理证实的95例食管癌患者资料。依据新辅助放化疗后病理缓解状态结果将患者疗效分为缓解组和未缓解组,在HRT2WI图像上勾画肿瘤感兴趣区（ROI）后采用A.K软件提取影像组学特征,采用最大相关最小冗余（mRMR）算法进行降维,采用逻辑回归模型对筛选出的影像组学特征及临床参数构建模型;采用受试者操作特征（ROC）曲线评估不同模型的预测效能,计算曲线下面积（AUC）、准确率、敏感度和特异度,并采用DeLong检验比较不同模型预测食管癌新辅助放化疗敏感性的效能。结果：缓解组与未缓解组年龄差异具有统计学意义（P=0.001）,其他临床特征差异无统计学意义（P＞0.05）。从1688个组学特征中逐层筛选出4个影像组学特征,构建两个预测模型：基于HRT2WI的影像组学模型、年龄-影像组学模型。HRT2WI影像组学模型在训练集与验证集预测食管癌新辅助放化疗是否缓解的AUC、准确率、敏感度和特异度分别为0.863、80.0%、88.2%、76.7%,0.809、81.5%、75.0%、84.2%。年龄-影像组学模型在训练集与验证集预测食管癌新辅助放化疗是否缓解的AUC、准确率、敏感度和特异度分别为0.888、81.7%、94.1%、76.7%,0.836、81.5%、87.5%、78.9%。年龄-影像组学模型预测食管癌新辅助放化疗是否缓解的AUC、准确率和敏感度均高于HRT2WI影像组学模型,而两者的特异度相仿。结论：基于HRT2WI影像组学模型对食管癌新辅助放化疗是否缓解具有较好的预测效能,且HRT2WI影像组学联合年龄特征模型显示出更高的预测价值。     

CT影像组学模型预测CNLC Ⅰ～Ⅱ期肝细胞癌切除术后早期复发的研究

目的 建立术前CT影像组学预测模型对中国肝癌分期(CNLC)Ⅰ～Ⅱ期肝细胞癌(HCC)切除术后早期复发进行预测。方法 回顾性分析接受手术切除的CNLCⅠ～Ⅱ期HCC患者153例的资料。用3D slicer软件勾画肿瘤感兴趣区(ROI),用pyradiomics包提取影像组学特征。基于LASSO算法进行特征筛选、并建立影像组学标签(Rad-score)。采用单因素Logistic回归和多因素Logistic逐步回归法确立独立预测因子,构建影像组学预测模型和临床预测模型。用受试者工作特征曲线(ROC)曲线下面积(AUC)来比较模型的区分度,用校准曲线评估模型的校准度,用临床决策曲线分析(DCA)评估模型的临床应用价值。结果 Rad-score、瘤内供血动脉、肝功能白蛋白-胆红素分级(ALBI分级)、性别是独立预测因子。影像组学模型具有良好的预测效能(AUC:训练组0.900,验证组0.853),优于临床模型(AUC:训练组0.823,验证组0.741)。校准曲线显示影像组学模型具有良好的校准度。DCA显示阈值概率在0.1～1.0时,影像组学模型的净获益要高于临床模型。结论 基于CT影像组学...     

基于增强CT影像组学预测食管鳞癌淋巴血管侵犯状态的价值

目的:探讨基于增强CT影像组学预测食管鳞癌(ESCC)淋巴血管侵犯(LVI)的价值。方法:回顾性搜集行根治性切除术并经术后病理证实的224例食管鳞癌患者,其中包括66例LVI阳性和158例LVI阴性患者。所有患者均在术前2周内进行胸部增强CT扫描。将入组的患者按照7:3的比例随机分为训练集和测试集。使用3D Slicer软件逐层勾画全肿瘤感兴趣区(ROI),采用Python软件的Pyradiomics包提取肿瘤组织的影像组学特征,建立影像组学模型用于预测食管鳞癌的LVI状态并进行验证。采用受试者工作特征(ROC)曲线的曲线下面积(AUC)、敏感度、特异度、准确度、阳性预测值和阴性预测值来评价影像组学模型的诊断效能,使用校准曲线评价影像组学模型在训练集和测试集中的拟合程度。使用决策曲线分析(DCA)评价影像组学模型的临床应用价值。结果:从全肿瘤ROI中提取了1130个组学特征,经过筛选最终保留了7个影像组学特征,并使用多因素logistic回归建立影像组学预测模型。在训练集中,影像组学模型预测LVI的AUC值为0.930,敏感度为0.851,特异度为0.919,准确度为0.899,阳性预...     

基于MRI影像组学构建新辅助放化疗后局部进展期直肠癌的预测模型

目的 探讨MRI影像组学模型、临床模型和综合模型对局部进展期直肠癌(LARC)新辅助放化疗(n CRT)疗效的预测价值。资料与方法 回顾性收集2017年1月—2021年12月在河南中医药大学第一附属医院行n CRT后行根治性手术的140例LARC患者的临床病理和影像资料,其中病理完全缓解(p CR)108例、无病理完全缓解(np CR)32例,以7∶3随机分为训练组99例和验证组41例。所有患者治疗前均行直肠MRI检查,收集、提取并筛选患者的：(1)临床特征,包括年龄、性别、癌胚抗原、糖类抗原199、血管通透性参数(K^trans、K_ep、V_e)等;(2)MRI影像组学特征;构建临床模型、影像组学模型及影像组学标签与临床特征相结合的综合模型。采用受试者工作特征曲线下面积(AUC)评估临床、影像组学和综合模型的预测效能,采用决策曲线分析法评价3种模型的临床获益情况,并构建疗效预测的诺模图。结果 训练组中,pCR和np CR患者的K_ep差异有统计学意义(t=3.862,P<0.000 1);验证组中...     

基于临床影像特征和多参数MRI影像组学特征评估儿童弥漫中线胶质瘤H3K27M突变

【摘要】目的:探讨基于临床影像特征和多参数MRI影像组学特征评估儿童弥漫中线胶质瘤(DMG)H3K27M突变状态的应用价值。方法:回顾性纳入经病理诊断为DMG的98例患儿,包括74例H3K27M突变型和24例H3K27M野生型。按照大约7:3的比例分为训练集（n=68）和测试集（n=30）。基于T2WI和增强T1WI(cT1WI)序列提取影像组学特征。应用最大相关最小冗余(mRMR)和最小绝对收缩算子(LASSO)在训练集中筛选最优影像组学特征并计算影像组学评分(Rad-Score)。将临床影像特征和Rad-Score纳入多因素logistics回归筛选独立风险因素。联合筛选出的临床影像特征和Rad-Score构建联合模型以预测DMG的H3K27M状态,应用列线图对联合模型进行可视化。通过受试者工作特征(ROC)曲线评估模型的诊断效能,使用决策曲线(DCA)评估模型的临床应用价值。结果:基于T2WI和cT1WI序列共提取1648个影像组学特征,最终选取5个影像组学特征用于构建影像组学模型,该模型在训练集和测试集中均表现出良好的预测能力,曲线下面积(AUC)分别为0.844和0.758。多因素logistics回归显示环形强化和最小表观扩散系数(ADCmin)是H3K27M状态相关的临床影像特征风险因素（P均<0.05）,两者构建的临床影像模型具备一定的预测H3K27M状态的能力,训练集和测试集的AUC分别为0.802和0.720。由环形强化、ADCmin和Rad-Score构建的联合模型评估H3K27M状态表现出最佳的预测效能,训练集和测试集的AUC分别为0.863和0.851。结论:基于临床影像特征和多参数MRI影像组学特征构建的联合模型可用于无创性评估儿童DMG的H3K27M突变状态,具有较好的临床应用价值。     

基于常规MRI影像组学模型预测软组织肉瘤复发

【摘要】目的：探讨基于常规MRI的影像组学模型对预测软组织肉瘤（STS）复发的价值。方法：回顾性分析2012年1月－2021年6月在本院经手术病理证实的92例STS患者的临床和影像资料。术后每3个月进行一次影像学检查,随访时间至少12个月以上,根据随访结果有无复发或远处转移分为复发组（27例）,无复发组（65例）。采用完全随机方法将所有患者按7:3的比例分为训练集（n=65）和验证集（n=27）。使用ITK-SNAP软件,分别在T1WI和压脂T2WI上逐层沿肿瘤边缘手动勾画ROI并进行三维融合（VOI）,然后使用AK软件提取纹理特征,使用最小冗余最大相关（mRMR）和最小绝对值收敛和选择算子（LASSO）回归分析方法分别对T1WI序列、压脂T2WI序列和联合序列的纹理特征进行降维和筛选,并建立影像组学模型,根据各个组学特征的权重系数计算影像组学评分（Radscore）,运用100次留组交叉验证（LGOCV）方法来评估模型的可靠性。将临床病理、常规MRI特征与预测效能最高的影像组学模型的Radscore相结合,采用多因素logistic回归（LR）、随机森林（RF）和支持向量机（SVM）三种机器学习算法分别建立机器学习模型。采用受试者工作特征（ROC）曲线评价各模型的预测效能,应用决策曲线分析（DCA）评估模型的临床应用价值。结果：临床模型在训练集和验证集中预测STS复发的ROC曲线下面积（AUC）分别为0.71（95%CI:0.58~0.85）和0.74（95%CI:0.52~0.97）。基于T1WI、压脂T2WI和联合序列的影像组学模型在训练集中预测STS复发的AUC分别为0.81（95%CI:0.70~0.93）、0.92（95%CI:0.86~0.99）和0.91（95%CI:0.84~0.99）,在验证集中分别为0.84（95%CI:0.63~1.00）、0.92（95%CI:0.81~1.00）和0.86（95%CI:0.72~1.00）。采用机器学习算法构建的LR、RF和SVM模型在训练集中预测STS复发的AUC分别为0.93（95%CI:0.87~0.99）、0.91（95%CI:0.84~0.99）和0.77（95%CI:0.63~0.91）,在验证集中分别为0.93（95%CI:0.83~1.00）、0.86（95%CI:0.71~1.00）和0.83（95%CI:0.66~1.00）。DCA分析结果表明,压脂T2WI和联合序列的影像组学模型、以及LR和RF模型的临床受益均较好。结论：基于常规MRI序列中的压脂T2WI和联合序列构建的影像组学模型对预测STS复发具有较高的预测效能和较好的临床受益,基于不同机器学习算法构建的预测模型的预测效能并无明显提高。     

磁共振影像组学对直肠癌新辅助治疗后病理完全反应的评估价值

目的:探讨基于M R T2 WI的影像组学方法对直肠癌接受新辅助治疗(nCRT)后病理完全反应(pC R)状态的评估价值.方法:回顾性分析2019年1月-2020年12月在我院接受新辅助放化疗(nCRT)后行手术切除的99例局部进展期直肠癌(locally-advanced rectal cancer,LARC)患者的...     

CT影像组学联合细胞角蛋白19片段预测EGFR突变阳性非小细胞肺癌患者EGFR-TKIs的疗效

目的:探讨CT影像组学联合细胞角蛋白19片段在预测表皮生长因子受体(EGFR)突变阳性的非小细胞肺癌(NSCLC)患者行表皮生长因子受体酪氨酸激酶抑制剂(EGFR-TKIs)治疗的疗效。方法:回顾性搜集在本院确诊为EGFR突变阳性随即接受EGFR-TKIs治疗的194例NSCLC患者的病例资料。在EGFR-TKIs治疗3个月后行CT检查来判断疗效。根据实体肿瘤疗效评价标准1.1(RECIST 1.1),治疗有效121例,无效73例。采用完全随机方法将患者按7∶3的比例分为训练集和验证集。在训练集中提取NSCLC病灶的组学特征,然后使用主成分分析(PCA)、kruskal-wallis(KW)法及逻辑回归分析结合最小绝对值收敛和选择算子(LR-LASSO)对影像组学特征进行降维及影像组学模型的构建,获得每例患者的影像组学标签值。利用临床资料、病灶的CT形态学特征和病理结果建立临床模型,联合临床资料和影像组学标签建立联合模型。使用受试者工作特征曲线(ROC)下面积(AUC)评价各个模型对EGFR-TKIs疗效的预测效能,应用决策曲线分析(DCA)评估模型的临床应用价值。结果:训练集中经PC...     

Radiomics signature as a new biomarker for preoperative prediction of neoadjuvant chemoradiotherapy response in locally advanced rectal cancer

PURPOSEWhether radiomics methods are useful in prediction of therapeutic response to neoadjuvant chemoradiotherapy (nCRT) is unclear. This study aimed to investigate multiple magnetic resonance imaging (MRI) sequence-based radiomics methods in evaluating therapeutic response to nCRT in patients with locally advanced rectal cancer (LARC).METHODSThis retrospective study enrolled patients with LARC (06/2014–08/2017) and divided them into nCRT-sensitive and nCRT-resistant groups according to postoperative tumor regression grading results. Radiomics features from preoperative MRI were extracted, followed by dimension reduction using the minimum redundancy maximum relevance filter. Three machine-learning classifiers and an ensemble classifier were used for therapeutic response prediction. Radiomics nomogram incorporating clinical parameters were constructed using logistic regression. The receiver operating characteristic (ROC), decision curves analysis (DCA) and calibration curves were also plotted to evaluate the prediction performance.RESULTSThe machine learning classifiers showed good prediction performance for therapeutic responses in LARC patients (n=189). The ROC curve showed satisfying performance (area under the curve [AUC], 0.830; specificity, 0.794; sensitivity, 0.815) in the validation group. The radiomics signature included 30 imaging features derived from axial T1-weighted imaging with contrast and sagittal T2-weighted imaging and exhibited good predictive power for nCRT. A radiomics nomogram integrating carcinoembryonic antigen levels and tumor diameter showed excellent performance with an AUC of 0.949 (95% confidence interval, 0.892–0.997; specificity, 0.909; sensitivity, 0.879) in the validation group. DCA confirmed the clinical usefulness of the nomogram model.CONCLUSIONThe radiomics method using multiple MRI sequences can be used to achieve individualized prediction of nCRT in patients with LARC before treatment.

Colorectal cancer is one of the most common malignancies. It ranks fourth for morbidity and third for mortality among malignant tumors, among which the proportion of rectal cancer with poor prognosis is over 60% (1, 2). Neoadjuvant therapy, combined with total mesorectal excision, has become a common strategy for rectal cancer (3). Response to neoadjuvant chemoradiotherapy (nCRT) is a marker of good prognosis in patients with locally advanced rectal cancer (LARC) (4). Tumor regression grading (TRG) is a reliable biomarker for evaluating the efficacy of nCRT (5, 6). TRG reflects the treatment effect of nCRT by evaluating fibrosis and the ratio of residual tumor cells (4). The accurate nCRT evaluation can only be achieved by postoperative histopathological TRG (3, 4), and there is still no technology that can noninvasively evaluate the therapeutic response.Magnetic resonance imaging (MRI) is commonly used in the diagnosis, preoperative staging, and therapeutic efficacy evaluation of rectal cancer. Prediction of the efficacy of nCRT by MRI has been rarely reported, partly due to the heterogeneity of the tumor combined with the prevalence of fibrosis and edema of lesions and surrounding tissue after nCRT. Over the recent years, a magnetic resonance TRG system was proposed for the evaluation of nCRT efficacy by using MRI and evaluating residual tumor and fibrosis. Nevertheless, the magnetic resonance TRG method has a low predictive value for pathological TRG and poor consistency, which hinders its clinical applications (7, 8).In recent years, radiomics has drawn increasing attention in oncology. Radiomics features selected from medical images have shown to be highly associated with the diagnosis and prognosis of cancers, and even with gene expression patterns (9). Studies highlighted the value of radiomics approaches in determining tumor status, preoperative staging, and efficacy evaluation (9, 10). Nevertheless, the application of the radiomics methods in evaluating therapeutic responses to nCRT is limited (11).Accordingly, the aim of the present study was to establish an nCRT prediction model based on multiple MRI sequences combined with tumor anatomy and biological characteristics so as to achieve a comprehensive preliminary prediction of nCRT efficacy for rectal cancer before treatment, to provide an essential basis for the rational formulation of clinical diagnosis and treatment decisions, and to avoid unnecessary exposure to radiotherapy and chemotherapy and the related risks such as toxicity and delayed definitive surgery.     

Development and validation of a MRI-based radiomics signature for prediction of KRAS mutation in rectal cancer

To develop a T2-weighted (T2W) image-based radiomics signature for the individual prediction of KRAS mutation status in patients with rectal cancer. Three hundred four consecutive patients from center I with pathologically diagnosed rectal adenocarcinoma (training dataset, n = 213; internal validation dataset, n = 91) were enrolled in our retrospective study. The patients from center II (n = 86) were selected as an external validation dataset. A total of 960 imaging features were extracted from high-resolution T2W images for each patient. Five steps, mainly univariate statistical tests, were applied for feature selection. Subsequently, three classification methods, i.e., logistic regression (LR), decision tree (DT), and support vector machine (SVM) algorithm, were applied to develop the radiomics signature for KRAS prediction in the training dataset. The predictive performance was evaluated by receiver operating characteristics curve (ROC) analysis, calibration curve, and decision curve analysis (DCA). Seven radiomics features were screened as a KRAS-associated radiomics signature of rectal cancer. Our best prediction model was obtained with SVM classifiers with AUC of 0.722 (95%CI, 0.654–0.790) in the training dataset. This was validated in the internal and external validation datasets with good calibration, and the corresponding AUCs were 0.682 (95% CI, 0.569–0.794) and 0.714 (95% CI, 0.602–0.827), respectively. DCA confirmed its clinical usefulness. The proposed T2WI-based radiomics signature has a moderate performance to predict KRAS status, and may be useful for supplementing genomic analysis to determine KRAS expression in rectal cancer patients. • T2WI-based radiomics showed a moderate diagnostic significance for KRAS status. • The best prediction model was obtained with SVM classifier. • The baseline clinical and histopathological characteristics were not associated with KRAS mutation.     

Classification of early stage non-small cell lung cancers on computed tomographic images into histological types using radiomic features: interobserver delineation variability analysis

Radiomics, which involves the extraction of large numbers of quantitative features from medical images, has attracted attention in cancer research. In radiomics analysis, tumor segmentation is a crucial step. In this study, we evaluated the potential application of radiomics for predicting the histology of early stage non-small cell lung cancer (NSCLC) by analyzing interobserver variability in tumor delineation. Forty patient datasets were included in this study, 21 involving adenocarcinomas and 19 involving squamous cell carcinomas. All patients underwent stereotactic body radiotherapy treatment. In total, 476 features were extracted from each dataset, representing treatment planning, computed tomography images, and gross tumor volume (GTV). The definition of GTV can significantly affect the histology prediction. Therefore, in the present study, the effect of interobserver tumor delineation variability on radiomic features was evaluated by preparing 4 volumes of interest (VOIs) for each patient, as follows: the original GTV (which was delineated at treatment planning); two GTVs delineated retrospectively by radiation oncologists; and a semi-automatic GTV contoured by a medical physicist. Radiomic features extracted from each VOI were then analyzed using a naïve Bayesian model. Area-under-the-curve (AUC) analysis showed that interobserver variability in delineation is a significant factor in radiomics performance. Nevertheless, with 8 selected features, AUC values averaged over the VOIs were high (0.725 ± 0.070). The present study indicated that radiomics has potential for predicting early stage NSCLC histology despite variability in delineation. The high prediction accuracy implies that noninvasive histology evaluation by radiomics is a promising clinical application.     

Combined value of apparent diffusion coefficient-standardized uptake value max in evaluation of post-treated locally advanced rectal cancer

AIM: To assess the clinical diagnostic value of functional imaging, combining quantitative parameters of apparent diffusion coefficient (ADC) and standardized uptake value (SUV)max, before and after chemo-radiation therapy, in prediction of tumor response of patients with rectal cancer, related to tumor regression grade at histology.METHODS: A total of 31 patients with biopsy proven diagnosis of rectal carcinoma were enrolled in our study. All patients underwent a whole body ¹⁸FDG positron emission tomography (PET)/computed tomography (CT) scan and a pelvic magnetic resonance (MR) examination including diffusion weighted (DW) imaging for staging (PET1, RM1) and after completion (6.6 wk) of neoadjuvant treatment (PET2, RM2). Subsequently all patients underwent total mesorectal excision and the histological results were compared with imaging findings. The MR scanning, performed on 1.5 T magnet (Philips, Achieva), included T2-weighted multiplanar imaging and in addition DW images with b-value of 0 and 1000 mm²/s. On PET/CT the SUVmax of the rectal lesion were calculated in PET1 and PET2. The percentage decrease of SUVmax (ΔSUV) and ADC (ΔADC) values from baseline to presurgical scan were assessed and correlated with pathologic response classified as tumor regression grade (Mandard’s criteria; TRG1 = complete regression, TRG5 = no regression).RESULTS: After completion of therapy, all the patients were submitted to surgery. According to the Mandard’s criteria, 22 tumors showed complete (TRG1) or subtotal regression (TRG2) and were classified as responders; 9 tumors were classified as non responders (TRG3, 4 and 5). Considering all patients the mean values of SUVmax in PET 1 was higher than the mean value of SUVmax in PET 2 (P < 0.001), whereas the mean ADC values was lower in RM1 than RM2 (P < 0.001), with a ΔSUV and ΔADC respectively of 60.2% and 66.8%. The best predictors for TRG response were SUV2 (threshold of 4.4) and ADC2 (1.29 × 10^-3 mm²/s) with high sensitivity and specificity. Combining in a single analysis both the obtained median value, the positive predictive value, in predicting the different group category response in related to TRG system, presented R² of 0.95.CONCLUSION: The functional imaging combining ADC and SUVmax in a single analysis permits to detect changes in cellular tissue structures useful for the assessment of tumour response after the neoadjuvant therapy in rectal cancer, increasing the sensitivity in correct depiction of treatment response than either method alone.     

基于MRI影像组学模型对食管癌淋巴结转移的预测价值

目的 探讨基于T₂WI及增强T₁WI序列MRI影像组学特征构建模型预测食管癌淋巴结转移的价值。 方法 回顾性收集经病理证实并行多模态MRI检查的食管癌病人120例，男89例，女31例，平均年龄（63.4±8.2）岁。将病人按7:3比例随机分为训练集84例和验证集36例。以手术病理为金标准将病人分为淋巴结转移阴性组（56例）和阳性组（64例）。采用A.K.软件基于T₂WI和增强T₁WI获取肿瘤兴趣区体积（VOI），提取影像组学特征并进行降维筛选，并采用Logistic回归分析法构建基于T₂WI、增强T₁WI、联合T₂WI+增强T₁WI序列的影像组学模型。2组间一般临床资料比较采用独立样本t检验和χ²检验。采用组内相关系数（ICC）分析2名医师获取VOI的一致性。采用受试者操作特征（ROC）曲线评估预测模型的诊断效能，计算其曲线下面积（AUC），并采用DeLong法比较不同模型的AUC值。 结果 淋巴结转移阴性和阳性组间病人的性别、年龄，肿瘤位置、病理类型及肿瘤长度的差异均无统计学意义（均P>0.05）。2名医师在T₂WI和增强T₁WI影像上获取VOI的一致性均较好（均P>0.8）。经筛选后，基于T₂WI、增强T₁WI、T₂WI+增强T₁WI联合序列获得的影像组学特征分别有5、6、9个。在训练集及验证集中联合模型的AUC高于增强T₁WI和T₂WI模型，且增强T₁WI模型的AUC高于T₂WI模型（均P<0.05）。 结论 基于MRI影像组学特征构建的模型对食管癌病人术前淋巴结转移具有良好的预测效能，且T₂WI+增强T₁WI联合模型较单序列模型的预测价值更高。     

Radiomics Analysis of Susceptibility Weighted Imaging for Hepatocellular Carcinoma: Exploring the Correlation between Histopathology and Radiomics Features

Purpose:No previous researches have extracted radiomics features from susceptibility weighted imaging (SWI) for biomedical applications. This research aimed to explore the correlation between histopathology of hepatocellular carcinoma (HCC) and radiomics features extracted from SWI.Methods:A total of 53 patients were ultimately enrolled into this retrospective study with MR examinations undertaken at a 3T scanner. About 107 radiomics features were extracted from SWI images of each patient. Then, the Spearman correlation test was performed to evaluate the correlation between the SWI-derived radiomics features and histopathologic indexes including histopathologic grade, microvascular invasion (MVI) as well as the expression status of cytokeratin 7 (CK-7), cytokeratin 19 (CK-19) and Glypican-3 (GPC-3). With SWI-derived radiomics features utilized as independent variables, four logistic regression-based diagnostic models were established for diagnosing patients with positive CK-7, CK-19, GPC-3 and high histopathologic grade, respectively. Then, receiver operating characteristic analysis was performed to evaluate the diagnostic performance.Results:A total of 11, 32, 18 and one SWI-derived radiomics features were significantly correlated with histopathologic grade, the expression of CK-7, the expression of CK-19 and the expression of GPC-3 (P < 0.05), respectively. None of the SWI-derived radiomics features was correlated with MVI status. The areas under the curve were 0.905, 0.837, 0.800 and 0.760 for diagnosing patients with positive CK-19, positive CK-7, high histopathologic grade and positive GPC-3.Conclusion:Extracting the radiomics features from SWI images was feasible to evaluate multiple histopathologic indexes of HCC.     

Ultrasound elastography in patients with rectal cancer treated with chemoradiation

Objective

The current literature has described several predictive markers in rectal cancer patients treated with chemoradiation, but so far none of them have been validated for clinical use. The purpose of the present study was to compare quantitative elastography based on ultrasound measurements in the course of chemoradiation with tumor response based on T stage classification and the Mandard tumor regression grading (TRG).

Materials and methods

We prospectively examined 31 patients with rectal cancer planned for high dose radiochemotherapy. The tumor and the mesorectal fat elasticity were measured using the Acoustic Radiation Force Impulse to generate information on the mechanical properties of the tissue. The objective quantitative elastography shear wave velocity was compared to the T stage classification and TRG.

Results

The baseline mean tumor elasticity was 3.13 m/s. Two and six weeks after the start of chemoradiation the velocities were 2.17 m/s and 2.11 m/s, respectively. The difference between baseline velocity and velocities during the treatment course was statistically significant, (p < 0.0001). Patients with tumor confined to the rectal wall at histopathology (ypT1-2) had a mean elasticity measurement after two weeks of treatment of 1.95 m/s, whereas tumors invading the mesorectal fat (ypT3-4) had a velocity of 2.47 m/s, (p < 0.05). The mean elasticity tended to be lower (1.99 m/s) after two weeks in patients with TRG 1–2 responses in contrast to 2.24 m/s in those with TRG 3–4.

Conclusion

Ultrasound elastography after two weeks of chemoradiation seems to hold early predictive information to the pathological T stage.     

Preoperative prediction of tumour deposits in rectal cancer by an artificial neural network–based US radiomics model

To develop a machine learning–based ultrasound (US) radiomics model for predicting tumour deposits (TDs) preoperatively. From December 2015 to December 2017, 127 patients with rectal cancer were prospectively enrolled and divided into training and validation sets. Endorectal ultrasound (ERUS) and shear-wave elastography (SWE) examinations were conducted for each patient. A total of 4176 US radiomics features were extracted for each patient. After the reduction and selection of US radiomics features , a predictive model using an artificial neural network (ANN) was constructed in the training set. Furthermore, two models (one incorporating clinical information and one based on MRI radiomics) were developed. These models were validated by assessing their diagnostic performance and comparing the areas under the curve (AUCs) in the validation set. The training and validation sets included 29 (33.3%) and 11 (27.5%) patients with TDs, respectively. A US radiomics ANN model was constructed. The model for predicting TDs showed an accuracy of 75.0% in the validation cohort. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and AUC were 72.7%, 75.9%, 53.3%, 88.0% and 0.743, respectively. For the model incorporating clinical information, the AUC improved to 0.795. Although the AUC of the US radiomics model was improved compared with that of the MRI radiomics model (0.916 vs. 0.872) in the 90 patients with both ultrasound and MRI data (which included both the training and validation sets), the difference was nonsignificant (p = 0.384). US radiomics may be a potential model to accurately predict TDs before therapy. • We prospectively developed an artificial neural network model for predicting tumour deposits based on US radiomics that had an accuracy of 75.0%. • The area under the curve of the US radiomics model was improved than that of the MRI radiomics model (0.916 vs. 0.872), but the difference was not significant (p = 0.384). • The US radiomics–based model may potentially predict TDs accurately before therapy, but this model needs further validation with larger samples.     

影像组学分析和机器学习在肺结节良恶性分类中的应用

目的:探讨CT纹理特征对良恶性肺结节的鉴别价值及在独立数据集上的泛化能力.方法:回顾性分析LIDC-IDRI和LUNGx数据库中共1428个肺结节(直径3～30 mm)的CT图像,其中良性1221个、恶性207个.将LIDC-IDRI数据库的1372个结节(良性1190个,恶性182个)作为训练集,LUNGx数据库的5...