In this paper, methods are proposed and validated to determine low and high thresholds to segment out gray matter and white matter for MR images of different pulse sequences of human brain. First, a two-dimensional re...In this paper, methods are proposed and validated to determine low and high thresholds to segment out gray matter and white matter for MR images of different pulse sequences of human brain. First, a two-dimensional reference image is determined to represent the intensity characteristics of the original three-dimensional data. Then a region of interest of the reference image is determined where brain tissues are present. The non-supervised fuzzy c-means clustering is employed to determine: the threshold for obtaining head mask, the low threshold for T2-weighted and PD-weighted images, and the high threshold for T1-weighted, SPGR and FLAIR images. Supervised range-constrained thresholding is employed to determine the low threshold for T1-weighted, SPGR and FLAIR images. Thresholding based on pairs of boundary pixels is proposed to determine the high threshold for T2-and PD-weighted images. Quantification against public data sets with various noise and inhomogeneity levels shows that the proposed methods can yield segmentation robust to noise and intensity inhomogeneity. Qualitatively the proposed methods work well with real clinical data.展开更多
Objective To qualitatively assess the diagnostic performance of dynamic contrast enhancement(DCE),diffusionweighted imaging(DWI),and T2-weighted imaging(T2WI),alone or in combination,in the evaluation of breast cancer...Objective To qualitatively assess the diagnostic performance of dynamic contrast enhancement(DCE),diffusionweighted imaging(DWI),and T2-weighted imaging(T2WI),alone or in combination,in the evaluation of breast cancer.Methods We retrospectively reviewed the records of 394 consecutive patients with pathologically confirmed breast lesions who had undergone 3-T magnetic resonance imaging(MRI).The morphological characteristics of breast lesions were evaluated using DCE,DWI,and T2WI based on BI-RADS lexicon descriptors by trained radiologists.Patients were categorized into mass and non-mass groups based on MRI characteristics of the lesions,and the differences between benign and malignant lesions in each group were compared.Clinical prediction models for breast cancer diagnosis were constructed using logistic regression analysis.Diagnostic efficacies were compared using the area under the receiver operating characteristic curve(AUC)and DeLong test.Results For mass-like lesions,all the morphological parameters significantly differentiated benign and malignant lesions on consensus DCE,DWI,and T2WI(P<0.05).The combined method(DCE+DWI+T2WI)had a higher AUC(0.865)than any of the individual modality(DCE:0.786;DWI:0.793;T2WI:0.809)(P<0.05).For non-mass-like lesions,DWI signal intensity was a significant predictor of malignancy(P=0.036),but the model using DWI alone had a low AUC(0.669).Conclusion Morphological assessment using the combination of DCE,DWI,and T2WI provides better diagnostic value in differentiating benign and malignant breast mass-like lesions than assessment with only one of the modalities.展开更多
Magnetic nanoparticles （Fe304） were prepared by chemical precipitation method using Fe^2＋ and Fe^3＋ salts with sodium hydroxide in the nitrogen atmosphere. Fe3O4 nanoparticles were coated with human serum albumin...Magnetic nanoparticles （Fe304） were prepared by chemical precipitation method using Fe^2＋ and Fe^3＋ salts with sodium hydroxide in the nitrogen atmosphere. Fe3O4 nanoparticles were coated with human serum albumin（HSA） for magnetic resonance imaging as contrast agent. Characteristics of magnetic particles coated or uncoated were carried out using scanning electron microscopy and X-ray diffraction. Zeta potentials, package effects and distributions of colloid particles were measured to confirm the attachment of HSA on magnetic particles. Effects of Fe3O4 nanoparticles coated with HSA on magnetic resonance imaging were investigated with rats. The experimental results show that the adsorption of HSA on magnetic particles is very favorable to dispersing of magnetic Fe3O4 particles, while the sizes of Fe3O4 particles coated are related to the molar ratio of Fe3O4 to HSA. The diameters of the majority of particles coated are less than 100 nm. Fe3O4 nanoparticle coated with HSA has a good biocompatibility and low toxicity. This new contrast agent has some effects on the nuclear magnetic resonance imaging of liver and the lowest dosage is 20μmol/kg for the demands of diagnosis.展开更多
目的本研究旨在构建一个基于临床和影像学特征的极端梯度提升(extreme gradient boosting,XGBoost)模型,以鉴别乳腺非肿块病变的良恶性。材料与方法收集2018年1月至2024年7月2个机构,2种乳腺X线设备检查的有病理结果的首诊乳腺非肿块病...目的本研究旨在构建一个基于临床和影像学特征的极端梯度提升(extreme gradient boosting,XGBoost)模型,以鉴别乳腺非肿块病变的良恶性。材料与方法收集2018年1月至2024年7月2个机构,2种乳腺X线设备检查的有病理结果的首诊乳腺非肿块病变480个。患者被分为建模组[n=310,数字乳腺X线摄影(digital mammography,DM)检查]、内部验证组(n=108,DM检查),和外部验证组[n=62,数字乳腺体层合成摄影(digital breast tomosynthesis,DBT)检查]。记录患者术前乳腺X线(DM或DBT),MRI以及临床特征。采用XGBoost算法和多因素逻辑回归分析,分别构建XGBoost模型和逻辑回归(logistic regression,LR)模型。使用受试者工作特征(receiver operating characteristic,ROC)曲线评估模型的诊断效能。结果在建模组中,患者以7∶3随机分为训练集(n=217)和测试集(n=93)。训练集、测试集、训练集的内部验证组及训练集的外部验证组中,恶性非肿块病灶分别为159(73%)、58(62%)、73(68%)和43(69%)。XGBoost模型的诊断效能明显优于LR模型,在独立的训练集、测试集、训练集的内部验证组及训练集的外部验证组中均表现出良好的诊断效能,曲线下面积(area under the curve,AUC)在0.884~0.913之间。XGBoost模型在四个队列中也表现出良好的校准能力和临床净获益。结论XGBoost模型能够准确鉴别乳腺非肿块病变的良恶性,具有推广应用的潜力。展开更多
目的观察基于相位对比(PC)MRI颅内血流动力学参数预测急性高原反应(AMS)的价值。方法前瞻性招募72名健康青年志愿者,于平原地区采集平静呼吸及轻、中及重度瓦尔萨尔瓦动作(VM)下的颈内动脉(ICA)及颈内静脉(IJV)PC MRI并记录ICA及IJV血...目的观察基于相位对比(PC)MRI颅内血流动力学参数预测急性高原反应(AMS)的价值。方法前瞻性招募72名健康青年志愿者,于平原地区采集平静呼吸及轻、中及重度瓦尔萨尔瓦动作(VM)下的颈内动脉(ICA)及颈内静脉(IJV)PC MRI并记录ICA及IJV血流动力学参数;根据急进海拔4411 m的高原地区10 h后路易斯湖评分(LLS)结果划分AMS组(n=9)与无AMS组(n=63);采用单因素及多因素logistic回归分析筛选各状态下AMS的独立预测因素,构建单一及联合VM状态预测模型;绘制受试者工作特征曲线,计算曲线下面积(AUC),评估各模型预测效能。结果轻度VM下ICA搏动指数(PI ICA)、中度VM下IJV面积(S IJV)及重度VM下IJV阻力指数(RI IJV)均为AMS独立预测因素(P均<0.05)。联合VM状态模型(AUC=0.869)预测AMS的效能高于单一VM状态模型(AUC=0.698~0.738)。结论基于轻度VM PI ICA、中度VM S IJV及重度VM RI IJV构建的模型可有效预测AMS。展开更多
文摘In this paper, methods are proposed and validated to determine low and high thresholds to segment out gray matter and white matter for MR images of different pulse sequences of human brain. First, a two-dimensional reference image is determined to represent the intensity characteristics of the original three-dimensional data. Then a region of interest of the reference image is determined where brain tissues are present. The non-supervised fuzzy c-means clustering is employed to determine: the threshold for obtaining head mask, the low threshold for T2-weighted and PD-weighted images, and the high threshold for T1-weighted, SPGR and FLAIR images. Supervised range-constrained thresholding is employed to determine the low threshold for T1-weighted, SPGR and FLAIR images. Thresholding based on pairs of boundary pixels is proposed to determine the high threshold for T2-and PD-weighted images. Quantification against public data sets with various noise and inhomogeneity levels shows that the proposed methods can yield segmentation robust to noise and intensity inhomogeneity. Qualitatively the proposed methods work well with real clinical data.
文摘Objective To qualitatively assess the diagnostic performance of dynamic contrast enhancement(DCE),diffusionweighted imaging(DWI),and T2-weighted imaging(T2WI),alone or in combination,in the evaluation of breast cancer.Methods We retrospectively reviewed the records of 394 consecutive patients with pathologically confirmed breast lesions who had undergone 3-T magnetic resonance imaging(MRI).The morphological characteristics of breast lesions were evaluated using DCE,DWI,and T2WI based on BI-RADS lexicon descriptors by trained radiologists.Patients were categorized into mass and non-mass groups based on MRI characteristics of the lesions,and the differences between benign and malignant lesions in each group were compared.Clinical prediction models for breast cancer diagnosis were constructed using logistic regression analysis.Diagnostic efficacies were compared using the area under the receiver operating characteristic curve(AUC)and DeLong test.Results For mass-like lesions,all the morphological parameters significantly differentiated benign and malignant lesions on consensus DCE,DWI,and T2WI(P<0.05).The combined method(DCE+DWI+T2WI)had a higher AUC(0.865)than any of the individual modality(DCE:0.786;DWI:0.793;T2WI:0.809)(P<0.05).For non-mass-like lesions,DWI signal intensity was a significant predictor of malignancy(P=0.036),but the model using DWI alone had a low AUC(0.669).Conclusion Morphological assessment using the combination of DCE,DWI,and T2WI provides better diagnostic value in differentiating benign and malignant breast mass-like lesions than assessment with only one of the modalities.
文摘Magnetic nanoparticles （Fe304） were prepared by chemical precipitation method using Fe^2＋ and Fe^3＋ salts with sodium hydroxide in the nitrogen atmosphere. Fe3O4 nanoparticles were coated with human serum albumin（HSA） for magnetic resonance imaging as contrast agent. Characteristics of magnetic particles coated or uncoated were carried out using scanning electron microscopy and X-ray diffraction. Zeta potentials, package effects and distributions of colloid particles were measured to confirm the attachment of HSA on magnetic particles. Effects of Fe3O4 nanoparticles coated with HSA on magnetic resonance imaging were investigated with rats. The experimental results show that the adsorption of HSA on magnetic particles is very favorable to dispersing of magnetic Fe3O4 particles, while the sizes of Fe3O4 particles coated are related to the molar ratio of Fe3O4 to HSA. The diameters of the majority of particles coated are less than 100 nm. Fe3O4 nanoparticle coated with HSA has a good biocompatibility and low toxicity. This new contrast agent has some effects on the nuclear magnetic resonance imaging of liver and the lowest dosage is 20μmol/kg for the demands of diagnosis.
文摘目的本研究旨在构建一个基于临床和影像学特征的极端梯度提升(extreme gradient boosting,XGBoost)模型,以鉴别乳腺非肿块病变的良恶性。材料与方法收集2018年1月至2024年7月2个机构,2种乳腺X线设备检查的有病理结果的首诊乳腺非肿块病变480个。患者被分为建模组[n=310,数字乳腺X线摄影(digital mammography,DM)检查]、内部验证组(n=108,DM检查),和外部验证组[n=62,数字乳腺体层合成摄影(digital breast tomosynthesis,DBT)检查]。记录患者术前乳腺X线(DM或DBT),MRI以及临床特征。采用XGBoost算法和多因素逻辑回归分析,分别构建XGBoost模型和逻辑回归(logistic regression,LR)模型。使用受试者工作特征(receiver operating characteristic,ROC)曲线评估模型的诊断效能。结果在建模组中,患者以7∶3随机分为训练集(n=217)和测试集(n=93)。训练集、测试集、训练集的内部验证组及训练集的外部验证组中,恶性非肿块病灶分别为159(73%)、58(62%)、73(68%)和43(69%)。XGBoost模型的诊断效能明显优于LR模型,在独立的训练集、测试集、训练集的内部验证组及训练集的外部验证组中均表现出良好的诊断效能,曲线下面积(area under the curve,AUC)在0.884~0.913之间。XGBoost模型在四个队列中也表现出良好的校准能力和临床净获益。结论XGBoost模型能够准确鉴别乳腺非肿块病变的良恶性,具有推广应用的潜力。
文摘目的观察基于相位对比(PC)MRI颅内血流动力学参数预测急性高原反应(AMS)的价值。方法前瞻性招募72名健康青年志愿者,于平原地区采集平静呼吸及轻、中及重度瓦尔萨尔瓦动作(VM)下的颈内动脉(ICA)及颈内静脉(IJV)PC MRI并记录ICA及IJV血流动力学参数;根据急进海拔4411 m的高原地区10 h后路易斯湖评分(LLS)结果划分AMS组(n=9)与无AMS组(n=63);采用单因素及多因素logistic回归分析筛选各状态下AMS的独立预测因素,构建单一及联合VM状态预测模型;绘制受试者工作特征曲线,计算曲线下面积(AUC),评估各模型预测效能。结果轻度VM下ICA搏动指数(PI ICA)、中度VM下IJV面积(S IJV)及重度VM下IJV阻力指数(RI IJV)均为AMS独立预测因素(P均<0.05)。联合VM状态模型(AUC=0.869)预测AMS的效能高于单一VM状态模型(AUC=0.698~0.738)。结论基于轻度VM PI ICA、中度VM S IJV及重度VM RI IJV构建的模型可有效预测AMS。
