The Wide Field Survey Telescope(WFST)is located at 4200 m on Saishiteng Mountain in Lenghu,Qinghai Province,China.It features a primary mirror with a diameter of 2.5 m and a camera equipped with nine CCDs,providing a ...The Wide Field Survey Telescope(WFST)is located at 4200 m on Saishiteng Mountain in Lenghu,Qinghai Province,China.It features a primary mirror with a diameter of 2.5 m and a camera equipped with nine CCDs,providing a wide field of view of approximately 3×3 square degrees.Calibration parameters are essential to ensure the precision of astrometric observations with the WFST.These parameters are derived from geometric distortion(GD)and gaps through astrometric modeling and are subsequently validated via the Yao’An High Precision Telescope(YAHPT).The GD solutions show maximum distortions between 1.18 and 10.29 pixels for the WFST chips,with central chips exhibiting lower distortion.After applying the GD correction,the precision of the WFST reaches 4 mas.The interchip gaps of the WFST range from 1.922 mm to 7.765 mm,corresponding to 10μm/pixel,aligning with the design and measurements.The calibrated parameters guarantee that the WFST can perform highly accurate astrometric measurements.Furthermore,as the WFST undergoes updates,the parameter model remains consistently applicable.展开更多
背景与目的肺癌是全球最常见的恶性肿瘤之一,也是癌症相关死亡的主要原因。早期肺癌通常表现为肺结节,准确评估其恶性风险对于延长生存期及避免过度诊疗至关重要。本研究旨在基于人工智能(artificial intelligence,AI)自动提取的影像学...背景与目的肺癌是全球最常见的恶性肿瘤之一,也是癌症相关死亡的主要原因。早期肺癌通常表现为肺结节,准确评估其恶性风险对于延长生存期及避免过度诊疗至关重要。本研究旨在基于人工智能(artificial intelligence,AI)自动提取的影像学特征参数构建模型,评估其在部分实性结节(part-solid nodule,PSN)恶性预测中的效能。方法回顾性分析2020年10月至2025年2月于兰州大学第二医院接受肺结节切除术的222例患者的229个PSN资料。根据病理结果,将45个良性病变及腺体前驱病变归为非恶性组,184个肺部恶性肿瘤归为恶性组。所有患者均接受胸部计算机断层扫描,使用AI软件提取影像学特征参数。通过单因素分析筛选显著变量,计算方差膨胀因子并剔除共线性较高的变量,LASSO回归进一步筛选关键特征,多因素逻辑回归确定独立危险因素。基于筛选结果,构建逻辑回归、随机森林、XGBoost、LightGBM、支持向量机5种模型,使用受试者工作特征(reciever operating characteristic,ROC)曲线评估模型性能。结果PSN良恶性的独立危险因素包括粗糙度(ngtdm)、依赖方差(gldm)和短运行低灰度重点(glrlm)。逻辑回归在训练集和测试集的曲线下面积(area under the curve,AUC)分别为0.86和0.89,表现较好。XGBoost的AUC分别为0.78和0.77,表现相对均衡,但准确度较低。支持向量机在训练集的AUC为0.93,测试集AUC降至0.80,表明该模型存在一定的过拟合。LightGBM在训练集表现优异,AUC为0.94,但在测试集上有所下降,AUC为0.88。随机森林模型在训练集和测试集上均表现稳定,训练集AUC为0.89,测试集AUC为0.91,具有较高的稳定性和良好的泛化能力。结论基于独立危险因素构建的随机森林模型在PSN良恶性预测中表现最佳,可以为临床医生提供有效的辅助预测,支持个体化治疗决策。展开更多
The internal and external flow fields during vented explosions of methane were characterized through numerical simulation,and the capability of numerical simulation thereof was validated by previous experimental data ...The internal and external flow fields during vented explosions of methane were characterized through numerical simulation,and the capability of numerical simulation thereof was validated by previous experimental data at three ignition positions.The venting mechanism was revealed by the simulated concentration distribution,temperature profile,and airflow velocity.The results show rear ignition results in the external methane mass distribution taking the form of"mushroom"and columnar flames in the external space,which can be expressed as a third-order polynomial relationship with distance;central ignition forms a relationship of the form y=AxB.Front ignition causes the temperature to show a tendency to repeated oscillations(rising,falling,and rising).Central ignition generates the maximum vented airflow velocity(V_(max)=320 m/s)upon vent opening.The results indicate that it is acceptable to apply numerical simulation of methane explosions in practice.展开更多
背景与目的肺癌是国内外致死率最高的恶性肿瘤,肺结节的精确检测是降低肺癌死亡率的关键。人工智能辅助诊断系统在肺结节检测、良恶性鉴别和浸润亚型诊断等领域发展迅速,对其效能进行验证是促进其应用于临床的前提。本研究旨在评估人工...背景与目的肺癌是国内外致死率最高的恶性肿瘤,肺结节的精确检测是降低肺癌死亡率的关键。人工智能辅助诊断系统在肺结节检测、良恶性鉴别和浸润亚型诊断等领域发展迅速,对其效能进行验证是促进其应用于临床的前提。本研究旨在评估人工智能辅助诊断系统预测肺结节早期肺腺癌浸润亚型的效能。方法回顾性分析2016年1月1日-2021年12月31日期间兰州大学第二医院收治的223例肺结节早期肺腺癌患者的临床资料,将早期肺腺癌分为浸润性腺癌组(n=170)和非浸润性腺癌组(n=53),其中非浸润性腺癌组又分为微浸润性腺癌组(n=31)和浸润前病变组(n=22)。比较各组的恶性概率和影像特征等信息,分析其对早期肺腺癌浸润亚型的预测能力,并对人工智能辅助诊断早期肺腺癌浸润亚型定性诊断的结果与术后病理进行一致性分析。结果早期肺腺癌不同浸润亚型肺结节的平均CT值(P<0.001)、直径(P<0.001)、体积(P<0.001)、恶性概率(P<0.001)、胸膜凹陷征(P<0.001)、分叶征(P<0.001)、毛刺征(P<0.001)差异均有统计学意义;随着早期肺腺癌不同浸润亚型浸润性增加,各组参数显性征象比例也逐渐升高;在二分类问题上,人工智能辅助诊断系统定性诊断早期肺腺癌浸润亚型的敏感性、特异性及曲线下面积(area under the curve,AUC)分别为81.76%、92.45%和0.871;在三分类问题上,人工智能辅助诊断系统定性诊断早期肺腺癌浸润亚型的准确率、召回率、F1分数及AUC分别为83.86%、85.03%、76.46%和0.879。结论该人工智能辅助诊断系统对肺结节早期肺腺癌浸润亚型具有一定的预测价值,随着算法的优化和数据的完善或可为患者个体化治疗提供指导。展开更多
基金supported by the Strategic Priority Research Program of the Chinese Academy of Sciences(XDA0350300)the National Natural Science Foundation of China(12203105,12103091,62394351,12073008)the China Manned Space Project(CMS-CSST-2021-A12,CMS-CSST-2021-B10).
文摘The Wide Field Survey Telescope(WFST)is located at 4200 m on Saishiteng Mountain in Lenghu,Qinghai Province,China.It features a primary mirror with a diameter of 2.5 m and a camera equipped with nine CCDs,providing a wide field of view of approximately 3×3 square degrees.Calibration parameters are essential to ensure the precision of astrometric observations with the WFST.These parameters are derived from geometric distortion(GD)and gaps through astrometric modeling and are subsequently validated via the Yao’An High Precision Telescope(YAHPT).The GD solutions show maximum distortions between 1.18 and 10.29 pixels for the WFST chips,with central chips exhibiting lower distortion.After applying the GD correction,the precision of the WFST reaches 4 mas.The interchip gaps of the WFST range from 1.922 mm to 7.765 mm,corresponding to 10μm/pixel,aligning with the design and measurements.The calibrated parameters guarantee that the WFST can perform highly accurate astrometric measurements.Furthermore,as the WFST undergoes updates,the parameter model remains consistently applicable.
文摘背景与目的肺癌是全球最常见的恶性肿瘤之一,也是癌症相关死亡的主要原因。早期肺癌通常表现为肺结节,准确评估其恶性风险对于延长生存期及避免过度诊疗至关重要。本研究旨在基于人工智能(artificial intelligence,AI)自动提取的影像学特征参数构建模型,评估其在部分实性结节(part-solid nodule,PSN)恶性预测中的效能。方法回顾性分析2020年10月至2025年2月于兰州大学第二医院接受肺结节切除术的222例患者的229个PSN资料。根据病理结果,将45个良性病变及腺体前驱病变归为非恶性组,184个肺部恶性肿瘤归为恶性组。所有患者均接受胸部计算机断层扫描,使用AI软件提取影像学特征参数。通过单因素分析筛选显著变量,计算方差膨胀因子并剔除共线性较高的变量,LASSO回归进一步筛选关键特征,多因素逻辑回归确定独立危险因素。基于筛选结果,构建逻辑回归、随机森林、XGBoost、LightGBM、支持向量机5种模型,使用受试者工作特征(reciever operating characteristic,ROC)曲线评估模型性能。结果PSN良恶性的独立危险因素包括粗糙度(ngtdm)、依赖方差(gldm)和短运行低灰度重点(glrlm)。逻辑回归在训练集和测试集的曲线下面积(area under the curve,AUC)分别为0.86和0.89,表现较好。XGBoost的AUC分别为0.78和0.77,表现相对均衡,但准确度较低。支持向量机在训练集的AUC为0.93,测试集AUC降至0.80,表明该模型存在一定的过拟合。LightGBM在训练集表现优异,AUC为0.94,但在测试集上有所下降,AUC为0.88。随机森林模型在训练集和测试集上均表现稳定,训练集AUC为0.89,测试集AUC为0.91,具有较高的稳定性和良好的泛化能力。结论基于独立危险因素构建的随机森林模型在PSN良恶性预测中表现最佳,可以为临床医生提供有效的辅助预测,支持个体化治疗决策。
基金supported by the Young Scientists Fund of National Natural Science Foundation of China(Grant Nos.12202202 and 12202494)the National Key Research and Development Program of China(Grant No.2021YFC3100700)。
文摘The internal and external flow fields during vented explosions of methane were characterized through numerical simulation,and the capability of numerical simulation thereof was validated by previous experimental data at three ignition positions.The venting mechanism was revealed by the simulated concentration distribution,temperature profile,and airflow velocity.The results show rear ignition results in the external methane mass distribution taking the form of"mushroom"and columnar flames in the external space,which can be expressed as a third-order polynomial relationship with distance;central ignition forms a relationship of the form y=AxB.Front ignition causes the temperature to show a tendency to repeated oscillations(rising,falling,and rising).Central ignition generates the maximum vented airflow velocity(V_(max)=320 m/s)upon vent opening.The results indicate that it is acceptable to apply numerical simulation of methane explosions in practice.
文摘背景与目的肺癌是国内外致死率最高的恶性肿瘤,肺结节的精确检测是降低肺癌死亡率的关键。人工智能辅助诊断系统在肺结节检测、良恶性鉴别和浸润亚型诊断等领域发展迅速,对其效能进行验证是促进其应用于临床的前提。本研究旨在评估人工智能辅助诊断系统预测肺结节早期肺腺癌浸润亚型的效能。方法回顾性分析2016年1月1日-2021年12月31日期间兰州大学第二医院收治的223例肺结节早期肺腺癌患者的临床资料,将早期肺腺癌分为浸润性腺癌组(n=170)和非浸润性腺癌组(n=53),其中非浸润性腺癌组又分为微浸润性腺癌组(n=31)和浸润前病变组(n=22)。比较各组的恶性概率和影像特征等信息,分析其对早期肺腺癌浸润亚型的预测能力,并对人工智能辅助诊断早期肺腺癌浸润亚型定性诊断的结果与术后病理进行一致性分析。结果早期肺腺癌不同浸润亚型肺结节的平均CT值(P<0.001)、直径(P<0.001)、体积(P<0.001)、恶性概率(P<0.001)、胸膜凹陷征(P<0.001)、分叶征(P<0.001)、毛刺征(P<0.001)差异均有统计学意义;随着早期肺腺癌不同浸润亚型浸润性增加,各组参数显性征象比例也逐渐升高;在二分类问题上,人工智能辅助诊断系统定性诊断早期肺腺癌浸润亚型的敏感性、特异性及曲线下面积(area under the curve,AUC)分别为81.76%、92.45%和0.871;在三分类问题上,人工智能辅助诊断系统定性诊断早期肺腺癌浸润亚型的准确率、召回率、F1分数及AUC分别为83.86%、85.03%、76.46%和0.879。结论该人工智能辅助诊断系统对肺结节早期肺腺癌浸润亚型具有一定的预测价值,随着算法的优化和数据的完善或可为患者个体化治疗提供指导。
