期刊文献+
共找到2篇文章
< 1 >
每页显示 20 50 100
Optically powered gas monitoring system using single-mode fibre for underground coal mines 被引量:1
1
作者 yingge chen Leonardo Silvestri +1 位作者 Xinyue Lei François Ladouceur 《International Journal of Coal Science & Technology》 EI CAS CSCD 2022年第2期36-47,共12页
We present an optically powered,intrinsically safe gas monitoring system to measure four essential environmental gases(CH_(4),CO_(2),CO and O_(2)),together with ambient temperature and pressure,for underground mines.T... We present an optically powered,intrinsically safe gas monitoring system to measure four essential environmental gases(CH_(4),CO_(2),CO and O_(2)),together with ambient temperature and pressure,for underground mines.The system is based on three key technologies developed at UNSW:(1)power-over-fbre(PoF)at 1550 nm using a single industry-standard,low-cost single-mode fbre(SMF)for both power delivery and information transmission,(2)liquid–crystal-based optical transducers for optical telemetry,and(3)ultra-low power consumption design of all electronics.Together,this approach allows each gas monitoring station to operate with less than 150 mW of optical power,meeting the intrinsic safety requirements specifed by the IEC60079-28 standard.A 2-month feld trial at BMA’s Broadmeadow underground mine proved the cabling compatibility to the mine’s existing optical network and the stability of the system performance.Compared with conventional electrically powered gas sensors,this technology bypasses the usual roadblocks of underground gas monitoring where electrical power is either unsafe or unavailable.Furthermore,using one fbre for both power delivery and communication enables longer distance coverage,reduces optical cabling and increases multiplexing possibilities and data throughput for better awareness of underground environment. 展开更多
关键词 Power-over-fbre Ultra-low power consumption Gas monitoring Intrinsic safety Single-mode fbre Optical transducer
下载PDF
超声影像组学标签对乳腺癌腋窝淋巴结转移的预测价值
2
作者 王瑛 陈英格 +4 位作者 叶素敏 陈东 刘宇 刘再毅 刘敏 《中华医学超声杂志(电子版)》 CSCD 北大核心 2022年第8期774-778,共5页
目的探讨基于常规超声的影像组学特征预测乳腺癌腋窝淋巴结转移的应用价值。方法回顾性收集2020年1月至2020年10月于中山大学肿瘤防治中心就诊经手术病理确诊的265例乳腺癌患者的临床资料和术前超声图像,按超声检查时间顺序,将患者分为... 目的探讨基于常规超声的影像组学特征预测乳腺癌腋窝淋巴结转移的应用价值。方法回顾性收集2020年1月至2020年10月于中山大学肿瘤防治中心就诊经手术病理确诊的265例乳腺癌患者的临床资料和术前超声图像,按超声检查时间顺序,将患者分为训练集(159例)和验证集(106例)。应用ImageJ软件手动勾画病灶区域,使用Pyradiomics从每个病灶区域中提取影像组学特征,采用多种方法逐步筛选特征,应用Logistic回归构建预测乳腺癌腋窝淋巴结转移的超声影像组学标签。在训练集和验证集上采用ROC曲线、校准曲线和决策曲线评估超声影像组学标签预测乳腺癌腋窝淋巴结转移的效能。结果最终筛选出8个关键超声影像组学特征用于构建超声影像组学标签。该标签在训练集和验证集中预测乳腺癌腋窝淋巴结转移的ROC曲线下面积分别为0.805(95%CI:0.734~0.876)、0.793(95%CI:0.706~0.880)。在校准曲线中,该标签在训练集和验证集均表现出较好的校准度(P=0.592、0.593),决策曲线分析进一步表明了该标签具有一定的临床实用性。结论基于超声的影像组学标签在预测乳腺癌腋窝淋巴结转移方面具有一定价值,可为治疗前乳腺癌的准确分期以及治疗方案的合理选择提供参考依据。 展开更多
关键词 超声检查 影像组学 乳腺肿瘤 腋窝淋巴结 预测模型 人工智能
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部