摘要
由一名新闻记者兼机器人与机器学习每日新闻编辑每日新闻-关于人工智能ce的详细数据已经呈现。根据NewsRx编辑从荷兰阿姆斯特丹发回的新闻报道，研究表明：“胸膜斑(PPs)是长期暴露石棉的形态学改变。PP和LUN G功能之间的关系尚不清楚，而PP去除体积的耗时性质阻碍了研究。”我们的新闻记者从荷兰癌症研究所的研究中获得了一句话，"为了使描绘的艰苦任务自动化，我们旨在开发一个自动人工智能（AI）驱动的PP分割。为探讨胸膜斑块体积(PPV)与肺功能的关系,放射科医生在石棉职业暴露患者(2014年11月至2019年11月)的Computed Tomography(CT)图像中手工绘制PPs。我们训练了一个无新UNET结构的AI模型。Dice相似系数量化了AI和放射科医生之间的重叠。使用Spearman相关系数(R)。为探讨肺活量与肺功能测试指标的相关性,记录肺活量(VC)、用力肺活量(FVC)和一氧化碳弥散量(DLCO)。我们对AI系统进行了5次的422次CT扫描,每次以不同的倍数(n=84~85)作为测试集,在这些独立的测试集上进行了对比。我们发现VC(n=80，r=-0.40)和FVC(n=82，r=-0.38)与PPV之间存在中等相关性，而DLCO(n=84，r=-0.09)之间没有相关性。我们观察到PV较高患者的VC（P=0.001）和FVC（P=0.04）值在统计学上明显较低，但DLCO没有（P=0.19）。我们成功开发了一种AI算法，自动分割CT图像中的PP，以实现快速的体积提取。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – Data detailed on Artificial Intelligen ce have been presented. According to news reporting out of Amsterdam, Netherland s, by NewsRx editors, research stated, “Pleural plaques (PPs) are morphologic ma nifestations of long-term asbestos exposure. The relationship between PP and lun g function is not well understood, whereas the time-consuming nature of PP delin eation to obtain volume impedes research.” Our news journalists obtained a quote from the research from Netherlands Cancer Institute, “To automate the laborious task of delineation, we aimed to develop a utomatic artificial intelligence (AI)-driven segmentation of PP. Moreover, we ai med to explore the relationship between pleural plaque volume (PPV) and pulmonar y function tests. Radiologists manually delineated PPs retrospectively in comput ed tomography (CT) images of patients with occupational exposure to asbestos (Ma y 2014 to November 2019). We trained an AI model with a no-new-UNet architecture . The Dice Similarity Coefficient quantified the overlap between AI and radiolog ists. The Spearman correlation coefficient (r) was used for the correlation betw een PPV and pulmonary function test metrics. When recorded, these were vital cap acity (VC), forced vital capacity (FVC), and diffusing capacity for carbon monox ide (DLCO). We trained the AI system on 422 CT scans in 5 folds, each time with a different fold (n = 84 to 85) as a test set. On these independent test sets co mbined, the correlation between the predicted volumes and the ground truth was r = 0.90, and the median overlap was 0.71 Dice Similarity Coefficient. We found w eak to moderate correlations with PPV for VC (n = 80, r = -0.40) and FVC (n = 82 , r = -0.38), but no correlation for DLCO (n = 84, r = -0.09). When the cohort w as split on the median PPV, we observed statistically significantly lower VC (P = 0.001) and FVC (P = 0.04) values for the higher PPV patients, but not for DLCO (P = 0.19). We successfully developed an AI algorithm to automatically segment PP in CT images to enable fast volume extraction.”
NETL
NSTL