摘要
由一名新闻记者-机器人与机器学习日报的工作人员新闻编辑每日新闻-调查人员发布关于人工智能的新报告。根据新闻报道来自肯塔基州路易斯维尔的NewsRx记者的研究表明，“这项研究提出了一个创新的深度磁共振成像中肾脏分割的勒阿宁框架(MRI)DAT A。我们的新闻记者从Lo Uisville大学的研究中获得了一句话：框架使用残差循环一致性整合肾脏外观和先前形状信息生成对抗网络（CycleGA N）。建立了一种基于外观的形状先验模型，利用基于快速行进水平集方法的肾质心等圆c形状提取。利用肾脏质心匹配交叉等圆轮廓线外观，所提出的基于外观的形状先验模型对平移、旋转、和缩放，消除对对齐的需要。另外，一种新的加权损失函数，即H-损失，为了提高分割性能和防止过拟合。建议的方法是在34个血氧水平依赖的(BOLD)g筏上进行测试，这些筏来自我们肾移植项目的患者，平均骰子得分达到92%。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning DailyNews Daily News – Investigators publish new report on ar tificial intelligence. According to news reportingfrom Louisville, Kentucky, by NewsRx journalists, research stated, “This study presents an innovative deeple arning framework for kidney segmentation in magnetic resonance imaging (MRI) dat a.”Our news correspondents obtained a quote from the research from University of Lo uisville: “Theframework integrates both kidney appearance and prior shape infor mation using a residual cycle-consistentgenerative adversarial network (CycleGA N). An appearance-based shape prior model is developed, utilizingiso-circular c ontours generated from the kidney centroid and employing the fast marching level sets methodfor shape extraction. By utilizing the kidney centroid and matching cross-circular iso-circular contours’appearance, the proposed appearance-based shape prior model remains invariant to translation, rotation,and scaling, elim inating the need for alignment. Additionally, a novel weighted loss function, th e H-Loss,is introduced to enhance segmentation performance and prevent overfitt ing. The proposed approach istested on 34 blood-oxygen-level-dependent (BOLD) g rafts from patients in our kidney transplant program,achieving an average dice score of 92%.”
NETL
NSTL