Method and System For Regression-Based Object Detection in Medical Images
    1.
    发明申请
    Method and System For Regression-Based Object Detection in Medical Images 有权
    医学图像中基于回归的物体检测方法与系统

    公开(公告)号:US20080085044A1

    公开(公告)日:2008-04-10

    申请号:US11866572

    申请日:2007-10-03

    IPC分类号: G06K9/00

    摘要: A method and system for regression-based object detection in medical images is disclosed. A regression function for predicting a location of an object in a medical image based on an image patch is trained using image-based boosting ridge regression (IBRR). The trained regression function is used to determine a difference vector based on an image patch of a medical image. The difference vector represents the difference between the location of the image patch and the location of a target object. The location of the target object in the medical image is predicted based on the difference vector determined by the regression function.

    摘要翻译: 公开了一种用于医学图像中基于回归的物体检测的方法和系统。 使用基于图像的增强脊回归(IBRR)训练用于基于图像块预测医学图像中的对象的位置的回归函数。 训练回归函数用于基于医学图像的图像块来确定差分矢量。 差分向量表示图像块的位置与目标对象的位置之间的差异。 基于由回归函数确定的差分向量来预测目标对象在医学图像中的位置。

    Method and system for regression-based object detection in medical images
    2.
    发明授权
    Method and system for regression-based object detection in medical images 有权
    医学图像中基于回归的物体检测方法与系统

    公开(公告)号:US07949173B2

    公开(公告)日:2011-05-24

    申请号:US11866572

    申请日:2007-10-03

    IPC分类号: G06K9/00 A61B6/00

    摘要: A method and system for regression-based object detection in medical images is disclosed. A regression function for predicting a location of an object in a medical image based on an image patch is trained using image-based boosting ridge regression (IBRR). The trained regression function is used to determine a difference vector based on an image patch of a medical image. The difference vector represents the difference between the location of the image patch and the location of a target object. The location of the target object in the medical image is predicted based on the difference vector determined by the regression function.

    摘要翻译: 公开了一种用于医学图像中基于回归的物体检测的方法和系统。 使用基于图像的增强脊回归(IBRR)训练用于基于图像块预测医学图像中的对象的位置的回归函数。 训练回归函数用于基于医学图像的图像块来确定差分矢量。 差分向量表示图像块的位置与目标对象的位置之间的差异。 基于由回归函数确定的差分向量来预测目标对象在医学图像中的位置。