Abstract: Systems and methods can include a method for training a deep convolutional neural network to provide a patient radiation treatment plan, the method comprising collecting patient data from a group of patients, the patient data including at least one image of patient anatomy and a prior treatment plan, wherein the treatment plan includes predetermined machine parameters, and training a deep convolution neural network for regression by using the prior treatment plans and the corresponding collected patient data to determine a new treatment plan.

Abstract: Systems and methods can include a method for training a deep convolutional neural network to provide a patient radiation treatment plan, the method comprising collecting patient data from a group of patients, the patient data including at least one image of patient anatomy and a prior treatment plan, wherein the treatment plan includes predetermined machine parameters, and training a deep convolution neural network for regression by using the prior treatment plans and the corresponding collected patient data to determine a new treatment plan.

Abstract: Methods and systems are disclosed for correcting segmentation errors in pre-existing contours and surfaces. Techniques are disclosed for receiving one or more edit contours from a user, identifying pre-existing data points that should be eliminated, and generating a new corrected surface. Embodiments disclosed herein relate to using received edit contours to generate a set of points on a pre-existing surface, and applying a proximity test to eliminate pre-existing constraint points that are undesirable.

Abstract: Disclosed herein are techniques for performing automated structure delineation on image data using trained landmark detectors and a shape refinement tool. The landmark detectors can be trained to detect a landmark in the image data based on image features that are indicative of intensity variations over a plurality of windows of the image data points. A machine-learning algorithm can be used to train the landmark detectors. The landmarks in the image data that are detected by the trained landmark detects can be used to initialize an iterative shape refinement to thereby compute a refined shape estimate for a structure of interest such as a prostate.

Abstract: Methods and systems are disclosed for correcting segmentation errors in pre-existing contours and surfaces. Techniques are disclosed for receiving one or more edit contours from a user, identifying pre-existing data points that should be eliminated, and generating a new corrected surface. Embodiments disclosed herein relate to using received edit contours to generate a set of points on a pre-existing surface, and applying a proximity test to eliminate pre-existing constraint points that are undesirable.

Abstract: Disclosed herein are techniques for performing automated structure delineation on image data using trained landmark detectors and a shape refinement tool. The landmark detectors can be trained to detect a landmark in the image data based on image features that are indicative of intensity variations over a plurality of windows of the image data points. A machine-learning algorithm can be used to train the landmark detectors. The landmarks in the image data that are detected by the trained landmark detects can be used to initialize an iterative shape refinement to thereby compute a refined shape estimate for a structure of interest such as a prostate.

Abstract: Methods and systems are disclosed for correcting segmentation errors in pre-existing contours and surfaces. Techniques are disclosed for receiving one or more edit contours from a user, identifying pre-existing data points that should be eliminated, and generating a new corrected surface. Embodiments disclosed herein relate to using received edit contours to generate a set of points on a pre-existing surface, and applying a proximity test to eliminate pre-existing constraint points that are undesirable.

Abstract: A technique is disclosed for generating a new contour and/or a 3D surface such as a variational implicit surface from contour data. In one embodiment, B-spline interpolation is used to efficiently generate a new contour (preferably a transverse contour), from a plurality of input contours (preferably, sagittal and/or coronal contours). In another embodiment, a point reduction operation is performed on data sets corresponding to any combination of transverse, sagittal, or coronal contour data prior to processing those data sets to generate a 3D surface such as a variational implicit surface. A new contour can also be generated by the intersection of this surface with an appropriately placed and oriented plane. In this manner, the computation of the variational implicit surface becomes sufficiently efficient to make its use for new contour generation practical.

Abstract: A technique is disclosed for generating a new contour and/or a 3D surface such as a variational implicit surface from contour data. In one embodiment, a point reduction operation is performed on data sets corresponding to any combination of transverse, sagittal, or coronal contour data prior to processing those data sets to generate a 3D surface such as a variational implicit surface. A new contour can also be generated by the intersection of this surface with an appropriately placed and oriented plane. In this manner, the computation of the variational implicit surface becomes sufficiently efficient to make its use for new contour generation practical.

Abstract: Methods and systems are disclosed for correcting segmentation errors in pre-existing contours and surfaces. Techniques are disclosed for receiving one or more edit contours from a user, identifying pre-existing data points that should be eliminated, and generating a new corrected surface. Embodiments disclosed herein relate to using received edit contours to generate a set of points on a pre-existing surface, and applying a proximity test to eliminate pre-existing constraint points that are undesirable.

Abstract: A technique is disclosed for generating a new contour and/or a 3D surface from contour data using a surface displacement field. The technique is preferably applied to un-contoured target images within a 4D image sequence for which contour data is desired. The technique can be initialized with contour data related to a reference image that has already been contoured, e.g. by a doctor who manually enters contour information into a computer. Image registration (calculation of correspondence between the reference image and target image) can be performed simultaneously with segmentation (identifying regions of interest in the target image). This allows one to make use of segmentation information in a reference image to improve the accuracy of contouring within the target image.

Abstract: A technique is disclosed for generating a new contour and/or a 3D surface such as a variational implicit surface from contour data. In one embodiment, B-spline interpolation is used to efficiently generate a new contour (preferably a transverse contour), from a plurality of input contours (preferably, sagittal and/or coronal contours). In another embodiment, a point reduction operation is performed on data sets corresponding to any combination of transverse, sagittal, or coronal contour data prior to processing those data sets to generate a 3D surface such as a variational implicit surface. A new contour can also be generated by the intersection of this surface with an appropriately placed and oriented plane. In this manner, the computation of the variational implicit surface becomes sufficiently efficient to make its use for new contour generation practical.

Abstract: A technique is disclosed for generating a new contour and/or a 3D surface such as a variational implicit surface from contour data. In one embodiment, a point reduction operation is performed on data sets corresponding to any combination of transverse, sagittal, or coronal contour data prior to processing those data sets to generate a 3D surface such as a variational implicit surface. A new contour can also be generated by the intersection of this surface with an appropriately placed and oriented plane. In this manner, the computation of the variational implicit surface becomes sufficiently efficient to make its use for new contour generation practical.

Abstract: A technique is disclosed for generating a new contour and/or a 3D surface such as a variational implicit surface from contour data. In one embodiment, B-spline interpolation is used to efficiently generate a new contour (preferably a transverse contour), from a plurality of input contours (preferably, sagittal and/or coronal contours). In another embodiment, a point reduction operation is performed on data sets corresponding to any combination of transverse, sagittal, or coronal contour data prior to processing those data sets to generate a 3D surface such as a variational implicit surface. A new contour can also be generated by the intersection of this surface with an appropriately placed and oriented plane. In this manner, the computation of the variational implicit surface becomes sufficiently efficient to make its use for new contour generation practical.

Abstract: A technique is disclosed for generating a new contour and/or a 3D surface from contour data using a surface displacement field. The technique is preferably applied to un-contoured target images within a 4D image sequence for which contour data is desired. The technique can be initialized with contour data related to a reference image that has already been contoured, e.g. by a doctor who manually enters contour information into a computer. Image registration (calculation of correspondence between the reference image and target image) can be performed simultaneously with segmentation (identifying regions of interest in the target image). This allows one to make use of segmentation information in a reference image to improve the accuracy of contouring within the target image.

Abstract: A technique is disclosed for generating a new contour and/or a 3D surface such as a variational implicit surface from contour data. In one embodiment, B-spline interpolation is used to efficiently generate a new contour (preferably a transverse contour), from a plurality of input contours (preferably, sagittal and/or coronal contours). In another embodiment, a point reduction operation is performed on data sets corresponding to any combination of transverse, sagittal, or coronal contour data prior to processing those data sets to generate a 3D surface such as a variational implicit surface. A new contour can also be generated by the intersection of this surface with an appropriately placed and oriented plane. In this manner, the computation of the variational implicit surface becomes sufficiently efficient to make its use for new contour generation practical.

Abstract: A method of approximating the boundary of an object in an image, the image being represented by a data set, the data set comprising a plurality of data elements, each data element having a data value corresponding to a feature of the image, the method comprising determining which one of a plurality of contours most closely matches the object boundary at least partially according to a divergence value for each contour, the divergence value being selected from the group consisting of Jensen-Shannon divergence and Jensen-Renyi divergence. Each contour Ci defines a zone ZIi and a zone ZOi, ZIi representing the data elements inside the contour and ZOi representing the data elements outside the contour, each zone having a corresponding probability distribution of data values for the data elements therein, and wherein the divergence value for each contour Ci represents a measure of the difference between the probability distributions for the zones ZIi and ZOi.

Abstract: A method of approximating the boundary of an object in an image, the image being represented by a data set, the data set comprising a plurality of data elements, each data element having a data value corresponding to a feature of the image, the method comprising determining which one of a plurality of contours most closely matches the object boundary at least partially according to a divergence value for each contour, the divergence value being selected from the group consisting of Jensen-Shannon divergence and Jensen-Renyi divergence. Each contour Ci defines a zone ZIi and a zone ZOi, ZIi representing the data elements inside the contour and ZOi representing the data elements outside the contour, each zone having a corresponding probability distribution of data values for the data elements therein, and wherein the divergence value for each contour Ci represents a measure of the difference between the probability distributions for the zones ZIi and ZOi.