Abstract: A method comprises (a) collecting (i) a set of chest computed tomography angiography (CTA) images scanned in the axial view and (ii) a manual segmentation of the images, for each one of multiple organs; (b) preprocessing the images such that they share the same field of view (FOV); (c) using both the images and their manual segmentation to train a supervised deep learning segmentation network, wherein loss is determined from a multi-dice score that is the summation of the dice scores for all the multiple organs, each dice score being computed as the similarity between the manual segmentation and the output of the network for one of the organs; (d) testing a given (input) pre-processed image on the trained network, thereby obtaining segmented output of the given image; and (e) smoothing the segmented output of the given image.

Abstract: Methods and systems for automatically linking entries in a medical image report to an image of a medical image study. One method includes identifying a description of pathology in text included in the medical image report using natural language processing and applying a model to select for the medical image report at least one modality and at least one image included in the at least one medical image study generated by the at least one modality. The method also includes creating a data link between the description of pathology and the at least one image included in the at least one medical image study generated by the at least one modality, and automatically inserting the data link into the medical image report, the data link being selectable to display the at least one image included in the at least one image generated by the at least one modality.

Abstract: Methods and systems for automatically linking entries in a medical image report to an image of a medical image study. One method includes identifying a first plurality of image features referenced in text included in the medical image report using natural language processing, identifying a second plurality of image features in images included in the medical image study, and comparing the first plurality of image features and the second plurality of image features. In response to a first image feature included in the first plurality of image features and a second image feature included in the second plurality of image features matching, the method includes creating a data link between the medical image report and an image included in the medical image study including the second image feature, and automatically inserting the data link into the medical image report, the data link selectable by a user to display the image.

Abstract: Methods and systems for automatically linking entries in a medical image report to an image of a medical image study. One method includes identifying a first plurality of image features referenced in text included in the medical image report using natural language processing, identifying a second plurality of image features in images included in the medical image study, and comparing the first plurality of image features and the second plurality of image features. In response to a first image feature included in the first plurality of image features and a second image feature included in the second plurality of image features matching, the method includes creating a data link between the medical image report and an image included in the medical image study including the second image feature, and automatically inserting the data link into the medical image report, the data link selectable by a user to display the image.

Abstract: Methods and systems for automatically linking entries in a medical image report to an image of a medical image study. One method includes identifying a description of pathology in text included in the medical image report using natural language processing and applying a model to select for the medical image report at least one modality and at least one image included in the at least one medical image study generated by the at least one modality. The method also includes creating a data link between the description of pathology and the at least one image included in the at least one medical image study generated by the at least one modality, and automatically inserting the data link into the medical image report, the data link being selectable to display the at least one image included in the at least one image generated by the at least one modality.

Abstract: A method comprises (a) collecting (i) a set of chest computed tomography angiography (CTA) images scanned in the axial view and (ii) a manual segmentation of the images, for each one of multiple organs; (b) preprocessing the images such that they share the same field of view (FOV); (c) using both the images and their manual segmentation to train a supervised deep learning segmentation network, wherein loss is determined from a multi-dice score that is the summation of the dice scores for all the multiple organs, each dice score being computed as the similarity between the manual segmentation and the output of the network for one of the organs; (d) testing a given (input) pre-processed image on the trained network, thereby obtaining segmented output of the given image; and (e) smoothing the segmented output of the given image.