Siemens Healthineers Ag

Athira Jane Jacob

Puneet Sharma

Mehmet Akif Gulsun

Gareth Funka-Lea

Ingmar Voigt

Noha Youssry El-Zehiry

Teodora Marina Chitiboi

Karim Amer

Sayan Ghosal

For training for and performance of LGE analysis, multi-task machine-learning model is trained to output various cardiac tissue characteristics based on input of LGE MR data. The use of segmentation may be avoided or limited, resulting in a greater number of available training data samples, by using radiology clinical reports with LGE information as a source for samples. The multi-task model may be trained to output cardiac tissue characteristics using radiology clinical reports with LGE information with no segmentation or with segmentation for only a subset of the training samples. By training for multiple tasks, the accuracy of prediction for each task benefits from the information for other tasks. The trained model outputs values of characteristics for multiple tasks, such as extent of enhancement, type of enhancement, and localization of enhancement. Other tasks may be included, such as disease classification. Other inputs may be used, such as also including sensor data and/or cardiac motion.

Late gadolinium enhancement analysis for magnetic resonance imaging

Systems and methods for performing a medical imaging analysis task are provided. An input medical image in a first modality is received. Features are extracted from the input medical image using a first machine learning based encoding network. A medical imaging analysis task is performed on the input medical image based on the extracted features by, in one embodiment, decoding the extracted features to generate results of the medical imaging analysis task using a machine learning based decoding network. Results of the medical imaging analysis task are output. In one embodiment, the first machine learning based encoding network is jointly trained with a second machine learning based encoding network with an unsupervised loss using unannotated pairs of training images. Each of the unannotated pairs comprise a first training image in the first modality and a second training image in a second modality. In one embodiment, the first machine learning based encoding network is also jointly trained with the machine learning based decoding network with a supervised loss using annotated training images.

Semi-supervised learning leveraging cross-domain data for medical imaging analysis

Systems and methods for graph based assessment of a patient are provided. Medical imaging data and non-imaging medical data of a patient are received. The medical imaging data and the non-imaging medical data are encoded into encoded features using a graph based machine learning network by comparing the patient with patients of a patient population based on a graph of the patient population. An assessment of the patient is determined based on the encoded features using a machine learning classifier network. The assessment of the patient is output.

Subpopulation based patient risk prediction using graph attention networks

A computer-implemented method for updating a boundary segmentation, the method including receiving image data and an original boundary segmentation including a plurality of boundary points. A plurality of edges in the image data is detected and used to generate an edge map. A confidence for a boundary point in the original boundary segmentation is computed where the confidence is based on a distance between the boundary point and an edge point associated with at least one of the plurality of edges of the edge map, and based on the confidence a classification of the boundary point is determined. An updated boundary segmentation based on the classification of the boundary point is generated and then output.

Growing community of inventors

Plainsboro, NJ, United States of America

Athira Jane Jacob