Siemens Healthineers International Ag

Varian Medical Systems International Ag

Angelo Genghi

Benjamin M Haas

Pascal Paysan

Tomasz Morgas

Simon Mathias Fluckiger

Thomas Coradi

Alexander Maslowski

Fernando Franco

Mário Joao Fartaria

Anna Siroki-Galambos

Anri Friman

Systems and methods for augmenting a training data set with annotated pseudo images for training machine learning models. The pseudo images are generated from corresponding images of the training data set and provide a realistic model of the interaction of image generating signals with the patient, while also providing a realistic patient model. The pseudo images are of a target imaging modality, which is different than the imaging modality of the training data set, and are generated using algorithms that account for artifacts of the target imaging modality. The pseudo images may include therein the contours and/or features of the anatomical structures contained in corresponding medical images of the training data set. The trained models can be used to generate contours in medical images of a patient of the target imaging modality or to predict an anatomical condition that may be indicative of a disease.

Systems and methods for pseudo image data augmentation for training machine learning models

Embodiments described herein provide for training a machine learning model for automatic organ segmentation. A processor executes a machine learning model using an image to output at least one predicted organ label for a plurality of pixels of the image. Upon transmitting the at least one predicted organ label to a correction computing device, the processor receives one or more image fragments identifying corrections to the at least one predicted organ label. Upon transmitting the one or more image fragments and the image to a plurality of reviewer computing devices, the processor receives a plurality of inputs indicating whether the one or more image fragments are correct. When a number of inputs indicating an image fragment of the image fragments is correct exceeds a threshold, the processor aggregates the image fragment into a training data set. The processor trains the machine learning model with the training data set.

Systems and methods for refining training data

Embodiments described herein provide a method for generating training data for Al based atlas mapping slice localization, and a system and method for using the training data to train a deep learning network. Training data development maps each slice of an input medical image to a position in a full body reference atlas along the longitudinal body axis. The method constructs a landmarking table of 2D slices indicating known anatomic landmarks of a reference subject, and interpolated slices. A final step for obtaining training data uses regression analysis techniques to create a vector of longitudinal axis coordinates of all slices from the input image. The training data is used to train a deep learning model to create an AI-based atlas mapping slice localizer model. The trained AI-based atlas mapping slice localizer model can be applied to generate mapping inputs to autosegmentation models to improve efficiency and reliability of contouring.

AI-based atlas mapping slice localizer for deep learning autosegmentation

Growing community of inventors

Wettingen, Switzerland

Angelo Genghi