System and method for evaluating the performance of a user in capturing an ultrasound image of an anatomical region

A computer implemented method for assessing whether an image of an internal anatomical region is usable for organ analysis, the method comprising:

• generating the image of the anatomical region based on an imaging probe position;
• processing the generated image using one or more automated machine learning models trained on datasets of labelled training images, the labelled training images being associated with an imaging probe positioned at different positions relative to a training manikin, a human or an animal, the labelled training images having been classified as valid or invalid images by assigning a label thereto, a valid image corresponding to a target cut plane of an organ, and at least some of the invalid images corresponding to a cut plane where the organ is at least partially hidden or incorrectly oriented, the automated machine learning model(s) determining whether the generated image is usable for organ analysis by outputting an image score indicative of a likelihood that the image is valid on the basis of similarities with the labelled training images classified as valid;
• and
• providing for output a visual or audible indication of whether the generated image corresponds to the target cut plane based on said determination.

A training platform for training users in capturing images of an internal anatomical region for the analysis of organs, the training platform comprising:

• a processing device comprising:
• a memory having stored thereon one or more automated machine learning models trained on a dataset of labelled training images associated with different imaging device positions, wherein the labelled training images have been classified as valid or invalid for organ analysis by assigning a label thereto, a valid image corresponding to a target cut plane of an organ, and at least some of the invalid images corresponding to a cut plane where the organ is at least partially hidden or incorrectly oriented;
• and
• a processor configured for:
• processing using the one or more automated machine learning models a generated image resulting from a user moving an imaging device at various imaging device positions relative to a training manikin, a human or an animal, to determine whether the generated image corresponds to a predefined target plane required for the analysis of the organ features shown therein, by outputting an image score indicative of a likelihood that the generated image is valid on the basis of similarities with the valid images;
• and
• providing an output indicative of whether the generated image corresponds to the target cut plane expected for organ analysis and measurements on the basis of similarities with the valid training images for the predefined view.