Abstract: Systems and methods facilitate privacy-preserving data curation in a federated learning system by transmitting a portion of a potential data sample to a remote location. The portion is inspected for quality to rule out data samples that do not satisfy data curation criteria. The remote examination focuses on checking the region of interest but maintains privacy as the examination is unable to parse any other identifiable subject information such as face, body shape etc. because pixels or voxels outside the portion are not included. The examination results are sent back to the collaborators so that inappropriate data samples can be excluded during federated learning rounds.

Abstract: Systems and methods for performing a medical imaging analysis task are provided. A plurality of 3D (three dimensional) patches extracted from a 3D input medical image is received. A set of local features is extracted from each of the plurality of 3D patches using a machine learning based local feature extractor network. Global features representing relationships between the sets of local features are determined. A medical imaging analysis task is performed on the 3D input medical image based on the global features. Results of the medical imaging analysis task are output.

Abstract: Systems and methods for generating synthesized medical images of a tumor are provided. A 3D mask of an anatomical structure generated from a 3D medical image and a 3D image of a plurality of concentric spheres are received. A 3D mask of a tumor is generated based on the 3D mask of the anatomical structure and the 3D image of the plurality of concentric spheres using a first 3D generator network. A 3D intensity map of the tumor is generated based on the 3D mask of the tumor and the 3D image of the plurality of concentric spheres using a second 3D generator network. A 3D synthesized medical image of the tumor is generated based on one or more 2D slices of the 3D intensity map of the tumor and one or more 2D slices of the 3D medical image using a 2D generator network. The 3D synthesized medical image of the tumor is output.

Abstract: Systems and methods for quantifying a shift of an anatomical object of a patient are provided. A 3D medical image of an anatomical object of a patient is received. An initial location of landmarks on the anatomical object in the 3D medical image is determined using a first machine learning network. A 2D slice depicting the initial location of the landmarks is extracted from the 3D medical image. The initial location of the landmarks in the 2D slice is refined using a second machine learning network. A shift of the anatomical object is quantified based on the refined location of the landmarks in the 2D slice. The quantified shift of the anatomical object is output.

Abstract: Systems and methods for assessing expansion of an abnormality are provided. A first input medical image of a patient depicting an abnormality at a first time and a second input medical image of the patient depicting the abnormality at a second time are received. The second input medical image is registered with the first input medical image. The abnormality is segmented from 1) the first input medical image to generate a first segmentation map and 2) the registered second input medical image to generate a second segmentation map. The first segmentation map and the second segmentation map are combined to generate a combined map. Features are extracted from the first input medical image and the registered second input medical image are based on the combined map. Expansion of the abnormality is assessed based on the extracted features using a trained machine learning based network. Results of the assessment are output.

Abstract: Systems and methods for occlusion detection in medical images are provided. An input medical image of one or more vessels in an anatomical object of a patient is received. One or more anatomical landmarks are identified in the input medical image. A first patch and one or more additional patches are extracted from the input medical image based on the identified one or more anatomical landmarks. The first patch and the one or more additional patches depict different portions of the anatomical object. Features are extracted from the first patch and the one or more additional patches using a machine learning based feature extractor network. An occlusion in the one or more vessels is detected in the first patch based on the extracted features with or without modeling features on a probability distribution function. Results of the detecting are output.

Abstract: Systems and methods for determining a segmentation of a hemorrhage are provided. An input medical image of a hemorrhage of a patient is received. A contour-sensitive segmentation of the hemorrhage from the input medical image is performed using a machine learning based contour-sensitive segmentation network. A detection-sensitive segmentation of the hemorrhage from the input medical image is performed using a machine learning based detection-sensitive segmentation network. A final segmentation of the hemorrhage from the input medical image is determined based on results of the contour-sensitive segmentation and results of the detection-sensitive segmentation. The final segmentation of the hemorrhage is output.

Abstract: Systems and methods for automatic segmentation of lesions from a 3D input medical image are provided. A 3D input medical image depicting one or more lesions is received. The one or more lesions are segmented from one or more 2D slices extracted from the 3D input medical image using a trained 2D segmentation network. 2D features are extracted from results of the segmentation of the one or more lesions from the one or more 2D slices. The one or more lesions are segmented from a 3D patch extracted from the 3D input medical image using a trained 3D segmentation network. 3D features are extracted from results of the segmentation of the one or more lesions from the 3D patch. The extracted 2D features and the extracted 3D features are fused to generate final segmentation results. The final segmentation results are output.

Abstract: A trained ML algorithm may be configured to process medical imaging data to generate a prediction of at least one diagnosis of a patient based on the medical imaging data. The prediction of the at least one diagnosis of the patient is compared with a validated label of the at least one diagnosis of the patient and the performance of the trained ML algorithm is determined based on the comparison. The validated label of the at least one diagnosis of the patient is obtained by parsing a validated radiology report of the patient and the medical imaging data is associated with the validated radiology report. If the performance of the trained ML algorithm is lower than a threshold, an update of parameters of the trained ML algorithm may be triggered based on the validated label.

Abstract: Systems and methods for generating a probabilistic tree of vessels are provided. An input medical image of vessels of a patient is received. Anatomical landmarks are identified in the input medical image. A centerline of the vessels in the input medical image is determined based on the anatomical landmarks. A probabilistic tree of the vessels is generated based on a probability of fit of the anatomical landmarks and the centerline of the vessels. The probabilistic tree of the vessels is output.

Abstract: Systems and methods facilitate privacy-preserving data curation in a federated learning system by transmitting a portion of a potential data sample to a remote location. The portion is inspected for quality to rule out data samples that do not satisfy data curation criteria. The remote examination focuses on checking the region of interest but maintains privacy as the examination is unable to parse any other identifiable subject information such as face, body shape etc. because pixels or voxels outside the portion are not included. The examination results are sent back to the collaborators so that inappropriate data samples can be excluded during federated learning rounds.

Abstract: Systems and methods for determining input data is out-of-domain of an AI (artificial intelligence) based system are provided. Input data for inputting into an AI based system is received. An in-domain feature space of the AI based system and an out-of-domain feature space of the AI based system are modelled. The in-domain feature space corresponds to features of data that the AI based system is trained to classify. The out-of-domain feature space corresponds to features of data that the AI based system is not trained to classify. Probability distribution functions in the in-domain feature space and the out-of-domain feature space are generated for the input data and for the data that the AI based system is trained to classify. It is determined whether the input data is out-of-domain of the AI based system based on the probability distribution functions for the input data and for the data that the AI based system is trained to classify.

Abstract: Systems and Methods for adaptive aggregation in a federated learning model. An aggregation server sends global model weights to all chosen collaborators for initialization. Each collaborator updates the model weights for certain epochs and then sends the updated model weights back to the aggregation server. The aggregation server adaptively aggregates the updated model weights using at least a computed model divergence value and sends the aggregated model weight to collaborators.

Abstract: Systems and methods for assessing expansion of an abnormality are provided. A first input medical image of a patient depicting an abnormality at a first time and a second input medical image of the patient depicting the abnormality at a second time are received. The second input medical image is registered with the first input medical image. The abnormality is segmented from 1) the first input medical image to generate a first segmentation map and 2) the registered second input medical image to generate a second segmentation map. The first segmentation map and the second segmentation map are combined to generate a combined map. Features are extracted from the first input medical image and the registered second input medical image are based on the combined map. Expansion of the abnormality is assessed based on the extracted features using a trained machine learning based network. Results of the assessment are output.

Abstract: Systems and method for performing a medical imaging analysis task for making a clinical decision are provided. One or more input medical images of a patient are received. A medical imaging analysis task is performed from the one or more input medical images using a machine learning based network. The machine learning based network generates a probability score associated with the medical imaging analysis task. An uncertainty measure associated with the probability score is determined. A clinical decision is made based on the probability score and the uncertainty measure.

Abstract: Systems and methods for automatic segmentation of lesions from a 3D input medical image are provided. A 3D input medical image depicting one or more lesions is received. The one or more lesions are segmented from one or more 2D slices extracted from the 3D input medical image using a trained 2D segmentation network. 2D features are extracted from results of the segmentation of the one or more lesions from the one or more 2D slices. The one or more lesions are segmented from a 3D patch extracted from the 3D input medical image using a trained 3D segmentation network. 3D features are extracted from results of the segmentation of the one or more lesions from the 3D patch. The extracted 2D features and the extracted 3D features are fused to generate final segmentation results. The final segmentation results are output.

Abstract: A scheduling system includes: a plurality of input devices configured to output medical data, a workforce storage, configured to store working characteristics of a plurality of doctors, and a scheduler configured to receive as input data related to the medical data and the working characteristics, and configured to provide as output a plurality of schedules for the plurality of doctors for analysing the medical data.

Abstract: Medical images may be classified by receiving a first medical image. The medical image may be applied to a machine learned classifier. The machine learned classifier may be trained on second medical images. A label of the medical image and a measure of uncertainty may be generated. The measure of uncertainty may be compared to a threshold. The first medical image and the label may be output when the measure of uncertainty is within the threshold.

Abstract: Systems and methods for quantifying a shift of an anatomical object of a patient are provided. A 3D medical image of an anatomical object of a patient is received. An initial location of landmarks on the anatomical object in the 3D medical image is determined using a first machine learning network. A 2D slice depicting the initial location of the landmarks is extracted from the 3D medical image. The initial location of the landmarks in the 2D slice is refined using a second machine learning network. A shift of the anatomical object is quantified based on the refined location of the landmarks in the 2D slice. The quantified shift of the anatomical object is output.

Abstract: For saliency mapping, a machine-learned classifier is used to classify input data. A perturbation encoder is trained and/or applied for saliency mapping of the machine-learned classifier. The training and/or application (testing) of the perturbation encoder uses less than all feature maps of the machine-learned classifier, such as selecting different feature maps of different hidden layers in a multiscale approach. The subset used is selected based on gradients from back-projection. The training of the perturbation encoder may be unsupervised, such as using an entropy score, or semi-supervised, such as using the entropy score and a difference of a perturbation mask from a ground truth segmentation.