Model combining and interaction for medical imaging

1. A computer-implemented method for facilitating medical image analysis, comprising:

obtaining a plurality of artificial intelligence (AI) models for medical image analysis from one or more model providers;

organizing the plurality of models into groups to form associations among the models, wherein at least a subset of the groups reference one another to associate models with one another;

in response to a user request received from a user device, providing one or more base models selected from the plurality of models;

searching, based on the one or more base models, the organized plurality of models in accordance with the models' associations to identify one or more additional models as candidates for combining with the one or more base models;

providing the one or more additional models selected from the organized plurality of models to combine with the one or more base models; and

causing presentation of medical image analysis results based, at least in part, on applying a combination of the one or more base models and the one or more additional models to target medical image data, wherein causing presentation of the medical image analysis results comprises causing presentation of image overlay features concurrently with one or more images of the target medical image data.

2. The method of claim 1 , wherein organizing the plurality of models comprises organizing the plurality of models into a hierarchy of the groups based on grouping criteria.

3. The method of claim 2 , wherein the grouping criteria includes at least one of a similarity of input between models, a similarity of output between models, or an overlap between a model's output and another model's input.

4. The method of claim 1 , wherein the user request indicates at least one of an analysis purpose, context, applicable medical data, model structure, model input or output, or performance requirement.

5. The method of claim 1 , further comprising selecting the one or more base models based, at least in part, on the associations among the models.

6. The method of claim 1 , wherein the one or more additional models and the one or more base models are designed to receive same or overlapping input features and to generate different output features.

7. The method of claim 1 , wherein the one or more additional models and the one or more base models are designed to generate same or overlapping output features.

8. The method of claim 1 , wherein the one or more additional models and the one or more base models are linkable to form a configurable workflow.

9. The method of claim 8 , wherein the configurable workflow includes at least one of a chain, tree, or lattice structure to link models.

10. The method of claim 1 , wherein causing presentation of the medical image analysis results comprises causing presentation of one or more user interfaces via the user device.

11. The method of claim 10 , wherein causing presentation of the medical image analysis results further comprises causing presentation of image overlay features corresponding to results from the combination of the one or more base models and the one or more additional models, via the one or more user interfaces.

12. One or more non-transitory computer-readable media collectively storing contents that, when executed by one or more processors, cause the one or more processors to perform actions comprising:

organizing a plurality of models for medical image analysis into groups to form associations among the models, wherein at least a subset of the groups reference one another to associate models with one another;

searching, based on one or more first models, the organized plurality of models in accordance with the models' associations to identify one or more second models for combining with the one or more first models;

combining at least a subset of the one or more first models with at least a subset of the one or more second models to form a combination of models based, at least in part, on a user request; and

causing presentation of medical image analysis results based, at least in part, on applying the combination of models to target medical image data, wherein causing presentation of the medical image analysis results comprises causing presentation of image overlay features concurrently with one or more images of the target medical image data.

13. The one or more non-transitory computer-readable media of claim 12 , wherein the image overlay features include at least one of bounding boxes for object detection and localization, or graded heat maps corresponding to clinical metrics.

14. The one or more non-transitory computer-readable media of claim 12 , wherein the image overlay features include a single feature that integrates results generated from individual models that form the combination of models.

15. The one or more non-transitory computer-readable media of claim 14 , wherein the single feature is generated by at least one of intersection, union, averaging, weighted averaging, or probabilistic sampling operation.

16. A system, comprising:

one or more processors; and

non-transitory memory storing contents that, when executed by the one or more processors, cause the system to:

organize a plurality of models into groups for medical image analysis to form associations among the models, wherein at least a subset of the groups reference one another to associate models with one another;

search, based on one or more first models, the organized plurality of models in accordance with the models' associations to identify one or more second models for combining with the one or more first models;

combine at least a subset of the one or more first models with at least a subset of the one or more second models to form a combination of models based, at least in part, on a user request; and

cause presentation of medical image analysis results based, at least in part, on applying the combination of models to target medical image data, wherein causing presentation of the medical image analysis results comprises causing presentation of image overlay features concurrently with one or more images of the target medical image data.

17. The system of claim 16 , wherein organizing the plurality of models comprises organizing the plurality of models into a hierarchy of the groups based on grouping criteria.

18. The system of claim 16 , wherein individual models that form the combination of models are designed to receive same or overlapping input features and to generate different output features.

19. The system of claim 16 , wherein individual models of that form the combination of models are designed to generate same or overlapping output features.

20. The system of claim 16 , wherein individual models of that form the combination of models are linkable to form a workflow including at least one of a chain, tree, or lattice structure to link models.