Matches in SemOpenAlex for { <https://semopenalex.org/work/W4285496103> ?p ?o ?g. }
Showing items 1 to 65 of
65
with 100 items per page.
- W4285496103 abstract "Purpose: To introduce a novel technique for pretraining deep neural networks on mammographic images, where the network learns to predict multiple metadata attributes and simultaneously to match images from the same patient and study. Further to demonstrate how this network can be used to produce explainable predictions. Methods: We trained a neural network on a dataset of 85,558 raw mammographic images and seven types of metadata, using a combination of supervised and self-supervised learning techniques. We evaluated the performance of our model on a dataset of 4,678 raw mammographic images using classification accuracy and correlation. We also designed an ablation study to demonstrate how the model can produce explainable predictions. Results: The model learned to predict all but one of the seven metadata fields with classification accuracy ranging from 78-99% on the validation dataset. The model was able to predict which images were from the same patient with over 93% accuracy on a balanced dataset. Using a simple X-ray system classifier built on top of the first model, representations learned on the initial X-ray system classification task showed by far the largest effect size on ablation, illustrating a method for producing explainable predictions. Conclusions: It is possible to train a neural network to predict several kinds of mammogram metadata simultaneously. The representations learned by the model for these tasks can be summed to produce an image representation that captures features unique to a patient and study. With such a model, ablation offers a promising method to enhance the explainability of deep learning predictions." @default.
- W4285496103 created "2022-07-15" @default.
- W4285496103 creator A5051485379 @default.
- W4285496103 creator A5067717824 @default.
- W4285496103 creator A5071070019 @default.
- W4285496103 creator A5086802773 @default.
- W4285496103 date "2022-07-13" @default.
- W4285496103 modified "2023-09-29" @default.
- W4285496103 title "Mammographic image metadata learning for model pretraining and explainable predictions" @default.
- W4285496103 cites W2183341477 @default.
- W4285496103 cites W2295739661 @default.
- W4285496103 cites W3148908072 @default.
- W4285496103 cites W4210718353 @default.
- W4285496103 cites W4236392513 @default.
- W4285496103 cites W4285552158 @default.
- W4285496103 doi "https://doi.org/10.1117/12.2626199" @default.
- W4285496103 hasPublicationYear "2022" @default.
- W4285496103 type Work @default.
- W4285496103 citedByCount "1" @default.
- W4285496103 countsByYear W42854961032023 @default.
- W4285496103 crossrefType "proceedings-article" @default.
- W4285496103 hasAuthorship W4285496103A5051485379 @default.
- W4285496103 hasAuthorship W4285496103A5067717824 @default.
- W4285496103 hasAuthorship W4285496103A5071070019 @default.
- W4285496103 hasAuthorship W4285496103A5086802773 @default.
- W4285496103 hasConcept C108583219 @default.
- W4285496103 hasConcept C111919701 @default.
- W4285496103 hasConcept C119857082 @default.
- W4285496103 hasConcept C124101348 @default.
- W4285496103 hasConcept C132964779 @default.
- W4285496103 hasConcept C153180895 @default.
- W4285496103 hasConcept C154945302 @default.
- W4285496103 hasConcept C199360897 @default.
- W4285496103 hasConcept C41008148 @default.
- W4285496103 hasConcept C50644808 @default.
- W4285496103 hasConcept C93518851 @default.
- W4285496103 hasConcept C95623464 @default.
- W4285496103 hasConceptScore W4285496103C108583219 @default.
- W4285496103 hasConceptScore W4285496103C111919701 @default.
- W4285496103 hasConceptScore W4285496103C119857082 @default.
- W4285496103 hasConceptScore W4285496103C124101348 @default.
- W4285496103 hasConceptScore W4285496103C132964779 @default.
- W4285496103 hasConceptScore W4285496103C153180895 @default.
- W4285496103 hasConceptScore W4285496103C154945302 @default.
- W4285496103 hasConceptScore W4285496103C199360897 @default.
- W4285496103 hasConceptScore W4285496103C41008148 @default.
- W4285496103 hasConceptScore W4285496103C50644808 @default.
- W4285496103 hasConceptScore W4285496103C93518851 @default.
- W4285496103 hasConceptScore W4285496103C95623464 @default.
- W4285496103 hasLocation W42854961031 @default.
- W4285496103 hasOpenAccess W4285496103 @default.
- W4285496103 hasPrimaryLocation W42854961031 @default.
- W4285496103 hasRelatedWork W3014300295 @default.
- W4285496103 hasRelatedWork W3164822677 @default.
- W4285496103 hasRelatedWork W4223943233 @default.
- W4285496103 hasRelatedWork W4225161397 @default.
- W4285496103 hasRelatedWork W4246751904 @default.
- W4285496103 hasRelatedWork W4312200629 @default.
- W4285496103 hasRelatedWork W4360585206 @default.
- W4285496103 hasRelatedWork W4364306694 @default.
- W4285496103 hasRelatedWork W4380075502 @default.
- W4285496103 hasRelatedWork W4380086463 @default.
- W4285496103 isParatext "false" @default.
- W4285496103 isRetracted "false" @default.
- W4285496103 workType "article" @default.