Multi-modal fusion approaches aim to integrate information from different data sources.
Ranked #1 on Phenotype classification on MIMIC-CXR, MIMIC-IV
The healthcare domain is characterized by heterogeneous data modalities, such as imaging and physiological data.
Here, we propose a multi-label generalized zero shot learning (CXR-ML-GZSL) network that can simultaneously predict multiple seen and unseen diseases in CXR images.
1 code implementation • 28 Nov 2020 • Ghadeer O. Ghosheh, Bana Alamad, Kai-Wen Yang, Faisil Syed, Nasir Hayat, Imran Iqbal, Fatima Al Kindi, Sara Al Junaibi, Maha Al Safi, Raghib Ali, Walid Zaher, Mariam Al Harbi, Farah E. Shamout
In test set B (225 patient encounters), the respective system achieves 0. 90 AUROC for AKI, elevated troponin, and elevated interleukin-6, and >0. 80 AUROC for most of the other complications.
The existing zero-shot detection approaches project visual features to the semantic domain for seen objects, hoping to map unseen objects to their corresponding semantics during inference.
Ranked #1 on Zero-Shot Object Detection on ImageNet Detection
Finally, we conclude with a set of recommendations on how to assess the results using a new AL algorithm to ensure results are reproducible and robust under changes in experimental conditions.
Ranked #6 on Active Learning on CIFAR10 (10,000)