It is a collection of AI radiology thoracic models, containers, and deployment pipelines that can be used in the fight against COVID-19.
The collection includes pre-trained models for CT Lung Segmentation, COVID Classification in chest studies, and a CT annotation lung model. The classification model for differentiation of pneumonia due to COVID-19 from other causes of pneumonia, may find significant use during the cold and flu season to help differentiate COVID-19 from other types of pneumonias.
The CT lung lesion segmentation model can be used to identify regions of the lung that are affected by COVID-19. The model also predicts the ratio between the volume of the area affected by COVID-19 and the volume of the lung, to indicate the severity of the disease which can be used for clinical triage.
The EXAM (EMR CXR AI Model) model predicts the oxygen needs of patients with COVID-19 symptoms, using chest X-rays and EHR data such as patient vitals and lab data.
The Clara Train container that is part of this collection enables researchers to adapt the above models to other datasets, or build new models from scratch.
Besides the COVID AI models, this collection also includes deployment pipelines. The Deploy Reference Pipeline for COVID-19 Chest CT Classification provides researchers with a reference deployment pipeline which can be seamlessly evaluated for localized data. Today, this pipeline runs the COVID classification model that can be used to predict if a patient has COVID-19 from radiology images.
The Deploy Reference Pipeline will be able to run the CT Lesion Segmentation model.