The collection of cases was acquired at Stony Brook University from patients who tested positive for COVID-19. The collection includes images from different modalities and organ sites (chest radiographs, chest CTs, brain MRIs, etc). Radiology imaging data is extremely important in COVID-19 from both a diagnostic and a monitoring perspective, given the crucial nature of COVID-19 pulmonary disease and its rapid phenotypic changes.

The datasets are available for non-commercial use by research communities for building AI systems for diagnostic and prognostic modeling. (KG & JK, should this be part of the Summary or is covered in the Licensing Agreement on the Citation Tab?)

Data collection was enabled by the Renaissance School of Medicine at Stony Brook University’s “COVID-19 Data Commons and Analytic Environment”, a data quality initiative instituted by the Office of the Dean, and supported by the Department of Biomedical Informatics. (KG & JK, this seems more like an Acknowledgement.  What do you think?  We can check with the site.)

Original Summary:

We have curated clinical and imaging data for covid19-positive patients admitted to the SBU hospital. The dataset consists of de-identified Radiology imaging data along with linked clinical data for each patient. The clinical data consists of diagnoses, procedures, lab tests, covid19 specific data values (e.g., intubation status, symptoms at admission) and a set of derived data elements, which were used in analyses of this data. The clinical data is stored as a set of csv files which comply with OMOP Common Data Model data elements.

Additional questions site was asked to address:

Please address the following items in your revision:

1. Add something similar to: "this collection of cases represents the entire Stony Brook COVID 19 experience over a period of several months there are quite a number of scanners with a number of DICOM tags relating to settings."
2. Update the abstract summarizing what body parts were imaged with each modality.  E.g. "COVID 19 impacts many organ sites.  This dataset includes all imaging studies carried out on the patient population during their hospitalization including CT's of the lung, MRI's of the brain, ..."
3. Update the abstract with a sentence along the lines of "This collection provides all scans carried out during the hospitalization of each Covid 19 positive patient admitted during the targeted time period, giving researchers a valuable opportunity to study disease progression."