Skip to main content
. 2023 Jul 1;13:10666. doi: 10.1038/s41598-023-37835-1

Figure 1.

Figure 1

Visualization of the data extraction pipeline. For training, we only make use of those patients who were admitted to the Intensive Care Unit (n = 53,150) and who had clinical data (clinical parameters—CP) with matching chest radiographs available (n = 6,125). The data is split into the training (n = 4,396 patients), validation (n = 472 patients), and test sets (n = 1,257 patients).