Figure 3 | Scientific Reports

Figure 3

From: Machine learning application for the prediction of SARS-CoV-2 infection using blood tests and chest radiograph

Figure 3

Case examples of human and machine learning model prediction. The cut-off threshold for SHAP model is 0.48 meaning that if the model output value is above this, then the prediction is positive. The relative contribution of each laboratory marker is shown in the individual SHAP value plot. (a) An elderly female with a positive prediction from chest X-ray (bilateral lower zones shadowing) and positive prediction from laboratory markers (WBC: 5.29, lymphocytes: 1.09, LDH: 247, and CRP: 1.63). The ground-truth COVID-19 RT-PCR result is positive. (b) An elderly male with a negative prediction from chest X-ray (normal radiographic appearance) and positive prediction from laboratory markers (LDH: 178, lymphocytes: 1.46, platelet: 146, and CRP: 1.033The ground-truth COVID-19 RT-PCR result is positive.

Back to article page