FIGURE 4.
Principal component screening and feature correlation analysis. Principal component analysis revealed the 4 features that had the greatest impact on prognosis (a). Covariance analysis showed the interdependence and relevance between the selected 18 features (b)