Skip to main content
. 2022 Jun 2;13:882651. doi: 10.3389/fimmu.2022.882651

Figure 3.

Figure 3

Parameters were selected to build LASSO regression and nomogram models. (A) The binomial deviation was plotted through the optimal log (Lambda) value. (B) 12 non-zero coefficient parameters are screened by the optimal log (Lambda) value. (C) The nomogram predicting the risk of spinal tuberculosis was plotted in the training set. (D) The nomogram predicting the risk of spinal tuberculosis was plotted in the external validation set.