4
Views
0
CrossRef citations to date
0
Altmetric
ORIGINAL RESEARCH

A Machine Learning Method for Differentiation Crohn’s Disease and Intestinal Tuberculosis

, , , , ORCID Icon, & ORCID Icon show all
Pages 3835-3847 | Received 27 May 2024, Accepted 29 Jul 2024, Published online: 07 Aug 2024

Figures & data

Figure 1 Study flow chart.

Figure 1 Study flow chart.

Table 1 Descriptive Statistics for Machine Learning

Table 2 Performance of All 6 ML Classification Methods

Figure 2 The average AUROC of the XGBoost method after 5-fold cross validation.

Figure 2 The average AUROC of the XGBoost method after 5-fold cross validation.

Figure 3 SHAP method to explain our ML models.

Figure 3 SHAP method to explain our ML models.

Figure 4 LIME method to explain our ML models.

Figure 4 LIME method to explain our ML models.

Table 3 Results of the ML Model and MDT in Clinical Practice

Figure 5 Results of the ML model and MDT in clinical practice.

Figure 5 Results of the ML model and MDT in clinical practice.

Figure 6 Confusion matrix used for calculating the kappa coefficient.

Figure 6 Confusion matrix used for calculating the kappa coefficient.

Data Sharing Statement

All relevant ML methods and codes can be freely accessed at https://github.com/philiplaw1984/IBD/.