Decoding Wilson disease: a machine learning approach to predict neurological symptoms

Abstract

Objectives: Wilson disease (WD) is a rare autosomal recessive disorder caused by a mutation in the ATP7B gene. Neurological symptoms are one of the most common symptoms of WD. This study aims to construct a model that can predict the occurrence of neurological symptoms by combining clinical multidimensional indicators with machine learning methods.

Methods: The study population consisted of WD patients who received treatment at the First Affiliated Hospital of Anhui University of Traditional Chinese Medicine from July 2021 to September 2023 and had a Leipzig score ≥ 4 points. Indicators such as general clinical information, imaging, blood and urine tests, and clinical scale measurements were collected from patients, and machine learning methods were employed to construct a prediction model for neurological symptoms. Additionally, the SHAP method was utilized to analyze clinical information to determine which indicators are associated with neurological symptoms.

Results: In this study, 185 patients with WD (of whom 163 had neurological symptoms) were analyzed. It was found that using the eXtreme Gradient Boosting (XGB) to predict achieved good performance, with an MCC value of 0.556, ACC value of 0.929, AUROC value of 0.835, and AUPRC value of 0.975. Brainstem damage, blood creatinine (Cr), age, indirect bilirubin (IBIL), and ceruloplasmin (CP) were the top five important predictors. Meanwhile, the presence of brainstem damage and the higher the values of Cr, Age, and IBIL, the more likely neurological symptoms were to occur, while the lower the CP value, the more likely neurological symptoms were to occur.

Conclusions: To sum up, the prediction model constructed using machine learning methods to predict WD cirrhosis has high accuracy. The most important indicators in the prediction model were brainstem damage, Cr, age, IBIL, and CP. It provides assistance for clinical decision-making.

Keywords: Wilson disease; eXtreme Gradient Boosting (XGB); machine learning; neurological symptom; prediction model.

Figure 1

The flowchart of this study.

Figure 2

Performance comparison between SVM, RF, XGB, LR, and BP on test sets. SVM, Support Vector Machine; RF, Random Forest; XGB, eXtreme Gradient Boosting; LR, logistic regression; BP, back propagation neural network. MCC, Matthews correlation coefficient; ACC, accuracy; AUC, area under the receiver operating characteristic curve; AUPRC, area under the precision-recall curve.

Figure 3

The ROC and PR curves of the prediction models. (A) The ROC curve of the prediction model; (B) the PR curve of the prediction model.

Figure 4

Feature analysis results. The abscissa represents the weight of the influence, instead of the specific value of the feature; the influence of feature value size on the results is represented by color (red represents a large value, blue represents a small value, and purple is adjacent to the mean). Cr, creatinine; IBIL, indirect bilirubin; CP, ceruloplasmin; DBIL, direct bilirubin; TT, thrombin time; WBC, white blood cell; BUN, blood urea nitrogen; Hb, hemoglobin; APTT, activated partial thromboplastin time; PLT, platelet count; GGT, γ-Glutamyl Transferase; PT, prothrombin time.

Figure 5

Magnitude of the influence of the indicator on neurological symptoms. Cr, creatinine; IBIL, indirect bilirubin; CP, ceruloplasmin; DBIL, direct bilirubin; TT, thrombin time; WBC, white blood cell; BUN, blood urea nitrogen; Hb, hemoglobin; APTT, activated partial thromboplastin time; PLT, platelet count; GGT, γ-Glutamyl Transferase; PT, prothrombin time.