Clinical decision support tool for breast cancer recurrence prediction using SHAP value in cooperative game theory

Abstract

Background: Recurrence remains the primary cause of death in patients with breast cancer. Although machine learning can efficiently predict the prognosis of breast cancer patients, the black-box nature of the model may result in a lack of evidence for clinicians when making critical decisions.

Methods: In this study, our main objective was twofold: (1) to develop a clinical decision support tool for predicting the prognosis of breast cancer and (2) to identify and explore the key factors that influence breast cancer recurrence. To achieve this, we employed an explainable ensemble learning method called Shapley additive explanation (SHAP), which leverages cooperative game theory. Using real-world data from 1629 breast cancer patients, we analyzed and uncovered the key factors associated with breast cancer recurrence. Subsequently, we used these identified factors to create a recurrence prediction model and establish a decision mechanism for the tool. The proposed method not only provides accurate recurrence predictions but also offers transparent explanations for these predictions.

Results: By utilizing four key factors, namely, tumor size, clinical stage III, number of lymph node metastases, and age, our decision support tool for predicting breast cancer recurrence achieved significant improvements. The extra-tree model exhibited an increased area under the receiver operating characteristic curve (AUC) of 0.97, while the Random Forest model demonstrated an improved AUC of 0.96. We also offer a decision mechanism for a recurrence prediction model based on the identified key factors. This transparent and interpretable decision-making process facilitated by our explainable ensemble learning model enhances trust and promotes its applicability in clinical settings.

Conclusions: The proposed explainable ensemble learning method shows promising results in predicting breast cancer recurrence, outperforming existing methods with high accuracy and transparency. This advancement has the potential to significantly improve clinical decision-making and patient outcomes in breast cancer treatment.

Keywords: Breast cancer; Cancer recurrence; Explainable ensemble learning; SHAP.

Fig. 1

Overview of study design.

Fig. 2

ROC curves of five ensemble learning models to predict breast cancer recurrence. Red, orange, green, blue, and purple represent the five models of AdaBoost, Bagging, Extra Trees, Gradient Boosting, and Random Forest, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

Comparison of five ensemble learning models in predicting breast cancer recurrence. ns: p≤1.00e+00, *: 1.00e-02 < p≤5.00e-02, **: 1.00e-03 < p≤1.00e-02, ***: 1.00e-04 < p≤1.00e-03, ****: p≤1.00e-04.

Fig. 4

Key features identified using the SHAP value analysis are represented by the colors red and blue, indicating labels 0 and 1, respectively. The horizontal axis represents the average SHAP value, while the vertical axis represents different clinical features. In this context, the term “Menstruation" refers to the age of menarche. “Basel-like" represents the Basel Type in molecular subtyping. “Total lymph nodes" refers to the total number of metastatic lymph nodes. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

Visualization of decision tree for predicting breast cancer recurrence using key clinical features. Yellow represents the non-recurrence patients, and green represents the recurrence population. Clinical state III, tumor size, age, and total lymph nodes are the reference indicators for decision-making. An arrow with ‘≤’ denotes that the classification is below the threshold of random classification decision, while an arrow with ‘>’ indicates it is above the threshold of classification decision. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

figs1

figs2