Utilizing machine learning algorithms for predicting risk factors for bone metastasis from right-sided colon carcinoma after complete mesocolic excision: a 10-year retrospective multicenter study
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.A machine learning model using XGBoost accurately predicts bone metastasis (BM) in right-sided colon cancer patients after complete mesocolonectomy (CME). Key risk factors include alkaline phosphatase, tumor size, and lymph node metastasis, aiding early intervention for this poor prognosis condition.
Area Of Science
- Oncology
- Computational Biology
- Surgical Oncology
Background
- Bone metastasis (BM) significantly worsens prognosis in colon cancer patients.
- Right-sided colon cancer with complete mesocolonectomy (CME) presents unique challenges.
- Identifying high-risk patients for BM is crucial for timely intervention.
Purpose Of The Study
- To develop a machine learning model for predicting BM risk in right-sided colon cancer patients undergoing CME.
- To identify key clinical and surgical variables associated with BM development.
Main Methods
- Utilized a cohort of 1,151 right-sided colon cancer patients, with 73 experiencing BM.
- Applied four machine learning algorithms: XGBoost, Random Forest, SVM, and KNN.
- Validated model performance using ROC curves, calibration curves, DCA, and SHAP analysis.
Main Results
- XGBoost demonstrated superior performance with high AUC, accuracy, sensitivity, and specificity in training and validation sets.
- External validation showed strong extrapolation capabilities (AUC=0.83).
- Significant risk factors identified: alkaline phosphatase (ALP), tumor size, invasion depth, lymph node metastasis, lung metastasis, and postoperative neutrophil-to-lymphocyte ratio (NLR).
Conclusions
- The XGBoost-based prediction model for BM in right-sided colon cancer is highly accurate and clinically valuable.
- This model can aid in identifying at-risk patients for enhanced monitoring and treatment strategies.
- The identified risk factors provide insights into BM pathogenesis and potential therapeutic targets.
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.