Construction and validation of a nomogram for predicting prolonged air leak after minimally invasive pulmonary resection

Interpretable prediction of cardiopulmonary complications after non-small cell lung cancer surgery based on machine learning and SHapley additive exPlanations

Introduction

Lung cancer is a prevalent malignancy globally, with approximately 20% of patients developing cardiopulmonary complications after lobectomy. In order to prevent complications, an accurate and personalized method based on machine learning (ML) is required.

Methods

During the period of 2017-2021, a retrospective analysis was conducted on the medical records of patients who had undergone lobectomy for non-small cell lung cancer (NSCLC). We performed logical regression, decision tree (DT), random forest (RF), gradient boost DT, and eXtreme gradient boosting analyses to establish an ML model. The ten-fold cross-validation was used to evaluate the performance of multiple ML models based on various evaluation metrics, including accuracy, precision, recall, F1 score, and area under the receiver operating (AUC). Additionally, we also calculated the Kappa value of these model. Each model used grid search to optimize hyper-parameters and then used the interpretability method to provide explanations for the model's Decisions.

Results

The study included 718 eligible patients, among whom the incidence of postoperative cardiopulmonary complications was 20.89%. The RF model showed the best comprehensive performance among all models, and its ten-fold cross-validation accuracy, precision, recall, F1 score, and AUC were (OR and 95% confidence interval [CI]) 0.786 (0.738-0.834), 0.803 (0.735-0.872), 0.738 (0.678-0.797), 0.766 (0.714-0.818), 0.856 (0.815-0.898), respectively. The kappa value of the RF model was 0.696 (0.617-0.768). The SHAP method showed that gender, age, and intraoperative blood loss were closely associated with postoperative cardiopulmonary complications.

Conclusion

The application of ML methods for predicting postoperative cardiopulmonary complications based on clinical data of patients with NSCLC showed a good performance. The results indicate that ML combined with the SHAP individualized interpretation method has practical clinical value.

Development and validation of a nomogram for predicting metachronous peritoneal metastasis in colorectal cancer: A retrospective study

BACKGROUND

Peritoneal metastasis (PM) after primary surgery for colorectal cancer (CRC) has the worst prognosis. Prediction and early detection of metachronous PM (m-PM) have an important role in improving postoperative prognosis of CRC. However, commonly used imaging methods have limited sensitivity to detect PM early. We aimed to establish a nomogram model to evaluate the individual probability of m-PM to facilitate early interventions for high-risk patients.

AIM

To establish and validate a nomogram model for predicting the occurrence of m-PM in CRC within 3 years after surgery.

METHODS

We used the clinical data of 878 patients at the Second Hospital of Jilin University, between January 1, 2014 and January 31, 2019. The patients were randomly divided into training and validation cohorts at a ratio of 2:1. The least absolute shrinkage and selection operator (LASSO) regression was performed to identify the variables with nonzero coefficients to predict the risk of m-PM. Multivariate logistic regression was used to verify the selected variables and to develop the predictive nomogram model. Harrell’s concordance index, receiver operating characteristic curve, Brier score, and decision curve analysis (DCA) were used to evaluate discrimination, distinctiveness, validity, and clinical utility of this nomogram model. The model was verified internally using bootstrapping method and verified externally using validation cohort.

RESULTS

LASSO regression analysis identified six potential risk factors with nonzero coefficients. Multivariate logistic regression confirmed the risk factors to be independent. Based on the results of two regression analyses, a nomogram model was established. The nomogram included six predictors: Tumor site, histological type, pathological T stage, carbohydrate antigen 125, v-raf murine sarcoma viral oncogene homolog B mutation and microsatellite instability status. The model achieved good predictive accuracy on both the training and validation datasets. The C-index, area under the curve, and Brier scores were 0.796, 0.796 [95% confidence interval (CI) 0.735-0.856], and 0.081 for the training cohort and 0.782, 0.782 (95%CI 0.690-0.874), and 0.089 for the validation cohort, respectively. DCA showed that when the threshold probability was between 0.01 and 0.90, using this model to predict m-PM achieved a net clinical benefit.

CONCLUSION

We have established and validated a nomogram model to predict m-PM in patients undergoing curative surgery, which shows good discrimination and high accuracy.