Predicting outcomes of pelvic exenteration using machine learning

Abstract

Aim We aim to compare machine learning with neural network performance in predicting R0 resection (R0), length of stay > 14 days (LOS), major complication rates at 30 days postoperatively (COMP) and survival greater than 1 year (SURV) for patients having pelvic exenteration for locally advanced and recurrent rectal cancer. Method A deep learning computer was built and the programming environment was established. The PelvEx Collaborative database was used which contains anonymized data on patients who underwent pelvic exenteration for locally advanced or locally recurrent colorectal cancer between 2004 and 2014. Logistic regression, a support vector machine and an artificial neural network (ANN) were trained. Twenty per cent of the data were used as a test set for calculating prediction accuracy for R0, LOS, COMP and SURV. Model performance was measured by plotting receiver operating characteristic (ROC) curves and calculating the area under the ROC curve (AUROC). Results Machine learning models and ANNs were trained on 1147 cases. The AUROC for all outcome predictions ranged from 0.608 to 0.793 indicating modest to moderate predictive ability. The models performed best at predicting LOS > 14 days with an AUROC of 0.793 using preoperative and operative data. Visualized logistic regression model weights indicate a varying impact of variables on the outcome in question. Conclusion This paper highlights the potential for predictive modelling of large international databases. Current data allow moderate predictive ability of both complex ANNs and more classic methods.

Is preoperative sarcopenia associated with postoperative complications after pelvic exenteration surgery?

“…Consistent with this, we noted an association between operative time and postoperative complications which is likely representative of the more complex surgical resections predisposing patients to major complications [ 27 ]. A large study from the PelvEx collaborative similarly identified prolonged operative time as negatively associated with 30-day complications [ 28 ].…”

Role of 18F-PET-CT to Predict Pathological Response After Neoadjuvant Treatment of Rectal Cancer

“…There are several studied that have investigated the predictive value of 18 F -FDG PET-CT in the LARC and in recurrent rectal cancer, and more are likely to be produced using collaborative, international research platforms [18][19][20][21][22][23][24][25][26][27]. However, these studies have some limits mainly due to the methodological heterogeneity secondary to their multicentric nature in terms of preoperative studies, chemoradiotherapy, patients' characteristics and PET study method (timing, technique and analysis of images).…”

Artificial intelligence in colorectal surgery: an AI-powered systematic review

“…The application of ML algorithms resulted in a moderate ability to predict outcomes of pelvic exenteration surgery, even by using more complex ANNs [23].…”

“…The PelvEx group trained ML models and ANNs to predict LOS with modest to moderate predictive ability [23]. Jo et al [63] developed an algorithm with ML that performed well for colon cancer (AUC 0.71).…”