Application of Artificial Neural Networks to Predict Prolonged Operative Timing during Laparoscopic Colorectal Cancer Surgery

Abstract

Aim: Prolonged operative timing is likely to negatively impact clinical outcomes and accurate preoperative prediction of those likely to undergo longer procedures can assist theatre planning and postoperative care. We aimed to apply artificial neural networks (ANN) as a predictive tool for prolonged operating time in laparoscopic colorectal surgery. Methods: A dedicated, prospectively populated database of elective laparoscopic colorectal cancer surgery with curative intent was utilised. Primary endpoint was the prediction of operative time. Variables included in the network were: age, gender, ASA, BMI, stage, location of cancer, and neoadjuvant therapy. A multi-layered perceptron ANN (MLPNN) model was trained and tested alongside unit and multivariate analyses. Results: Data from 554 patients were included. 400 (72.2%) were used for ANN training and 154 (27.8%) to test predictive accuracy. 59.3% male, mean age 70 years, and BMI of 26. 161 (29%) were ASA III. 261 (47%) had rectal cancer and 8.5% underwent neoadjuvant treatment. Mean operative time was 218 minutes (95% CI 210-226) with 436 (78.7%) of less than 5 hours and 16% conversion rate. ANN accurately identified and predicted operative timing overall 87%, and those having surgery less than 5 hours with an accuracy of 93.3%; AUC 0.843 and 93.3%. The ANN findings were accurately cross-validated with a logistic regression model. Conclusion: Artificial neural network using patient demographic and tumour data successfully predicted the timing of surgery and the likelihood of prolonged laparoscopic procedures. This finding could assist the personalisation of peri-operative care to enhance the efficiency of theatre utilisation.