1. Machine learning driven prediction of cerebrospinal fluid rhinorrhoea following endonasal skull base surgery: A multicentre prospective observational study
- Author
-
CRANIAL Consortium, Adrito Das, Danyal Z. Khan, Danail Stoyanov, Hani J. Marcus, Soham Bandyopadhyay, Benjamin E. Schroeder, Vikesh Patel, Alice O’Donnell, Neurology and Neurosurgery Interest Group, British Neurosurgical Trainee Research Collaborative, Anastasios Giamouriadis, Pragnesh Bhatt, Bhaskar Ram, Adithya Varma, Philip Weir, Brendan Hanna, Theodore C. Hirst, Patrick McAleavey, Alessandro Paluzzi, Georgios Tsermoulas, Shahzada Ahmed, Wai Cheong Soon, Yasir Arafat Chowdhury, Suhaib Abualsaud, Shumail Mahmood, Paresh Naik, Zohra Haiderkhan, Rafid Al-Mahfoudh, Andrea Perera, Mircea Rus, Adam Williams, Charles Hand, Kumar Abhinav, Cristina Cernei, Aiman Dilnawaz, Richard Mannion, Thomas Santarius, James Tysome, Rishi Sharma, Angelos G. Kolias, Neil Donnelly, Ashwin Venkatesh, Caroline Hayhurst, Amr Mohamed, Benjamin Stew, Joseph Merola, Setthasorn Zhi Yang, Mahmoud Kamel, Mohammad Habibullah Khan, Sahibzada Abrar, Christopher Mckeon, Daniel McSweeney, Mohsen Javadpour, Peter Lacy, Daniel Murray, Elena Roman, Kismet Hossain-Ibrahim, David Bennett, Nathan McSorley, Adam Hounat, Patrick Statham, Mark Hughes, Alhafidz Hamdan, Caroline Scott, Jigi Moudgil-Joshi, Anuj Bahl, Anna Bjornson, Daniel Gatt, Nick Phillips, Neeraj Kalra, Melissa Bautista, Seerat Shirazi, Catherine E. Gilkes, Christopher P. Millward, Ahmad MS. Ali, Dimitris Paraskevopoulos, Jarnail Bal, Samir Matloob, Rhannon Lobo, Nigel Mendoza, Ramesh Nair, Arthur Dalton, Adarsh Nadig, Lucas Hernandez, Nick Thomas, Eleni Maratos, Jonathan Shapey, Sinan Al-Barazi, Asfand Baig Mirza, Mohamed Okasha, Prabhjot Singh Malhotra, Razna Ahmed, Neil L. Dorward, Joan Grieve, Parag Sayal, David Choi, Ivan Cabrilo, Hugo Layard Horsfall, Jonathan Pollock, Alireza Shoakazemi, Oscar Maccormac, Guru N K. Amirthalingam, Andrew Martin, Simon Stapleton, Florence Hogg, Daniel Richardson, Kanna Gnanalingham, Omar Pathmanaban, Daniel M. Fountain, Raj Bhalla, Cathal J. Hannan, Annabel Chadwick, Alistair Jenkins, Claire Nicholson, Syed Shumon, Mohamed Youssef, Callum Allison, Graham Dow, Iain Robertson, Laurence Johann Glancz, Murugan Sitaraman, Ashwin Kumaria, Ananyo Bagchi, Simon Cudlip, Jane Halliday, Rory J. Piper, Alexandros Boukas, Meriem Amarouche, Damjan Veljanoski, Samiul Muquit, Ellie Edlmann, Haritha Maripi, Yi Wang, Mehnaz Hossain, Andrew Alalade, Syed Maroof, Pradnya Patkar, Saurabh Sinha, Showkat Mirz, Duncan Henderson, Mohammad Saud Khan, Nijaguna Mathad, Jonathan Hempenstall, Difei Wang, Pavan Marwaha, Simon Shaw, Georgios Solomou, and Alina Shrestha
- Subjects
cerebrospinal fluid leak ,cerebrospinal fluid rhinorrhoea ,CSF ,endoscopic endonasal ,skull base surgery ,machine learning - ML ,Neoplasms. Tumors. Oncology. Including cancer and carcinogens ,RC254-282 - Abstract
BackgroundCerebrospinal fluid rhinorrhoea (CSFR) is a common complication following endonasal skull base surgery, a technique that is fundamental to the treatment of pituitary adenomas and many other skull base tumours. The CRANIAL study explored CSFR incidence and related risk factors, particularly skull base repair techniques, via a multicentre prospective observational study. We sought to use machine learning to leverage this complex multicentre dataset for CSFR prediction and risk factor analysis.MethodsA dataset of 865 cases - 725 transsphenoidal approach (TSA) and 140 expanded endonasal approach (EEA) - with cerebrospinal fluid rhinorrhoea as the primary outcome, was used. Relevant variables were extracted from the data, and prediction variables were divided into two categories, preoperative risk factors; and repair techniques, with 6 and 11 variables respectively. Three types of machine learning models were developed in order to predict CSFR: logistic regression (LR); decision tree (DT); and neural network (NN). Models were validated using 5-fold cross-validation, compared via their area under the curve (AUC) evaluation metric, and key prediction variables were identified using their Shapley additive explanations (SHAP) score.ResultsCSFR rates were 3.9% (28/725) for the transsphenoidal approach and 7.1% (10/140) for the expanded endonasal approach. NNs outperformed LR and DT for CSFR prediction, with a mean AUC of 0.80 (0.70-0.90) for TSA and 0.78 (0.60-0.96) for EEA, when all risk factor and intraoperative repair data were integrated into the model. The presence of intraoperative CSF leak was the most prominent risk factor for CSFR. Elevated BMI and revision surgery were also associated with CSFR for the transsphenoidal approach. CSF diversion and gasket sealing appear to be strong predictors of the absence of CSFR for both approaches.ConclusionNeural networks are effective at predicting CSFR and uncovering key CSFR predictors in patients following endonasal skull base surgery, outperforming traditional statistical methods. These models will be improved further with larger and more granular datasets, improved NN architecture, and external validation. In the future, such predictive models could be used to assist surgical decision-making and support more individualised patient counselling.
- Published
- 2023
- Full Text
- View/download PDF