Machine learning to predict curative multidisciplinary team treatment decisions in oesophageal cancer
Machine learning to predict curative multidisciplinary team treatment decisions in oesophageal cancer
Background: Rising workflow pressures within the oesophageal cancer (OC) multidisciplinary team (MDT) can lead to variability in decision-making, and health inequality. Machine learning (ML) offers a potential automated data-driven approach to address inconsistency and standardize care. The aim of this experimental pilot study was to develop ML models able to predict curative OC MDT treatment decisions and determine the relative importance of underlying decision-critical variables. Methods: Retrospective complete-case analysis of oesophagectomy patients ± neoadjuvant chemotherapy (NACT) or chemoradiotherapy (NACRT) between 2010 and 2020. Established ML algorithms (Multinomial Logistic regression (MLR), Random Forests (RF), Extreme Gradient Boosting (XGB)) and Decision Tree (DT) were used to train models predicting OC MDT treatment decisions: surgery (S), NACT + S or NACRT + S. Performance metrics included Area Under the Curve (AUC), Accuracy, Kappa, LogLoss, F1 and Precision -Recall AUC. Variable importance was calculated for each model. Results: We identified 399 cases with a male-to-female ratio of 3.6:1 and median age of 66.1yrs (range 32–83). MLR outperformed RF, XGB and DT across performance metrics (mean AUC of 0.793 [±0.045] vs 0.757 [±0.068], 0.740 [±0.042], and 0.709 [±0.021] respectively). Variable importance analysis identified age as a major factor in the decision to offer surgery alone or NACT + S across models (p < 0.05). Conclusions: ML techniques can use limited feature-sets to predict curative UGI MDT treatment decisions. Explainable Artificial Intelligence methods provide insight into decision-critical variables, highlighting underlying subconscious biases in cancer care decision-making. Such models may allow prioritization of caseload, improve efficiency, and offer data-driven decision-assistance to MDTs in the future.
Artificial intelligence, Machine learning, Oesophageal cancer multidisciplinary team
Thavanesan, Navamayooran
94f0a216-9131-431b-809c-f5a83146343d
Bodala, Indu
aa030b32-7159-4bc7-beb4-50df4ec84944
Walters, Zoe
e1ccd35d-63a9-4951-a5da-59122193740d
Ramchurn, Sarvapali
1d62ae2a-a498-444e-912d-a6082d3aaea3
Underwood, Timothy J.
8e81bf60-edd2-4b0e-8324-3068c95ea1c6
Vigneswaran, Ganesh
4e3865ad-1a15-4a27-b810-55348e7baceb
November 2023
Thavanesan, Navamayooran
94f0a216-9131-431b-809c-f5a83146343d
Bodala, Indu
aa030b32-7159-4bc7-beb4-50df4ec84944
Walters, Zoe
e1ccd35d-63a9-4951-a5da-59122193740d
Ramchurn, Sarvapali
1d62ae2a-a498-444e-912d-a6082d3aaea3
Underwood, Timothy J.
8e81bf60-edd2-4b0e-8324-3068c95ea1c6
Vigneswaran, Ganesh
4e3865ad-1a15-4a27-b810-55348e7baceb
Thavanesan, Navamayooran, Bodala, Indu, Walters, Zoe, Ramchurn, Sarvapali, Underwood, Timothy J. and Vigneswaran, Ganesh
(2023)
Machine learning to predict curative multidisciplinary team treatment decisions in oesophageal cancer.
European Journal of Surgical Oncology, 49 (11), [106986].
(doi:10.1016/j.ejso.2023.106986).
Abstract
Background: Rising workflow pressures within the oesophageal cancer (OC) multidisciplinary team (MDT) can lead to variability in decision-making, and health inequality. Machine learning (ML) offers a potential automated data-driven approach to address inconsistency and standardize care. The aim of this experimental pilot study was to develop ML models able to predict curative OC MDT treatment decisions and determine the relative importance of underlying decision-critical variables. Methods: Retrospective complete-case analysis of oesophagectomy patients ± neoadjuvant chemotherapy (NACT) or chemoradiotherapy (NACRT) between 2010 and 2020. Established ML algorithms (Multinomial Logistic regression (MLR), Random Forests (RF), Extreme Gradient Boosting (XGB)) and Decision Tree (DT) were used to train models predicting OC MDT treatment decisions: surgery (S), NACT + S or NACRT + S. Performance metrics included Area Under the Curve (AUC), Accuracy, Kappa, LogLoss, F1 and Precision -Recall AUC. Variable importance was calculated for each model. Results: We identified 399 cases with a male-to-female ratio of 3.6:1 and median age of 66.1yrs (range 32–83). MLR outperformed RF, XGB and DT across performance metrics (mean AUC of 0.793 [±0.045] vs 0.757 [±0.068], 0.740 [±0.042], and 0.709 [±0.021] respectively). Variable importance analysis identified age as a major factor in the decision to offer surgery alone or NACT + S across models (p < 0.05). Conclusions: ML techniques can use limited feature-sets to predict curative UGI MDT treatment decisions. Explainable Artificial Intelligence methods provide insight into decision-critical variables, highlighting underlying subconscious biases in cancer care decision-making. Such models may allow prioritization of caseload, improve efficiency, and offer data-driven decision-assistance to MDTs in the future.
Text
manuscript revised
- Accepted Manuscript
Text
1-s2.0-S0748798323006121-main
- Proof
More information
Accepted/In Press date: 11 July 2023
e-pub ahead of print date: 13 July 2023
Published date: November 2023
Additional Information:
Publisher Copyright:
© 2023 The Author(s)
Keywords:
Artificial intelligence, Machine learning, Oesophageal cancer multidisciplinary team
Identifiers
Local EPrints ID: 479497
URI: http://eprints.soton.ac.uk/id/eprint/479497
ISSN: 0748-7983
PURE UUID: eca9fccc-c878-4a73-9db4-9b0a5ac52113
Catalogue record
Date deposited: 25 Jul 2023 16:47
Last modified: 08 Aug 2024 02:09
Export record
Altmetrics
Contributors
Author:
Navamayooran Thavanesan
Author:
Indu Bodala
Author:
Sarvapali Ramchurn
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics