2020
DOI: 10.1002/bjs.11461
|View full text |Cite
|
Sign up to set email alerts
|

Machine learning to predict early recurrence after oesophageal cancer surgery

Abstract: Background Early cancer recurrence after oesophagectomy is a common problem, with an incidence of 20–30 per cent despite the widespread use of neoadjuvant treatment. Quantification of this risk is difficult and existing models perform poorly. This study aimed to develop a predictive model for early recurrence after surgery for oesophageal adenocarcinoma using a large multinational cohort and machine learning approaches. Methods Consecutive patients who underwent oesophagectomy for adenocarcinoma and had neoadj… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

1
18
0

Year Published

2021
2021
2022
2022

Publication Types

Select...
10

Relationship

0
10

Authors

Journals

citations
Cited by 42 publications
(19 citation statements)
references
References 45 publications
1
18
0
Order By: Relevance
“…However, after multivariate adjustment, it was only the presence of lymph node metastasis that was an independent predictor of OS (HR: 3.36; 95% CI, 1.70 to 6.63). Interestingly, a machine learning model developed to predict the risk of recurrence following neoadjuvant treatment found that the number of lymph node metastases was the most important variable in the model, with lymphovascular invasion second [35]. This is similar to our findings as we demonstrate that nodal status has equivalent prognostic value compared to pT and TRG in patients treated with both chemoradiotherapy and perioperative chemotherapy.…”
Section: Discussionsupporting
confidence: 89%
“…However, after multivariate adjustment, it was only the presence of lymph node metastasis that was an independent predictor of OS (HR: 3.36; 95% CI, 1.70 to 6.63). Interestingly, a machine learning model developed to predict the risk of recurrence following neoadjuvant treatment found that the number of lymph node metastases was the most important variable in the model, with lymphovascular invasion second [35]. This is similar to our findings as we demonstrate that nodal status has equivalent prognostic value compared to pT and TRG in patients treated with both chemoradiotherapy and perioperative chemotherapy.…”
Section: Discussionsupporting
confidence: 89%
“…For example, in oncology, research has demonstrated that ML applications can be of great help for the diagnosis or detection of cancer. 42 - 44 In cardiology, AI techniques are capable of reading electrocardiograms, and by integration with electronic medical records of patients, heart failure can be detected early on with reduced mortality as outcome. 45 47 In anesthesiology, ANNs are used to monitor the depth of anesthesia, and ML techniques are able to predict hypotension during surgery.…”
Section: Discussionmentioning
confidence: 99%
“…We also have the advantage of analyzing the largest number of CTS hands compared with the aforementioned studies. XGB is superior to other ML models for building prediction models based on regression or classification 25,26 . Here, multi-class classification suggested that the XGB model had the highest accuracy.…”
Section: Discussionmentioning
confidence: 99%