51. Prediction of endometrial cancer recurrence by using a novel machine learning algorithm: An Israeli gynecologic oncology group study.
- Author
-
Houri O, Gil Y, Gemer O, Helpman L, Vaknin Z, Lavie O, Arie AB, Amit A, Levy T, Namazov A, Shachar IB, Atlas I, Bruchim I, and Eitan R
- Subjects
- Female, Humans, Israel, Retrospective Studies, Machine Learning, Albumins, Neoplasm Recurrence, Local pathology, Endometrial Neoplasms diagnosis, Endometrial Neoplasms therapy, Endometrial Neoplasms pathology
- Abstract
Objectives: Endometrial cancer is the most common gynecologic malignancy in developed countries. The overall risk of recurrence is associated with traditional risk factors., Methods: Machine learning was used to predict recurrence among women who were diagnosed and treated for endometrial cancer between 2002 and 2012 at elven university-affiliated centers. The median follow-up time was 5 years. The following data were retrieved from the medical records and fed into the algorithm: age, chronic metabolic diseases, family and personal cancer history, hormone replacement therapy use, endometrial thickness, uterine polyp presence, complete blood count results, albumin, Ca-125 level, surgical staging, histology, depth of myometrial invasion, LVSI, grade, pelvic washing cytology, and adjuvant treatment. We used XGBoost algorithm, which fits the training data using decision trees, and can also rate the factors according to their influence on the prediction., Results: 1935 women were identified of whom 325 had recurrent disease. On the randomly picked samples, the specificity was 55% and the sensitivity was 98%. Our model showed an operating characteristic curve with AUC of 0.84., Conclusions: A machine learning algorithm presented promising ability to predict recurrence of endometrial cancer. The algorithm provides an opportunity to identify at-risk patients who may benefit from adjuvant therapy, tighter surveillance, and early intervention., Competing Interests: Declaration of Competing Interest The authors whose names are listed in the first page,certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript., (Copyright © 2022. Published by Elsevier Masson SAS.)
- Published
- 2022
- Full Text
- View/download PDF