Related Experiment Video
Updated: Jul 23, 2025
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
Construction of an auxiliary scoring model for myelosuppression in patients with lung cancer chemotherapy based on
Yingjun Dong1, Changqing Hu2, Jun Liu3
1Intensive Care Unit, Shanxi Provincial Cancer Hospital/Shanxi Hospital Affiliated to Cancer Hospital, Chinese Academy of Medical Sciences/Cancer Hospital Affiliated to Shanxi Medical University Taiyuan 030013, Shanxi, China.
Objective:
To construct an auxiliary scoring model for myelosuppression in patients with lung cancer undergoing chemotherapy based on a random forest algorithm, and to evaluate the predictive performance of the model.
Methods:
Patients with lung cancer who received chemotherapy in Shanxi Province Cancer Hospital from January 2019 to January 2022 were retrospectively selected as research subjects, and their general demographic information, disease-related indicators, and laboratory test results before chemotherapy were collected. Patients were divided into a training set (136 cases) and a validation set (68 cases) in a ratio of 2:1. R software was used to establish a scoring model of myelosuppression in lung cancer patients in the training set, and the receiver operating characteristic curve, accuracy, sensitivity, and balanced F-score were used in the two data sets to evaluate the predictive performance of the model.
Results:
Among the 204 lung cancer patients enrolled, 75 patients developed myelosuppression during the follow-up period after chemotherapy, with an incidence of 36.76%. The factors in the constructed random forest model were ranked in order of age (23.233), bone metastasis (21.704), chemotherapy course (19.259), Alb (13.833), and gender (11.471) according to the mean decrease accuracy. The areas under the curve of the model in the training and validation sets were 0.878 and 0.885, respectively (all P < 0.05). The predictive accuracy of the validated model was 82.35%, the sensitivity and specificity were 84.00% and 81.40%, respectively, and the balanced F-score was 77.78% (all P < 0.05).
Conclusion:
The risk assessment model for the occurrence of myelosuppression in patients with lung cancer chemotherapy based on a random forest algorithm can provide a reference for the accurate identification of high-risk patients.
