1st International and 10th National Iranian Conference on Bioinformatics
Home Accepted Papers
Machine Learning and Feature Selection to SEER Data to Novel Diagnosis Thyroid Cancer
Paper ID : 1055-ICB10
Authors:
ali abedini *1, shirim malehmir2
1Department of Bioinformatics, Segal biotechnology , Tehran, Iran.
2Department of Microbiology, Karaj Branch, Islamic azad university, Karaj, Iran.
Abstract:
In this research to create a machine learning prediction model that can be used to predict bone metastasis
in thyroid cancer. Demographic and clinicopathologic variables of thyroid cancer patients in the
Surveillance, Epidemiology, and End Results database from 2010 to 2016 were retrospectively analyzed.
On this basis, we developed a random forest algorithm model based on machine-learning. The area under
receiver operating characteristic curve (AUC), accuracy score, recall rate, and specificity are used to
evaluate and compare the prediction performance of the random forestmodel and the other model. A total
of 17,138 patients were included in the study, with 166 (0.97%) developed bone metastases. Grade, T stage,
histology, race, sex, age, and N stage were the important prediction features of bone metastasis. The random
forestmodel has better predictive performance than the other model (AUC: 0.917, accuracy: 0.904, recall
rate: 0.833, and specificity: 0.905). The random forestmodel constructed in this study could accurately
predict bone metastases in thyroid cancer patients, which may provide clinicians with more personalized
clinical decision-making recommendations. In conclusion, here, we developed a random forest prediction
model for bone metastases in thyroid cancer patients that outperformed traditional logistic regression
models. This facilitates personalized diagnosis and refined clinical decision making for bone metastasis in
thyroid cancer patients.
Keywords:
bone metastasis, machine learning, random forest, SEER.
Status : Paper Accepted (Poster Presentation)