Evaluation of risk factors and survival rates of patients with early-stage breast cancer with machine learning and traditional methods

Aug 3, 2026

Upload by Ivy

This study develops and compares prognostic prediction approaches for early-stage breast cancer by evaluating risk factors and survival probabilities after treatment. Cox proportional hazards regression (CPH) is used alongside an Accelerated Failure Time (AFT) model, then machine learning methods are applied in a multi-stage workflow. The dataset includes 697 patients treated at Marmara University Hospital between 1994 and 2009, with models assessed using C-index, 5-year survival, and 10-year survival. Results identify differing key risk factors across methods and show comparable-to-improved predictive performance, supporting more precise prognosis.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages17

Reading Time43 min

AccessFree to download

Background

Material and Methods

Data set and patient cohort

Modeling approaches (CPH, AFT, machine learning)

Results and Conclusion

Identified risk factors

Model performance (C-index, survival rates)

Journal Pre-proofs Evaluation of risk factors and survival rates of patients with early-stage breast cancer with machine learning and traditional methods Emrah Gökay Özgür, Ayşe Ülgen, Sinan Uzun, Nural Bekiroğlu PII: S1386-5056(24)00211-9 DOI: [https://doi.org/10.1016/j.ijmedinf.2024.105548](https://doi.org/10.1016/j.ijmedinf.2024.105548) Reference: IJB 105548 To appear in: International Journal of Medical Informatics Received Date: 29 May 2024 Revised Date: 4 July 2024 Accepted Date: 9 July 2024 Please cite this article as: E.G. Özgür, A. Ülgen, S. Uzun, N. Bekiroğlu, Evaluation of risk factors and survival rates of patients with early-stage breast cancer with machine learning and traditional methods, International Journal of Medical Informatics (2024), doi: [https://doi.org/10.1016/j.ijmedinf.2024.105548](https://doi.org/10.1016/j.ijmedinf.2024.105548) This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2024 Published by Elsevier B.V. Evaluation of risk factors and survival rates of patients with early-stage breast cancer with machine learning and traditional methods Emrah Gökay ÖZGÜR1, Ayşe ÜLGEN2, Sinan UZUN3, Nural BEKİROĞLU1 1Marmara University, School of Medical, Department of Biostatistics 2Girne American University, Faculty of Medicine, Department of Biostatistics 3Marmara University, Institute of Health Sciences, Department of Biostatistics Correspondence Author Asist. Prof. Emrah Gökay ÖZGÜR Marmara University, School of Medical, Department of Biostatistics, Başıbüyük mah. Başıbüyük yolu sok. Maltepe/İstanbul/Türkiye ORCID Number: 0000-0002-3966-4184 [emrahgokayozgur@gmail.com](emrahgokayozgur@gmail.com) • Identification of risk factors of patients with early stage breast cancer. • Determination of risk factors by traditional statistical methods • Determination of risk factors with machine learning algorithms • Comparison of prediction performances of traditional statistical methods and machine learning algorithms in terms of survival rate and c index values Funding No financial support is received from any institution, organization or person for this article. Disclosure statement The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability statement The data that support the findings of this study are available from the corresponding author,[E.G.O], upon reasonable request. Abstract Background: This article is aimed to make predictions in terms of prognostic factors and compare prediction methods by using Cox proportional hazards regression analysis (CPH), some machine learning techniques and Accelerated Failure Time (AFT) model for posttreatment survival probabilities according to clinical presentations and pathological information of early-stage breast cancer patients. Material and Methods: The study was carried out in three stages. In the first stage, the CPH method was applied. In the second stage, the AFT model and in the last stage, machine learning methods were applied. The data set consists of 697 breast cancer patients who applied to Marmara University Hospital oncology clinic between 01.01.1994 and 31.12.2009. The models obtained by using various parameters of the patients were compared according to the C index, 5-year survival rate and 10-year survival rate. Results and Conclusion: According

FAQs

What prediction methods are compared for early-stage breast cancer prognosis?

The study compares Cox proportional hazards regression (CPH), an Accelerated Failure Time (AFT) model, and machine learning approaches for posttreatment survival prediction.

How are model performances evaluated?

Models are compared using C-index values and survival rates at 5 years and 10 years.

Which risk factors are reported as significant across different modeling strategies?

CPH and AFT analyses highlight MetLN and age, while machine learning methods identify MetLN, age, tumor size, LV1, and extracapsular involvement as risk factors.

Aug 3, 2026

Upload by Ivy