Lung Cancer Survival Prediction Using Machine Learning and Statistical Methods
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911188564049920 |
|---|---|
| author | Nair, Varun Vishwanathan Soberanis, Victor Miranda |
| author_facet | Nair, Varun Vishwanathan Soberanis, Victor Miranda |
| contents | Lung cancer remains one of the leading causes of cancer-related mortality, yet most survival models rely only on baseline factors and overlook posttreatment variables that reflect disease progression. To address this gap, we applied Cox Proportional Hazards and Random Survival Forests, integrating baseline features with post-treatment predictors such as progression-free interval (PFI.time) and residual tumor status. The Cox model achieved a concordance index (C-index) of 0.90, while the RSF model reached 0.86, both outperforming previous studies. Beyond statistical gains, the integration of post-treatment variables provides oncologists with more clinically meaningful and reliable survival estimates. This enables improved treatment planning, more personalized patient counseling, and better-informed follow-up strategies. From a practical standpoint, these results demonstrate how routinely collected clinical variables can be transformed into actionable survival predictions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_01267 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Lung Cancer Survival Prediction Using Machine Learning and Statistical Methods Nair, Varun Vishwanathan Soberanis, Victor Miranda Applications Lung cancer remains one of the leading causes of cancer-related mortality, yet most survival models rely only on baseline factors and overlook posttreatment variables that reflect disease progression. To address this gap, we applied Cox Proportional Hazards and Random Survival Forests, integrating baseline features with post-treatment predictors such as progression-free interval (PFI.time) and residual tumor status. The Cox model achieved a concordance index (C-index) of 0.90, while the RSF model reached 0.86, both outperforming previous studies. Beyond statistical gains, the integration of post-treatment variables provides oncologists with more clinically meaningful and reliable survival estimates. This enables improved treatment planning, more personalized patient counseling, and better-informed follow-up strategies. From a practical standpoint, these results demonstrate how routinely collected clinical variables can be transformed into actionable survival predictions. |
| title | Lung Cancer Survival Prediction Using Machine Learning and Statistical Methods |
| topic | Applications |
| url | https://arxiv.org/abs/2510.01267 |