Application of machine learning in breast cancer survival prediction using a multimethod approach
Problem
Predicting 5-year survival of women with breast cancer is needed to set appropriate care plans in Iran.
Why it matters
Breast cancer is increasingly prevalent in Iran, and accurate survival prediction guides treatment planning.
What the AI did
A Deep Neural Network plus 11 conventional machine learning methods (108 models total) were trained on 2644 records from two centers, with external validation.
Results
DNN trained on Shiraz dataset with all features achieved highest accuracy of 85.56% on external validation.
Impact
External validation supports the prediction accuracy claim, though clinical adoption is not stated.
Record
| Venue | Scientific Reports |
|---|---|
| Published | 2024 |
| Work type | article |
| Authors | Seyedeh Zahra Hamedi, Hassan Emami, Maryam Khayamzadeh, Reza Rabiei, Mehrad Aria, Majid Akrami, Vahid Zangouri |
| Citations | 23 (OpenAlex, at import time) |
| References | 65 |
| Model or method | Deep Neural Network (DNN) plus 11 conventional ML methods |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41598-024-81734-y |
| Open access | gold |
| Keywords | breast cancer, survival prediction, machine learning, DNN, external validation |
Importance rating
real but narrower advance
Abstract states breast cancer is one of the most prevalent cancers with increasing incidence and mortality in Iran; survival prediction guides care plans. No quantified burden figures given.