Applying Explainable Machine Learning Models for Detection of Breast Cancer Lymph Node Metastasis in Patients Eligible for Neoadjuvant Treatment

Due to recent changes in breast cancer treatment strategy, significantly more patients are treated with neoadjuvant systemic therapy (NST). Radiological methods do not precisely determine axillary lymph node status, with up to 30% of patients being misdiagnosed. Hence, supplementary methods for lymp...

Full description

Saved in:
Bibliographic Details
Published in:Cancers 2023-01, Vol.15 (3), p.634
Main Authors: Vrdoljak, Josip, Boban, Zvonimir, Barić, Domjan, Šegvić, Darko, Kumrić, Marko, Avirović, Manuela, Perić Balja, Melita, Periša, Marija Milković, Tomasović, Čedna, Tomić, Snježana, Vrdoljak, Eduard, Božić, Joško
Format: Article
Language:English
Subjects:
Algorithms
Biomarkers
Breast cancer
Cancer therapies
Care and treatment
Complications and side effects
Decision making
Diagnosis
Learning algorithms
Lymph nodes
Lymphatic metastasis
Lymphatic system
Machine learning
Mammography
Mathematical functions
Metastases
Metastasis
Methods
Neoadjuvant therapy
Optimization
Patients
Population studies
Predictions
Regression analysis
Risk factors
Tumors
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Due to recent changes in breast cancer treatment strategy, significantly more patients are treated with neoadjuvant systemic therapy (NST). Radiological methods do not precisely determine axillary lymph node status, with up to 30% of patients being misdiagnosed. Hence, supplementary methods for lymph node status assessment are needed. This study aimed to apply and evaluate machine learning models on clinicopathological data, with a focus on patients meeting NST criteria, for lymph node metastasis prediction. From the total breast cancer patient data ( = 8381), 719 patients were identified as eligible for NST. Machine learning models were applied for the NST-criteria group and the total study population. Model explainability was obtained by calculating Shapley values. In the NST-criteria group, random forest achieved the highest performance (AUC: 0.793 [0.713, 0.865]), while in the total study population, XGBoost performed the best (AUC: 0.762 [0.726, 0.795]). Shapley values identified tumor size, Ki-67, and patient age as the most important predictors. Tree-based models achieve a good performance in assessing lymph node status. Such models can lead to more accurate disease stage prediction and consecutively better treatment selection, especially for NST patients where radiological and clinical findings are often the only way of lymph node assessment.
ISSN:2072-6694
2072-6694
DOI:10.3390/cancers15030634