Related Experiment Video
Updated: May 31, 2025
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
Machine Learning-Based Radiomics Analysis for Identifying KRAS Mutations in Non-Small-Cell Lung Cancer from CT
Mirjam Schöneck1, Nicolas Rehbach1, Lars Lotter-Becker1
1Institute for Diagnostic and Interventional Radiology, Faculty of Medicine and University Hospital Cologne, University of Cologne, 50937 Cologne, Germany.
Abstract:
Kirsten Rat Sarcoma viral oncogene homolog (KRAS) is a frequently occurring mutation in non-small-cell lung cancer (NSCLC) and influences cancer treatment and disease progression. In this study, a machine learning (ML) pipeline was applied to radiomic features extracted from public and internal CT images to identify KRAS mutations in NSCLC patients. Both datasets were analyzed using parametric (t test) and non-parametric statistical tests (Mann-Whitney U test) and dimensionality reduction techniques. Afterwards, the proposed ML pipeline was applied to both datasets using a five-fold cross-validation on the training set (70/30 train/test split) before being validated on the other dataset. The results show that the radiomic features are significantly different (Mann-Whitney U test; p < 0.05) between the two datasets, despite the use of identical feature extraction methods. Model transferability is therefore difficult to achieve, which became evident during external testing (F1 score = 0.41). Oversampling, undersampling, clustering and harmonization techniques were applied to balance and harmonize the datasets, but did not improve the classification of KRAS mutation presence. In general, due to only a single moderate result (highest test F1 score = 0.67), the accuracy of KRAS prediction is not sufficient for clinical application. In future work, the complexity of KRAS mutation might be addressed by taking submutations into consideration. Larger multicentric datasets with balanced tumor stages, including multi-scanner datasets, seem to be necessary for building robust predictive models.
Insights
Machine learning struggled to predict Kirsten Rat Sarcoma viral oncogene homolog (KRAS) mutations in non-small-cell lung cancer using CT radiomics. Model transferability was poor, indicating current methods are insufficient for clinical use.
Area of Science:
- Oncology
- Radiology
- Machine Learning
Background:
- Kirsten Rat Sarcoma viral oncogene homolog (KRAS) mutations are common in non-small-cell lung cancer (NSCLC), impacting treatment and prognosis.
- Radiomic features from CT images offer potential for non-invasive mutation detection.
Purpose of the Study:
- To develop and validate a machine learning pipeline for identifying KRAS mutations in NSCLC patients using CT-derived radiomic features.
- To assess the transferability and clinical applicability of the developed machine learning model.
Main Methods:
- A machine learning pipeline was applied to radiomic features extracted from public and internal CT datasets.
- Statistical analyses (t-test, Mann-Whitney U test) and dimensionality reduction were performed.
- A five-fold cross-validation was used for training and external validation between datasets.
- Data balancing and harmonization techniques were explored to improve model performance.
Main Results:
- Significant differences in radiomic features were observed between the two datasets (p < 0.05), hindering model transferability (F1 score = 0.41).
- Data balancing and harmonization techniques did not improve KRAS mutation classification accuracy.
- The highest achieved test F1 score was 0.67, deemed insufficient for clinical application.
Conclusions:
- Current machine learning approaches using radiomic features alone are not accurate enough for predicting KRAS mutations in NSCLC.
- Model transferability remains a significant challenge due to dataset variability.
- Future research should consider KRAS submutations and utilize larger, multicentric, and multi-scanner datasets for robust predictive model development.
