The effect of feature normalization methods in radiomics

OBJECTIVES

In radiomics, different feature normalization methods, such as z-Score or Min-Max, are currently utilized, but their specific impact on the model is unclear. We aimed to measure their effect on the predictive performance and the feature selection.

METHODS

We employed fifteen publicly available radiomics datasets to compare seven normalization methods. Using four feature selection and classifier methods, we used cross-validation to measure the area under the curve (AUC) of the resulting models, the agreement of selected features, and the model calibration. In addition, we assessed whether normalization before cross-validation introduces bias.

RESULTS

On average, the difference between the normalization methods was relatively small, with a gain of at most + 0.012 in AUC when comparing the z-Score (mean AUC: 0.707 ± 0.102) to no normalization (mean AUC: 0.719 ± 0.107). However, on some datasets, the difference reached + 0.051. The z-Score performed best, while the tanh transformation showed the worst performance and even decreased the overall predictive performance. While quantile transformation performed, on average, slightly worse than the z-Score, it outperformed all other methods on one out of three datasets. The agreement between the features selected by different normalization methods was only mild, reaching at most 62%. Applying the normalization before cross-validation did not introduce significant bias.

CONCLUSION

The choice of the feature normalization method influenced the predictive performance but depended strongly on the dataset. It strongly impacted the set of selected features.

CRITICAL RELEVANCE STATEMENT

Feature normalization plays a crucial role in the preprocessing and influences the predictive performance and the selected features, complicating feature interpretation.

KEY POINTS

• The impact of feature normalization methods on radiomic models was measured. • Normalization methods performed similarly on average, but differed more strongly on some datasets. • Different methods led to different sets of selected features, impeding feature interpretation. • Model calibration was not largely affected by the normalization method.

Complex features extraction with deep learning model for the detection of COVID19 from CT scan images using ensemble based machine learning approach

Recently the most infectious disease is the novel Coronavirus disease (COVID 19) creates a devastating effect on public health in more than 200 countries in the world. Since the detection of COVID19 using reverse transcription-polymerase chain reaction (RT-PCR) is time-consuming and error-prone, the alternative solution of detection is Computed Tomography (CT) images. In this paper, Contrast Limited Histogram Equalization (CLAHE) was applied to CT images as a preprocessing step for enhancing the quality of the images. After that, we developed a novel Convolutional Neural Network (CNN) model that extracted 100 prominent features from a total of 2482 CT scan images. These extracted features were then deployed to various machine learning algorithms - Gaussian Naive Bayes (GNB), Support Vector Machine (SVM), Decision Tree (DT), Logistic Regression (LR), and Random Forest (RF). Finally, we proposed an ensemble model for the COVID19 CT image classification. We also showed various performance comparisons with the state-of-art methods. Our proposed model outperforms the state-of-art models and achieved an accuracy, precision, and recall score of 99.73%, 99.46%, and 100%, respectively.

Optimized stacking, a new method for constructing ensemble surrogate models applied to DNAPL-contaminated aquifer remediation

Surfactant-enhanced aquifer remediation (SEAR) is an appropriate method for DNAPL-contaminated aquifer remediation; However, due to the high cost of the SEAR method, finding the optimal remediation scenario is usually essential. Embedding numerical simulation models of DNAPL remediation within the optimization routines are computationally expensive, and in this situation, using surrogate models instead of numerical models is a proper alternative. Ensemble methods are also utilized to enhance the accuracy of surrogate models, and in this study, the Stacking ensemble method was applied and compared with conventional methods. First, Six machine learning methods were used as surrogate models, and various feature scaling techniques were employed, and their impact on the models' performance was evaluated. Also, Bagging and Boosting homogeneous ensemble methods were used to improve the base models' accuracy. A total of six stand-alone surrogate models and 12 homogeneous ensemble models were used as the base input models of the Stacking ensemble model. Due to the large size of the Stacking model, Bayesian hyper-parameter optimization method was used to find its optimal hyper-parameters. The results showed that the Bayesian hyper-parameter optimization method had better performance than common methods such as random search and grid search. The artificial neural network model, whose input data was scaled by the power transformer method, had the best performance with a cross-validation RMSE of 0.065. The Boosting method increased the base models' accuracy more than other homogeneous methods, and the best Boosting model had a test RMSE of 0.039. The Stacking ensemble method significantly increased the base models' accuracy and performed better than other ensemble methods. The best ensemble surrogate model constructed with Stacking had a cross-validation RMSE of 0.016. Finally, a differential evolution optimization model was used by substituting the Stacking ensemble model with the numerical model, and the optimal remediation strategy was obtained at a total cost of $ 72,706.