GLIMPSE: a glioblastoma prognostication model using ensemble learning-a surveillance, epidemiology, and end results study

New perspectives on cancer clinical research in the era of big data and machine learning

In the 21st century, the development of medical science has entered the era of big data, and machine learning has become an essential tool for mining medical big data. The establishment of the SEER database has provided a wealth of epidemiological data for cancer clinical research, and the number of studies based on SEER and machine learning has been growing in recent years. This article reviews recent research based on SEER and machine learning and finds that the current focus of such studies is primarily on the development and validation of models using machine learning algorithms, with the main directions being lymph node metastasis prediction, distant metastasis prediction, and prognosis-related research. Compared to traditional models, machine learning algorithms have the advantage of stronger adaptability, but also suffer from disadvantages such as overfitting and poor interpretability, which need to be weighed in practical applications. At present, machine learning algorithms, as the foundation of artificial intelligence, have just begun to emerge in the field of cancer clinical research. The future development of oncology will enter a more precise era of cancer research, characterized by larger data, higher dimensions, and more frequent information exchange. Machine learning is bound to shine brightly in this field.

Survival prediction of glioblastoma patients using modern deep learning and machine learning techniques

In this study, we utilized data from the Surveillance, Epidemiology, and End Results (SEER) database to predict the glioblastoma patients' survival outcomes. To assess dataset skewness and detect feature importance, we applied Pearson's second coefficient test of skewness and the Ordinary Least Squares method, respectively. Using two sampling strategies, holdout and five-fold cross-validation, we developed five machine learning (ML) models alongside a feed-forward deep neural network (DNN) for the multiclass classification and regression prediction of glioblastoma patient survival. After balancing the classification and regression datasets, we obtained 46,340 and 28,573 samples, respectively. Shapley additive explanations (SHAP) were then used to explain the decision-making process of the best model. In both classification and regression tasks, as well as across holdout and cross-validation sampling strategies, the DNN consistently outperformed the ML models. Notably, the accuracy were 90.25% and 90.22% for holdout and five-fold cross-validation, respectively, while the corresponding R2 values were 0.6565 and 0.6622. SHAP analysis revealed the importance of age at diagnosis as the most influential feature in the DNN's survival predictions. These findings suggest that the DNN holds promise as a practical auxiliary tool for clinicians, aiding them in optimal decision-making concerning the treatment and care trajectories for glioblastoma patients.

Predicting Survival in Glioblastoma Using Gene Expression Databases: A Neural Network Analysis

Glioblastoma (GBM) is the most aggressive and lethal brain tumor. Artificial neural networks (ANNs) have the potential to make accurate predictions and improve decision making. The aim of this study was to create an ANN model to predict 15-month survival in GBM patients according to gene expression databases. Genomic data of GBM were downloaded from the CGGA, TCGA, MYO, and CPTAC. Logistic regression (LR) and ANN model were used. Age, gender, IDH wild-type/mutant and the 31 most important genes from our previous study, were determined as input factors for the established ANN model. 15-month survival time was used to evaluate the results. The normalized importance scores of each covariate were calculated using the selected ANN model. The area under a receiver operating characteristic (ROC) curve (AUC), Hosmer-Lemeshow (H-L) statistic and accuracy of prediction were measured to evaluate the two models. SPSS 26 was utilized. A total of 551 patients (61% male, mean age 55.5 ± 13.3 years) patients were divided into training, testing, and validation datasets of 441, 55 and 55 patients, respectively. The main candidate genes found were: FN1, ICAM1, MYD88, IL10, and CCL2 with the ANN model; and MMP9, MYD88, and CDK4 with LR model. The AUCs were 0.71 for the LR and 0.81 for the ANN analysis. Compared to the LR model, the ANN model showed better results: Accuracy rate, 83.3 %; H-L statistic, 6.5 %; and AUC, 0.81 % of patients. The findings show that ANNs can accurately predict the 15-month survival in GBM patients and contribute to precise medical treatment.

Joint modeling strategy for using electronic medical records data to build machine learning models: an example of intracerebral hemorrhage

Background

Outliers and class imbalance in medical data could affect the accuracy of machine learning models. For physicians who want to apply predictive models, how to use the data at hand to build a model and what model to choose are very thorny problems. Therefore, it is necessary to consider outliers, imbalanced data, model selection, and parameter tuning when modeling.

Methods

This study used a joint modeling strategy consisting of: outlier detection and removal, data balancing, model fitting and prediction, performance evaluation. We collected medical record data for all ICH patients with admissions in 2017–2019 from Sichuan Province. Clinical and radiological variables were used to construct models to predict mortality outcomes 90 days after discharge. We used stacking ensemble learning to combine logistic regression (LR), random forest (RF), artificial neural network (ANN), support vector machine (SVM), and k-nearest neighbors (KNN) models. Accuracy, sensitivity, specificity, AUC, precision, and F1 score were used to evaluate model performance. Finally, we compared all 84 combinations of the joint modeling strategy, including training set with and without cross-validated committees filter (CVCF), five resampling techniques (random under-sampling (RUS), random over-sampling (ROS), adaptive synthetic sampling (ADASYN), Borderline synthetic minority oversampling technique (Borderline SMOTE), synthetic minority oversampling technique and edited nearest neighbor (SMOTEENN)) and no resampling, seven models (LR, RF, ANN, SVM, KNN, Stacking, AdaBoost).

Results

Among 4207 patients with ICH, 2909 (69.15%) survived 90 days after discharge, and 1298 (30.85%) died within 90 days after discharge. The performance of all models improved with removing outliers by CVCF except sensitivity. For data balancing processing, the performance of training set without resampling was better than that of training set with resampling in terms of accuracy, specificity, and precision. And the AUC of ROS was the best. For seven models, the average accuracy, specificity, AUC, and precision of RF were the highest. Stacking performed best in F1 score. Among all 84 combinations of joint modeling strategy, eight combinations performed best in terms of accuracy (0.816). For sensitivity, the best performance was SMOTEENN + Stacking (0.662). For specificity, the best performance was CVCF + KNN (0.987). Stacking and AdaBoost had the best performances in AUC (0.756) and F1 score (0.602), respectively. For precision, the best performance was CVCF + SVM (0.938).

Conclusion

This study proposed a joint modeling strategy including outlier detection and removal, data balancing, model fitting and prediction, performance evaluation, in order to provide a reference for physicians and researchers who want to build their own models. This study illustrated the importance of outlier detection and removal for machine learning and showed that ensemble learning might be a good modeling strategy. Due to the low imbalanced ratio (IR, the ratio of majority class and minority class) in this study, we did not find any improvement in models with resampling in terms of accuracy, specificity, and precision, while ROS performed best on AUC.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12911-022-02018-x.

AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?

Radiomic models outperform clinical data for outcome prediction in high-grade gliomas (HGG). However, lack of parameter standardization limits clinical applications. Many machine learning (ML) radiomic models employ single classifiers rather than ensemble learning, which is known to boost performance, and comparative analyses are lacking in the literature. We aimed to compare ML classifiers to predict clinically relevant tasks for HGG: overall survival (OS), isocitrate dehydrogenase (IDH) mutation, O-6-methylguanine-DNA-methyltransferase (MGMT) promoter methylation, epidermal growth factor receptor vIII (EGFR) amplification, and Ki-67 expression, based on radiomic features from conventional and advanced magnetic resonance imaging (MRI). Our objective was to identify the best algorithm for each task. One hundred fifty-six adult patients with pathologic diagnosis of HGG were included. Three tumoral regions were manually segmented: contrast-enhancing tumor, necrosis, and non-enhancing tumor. Radiomic features were extracted with a custom version of Pyradiomics and selected through Boruta algorithm. A Grid Search algorithm was applied when computing ten times K-fold cross-validation (K=10) to get the highest mean and lowest spread of accuracy. Model performance was assessed as AUC-ROC curve mean values with 95% confidence intervals (CI). Extreme Gradient Boosting (xGB) obtained highest accuracy for OS (74,5%), Adaboost (AB) for IDH mutation (87.5%), MGMT methylation (70,8%), Ki-67 expression (86%), and EGFR amplification (81%). Ensemble classifiers showed the best performance across tasks. High-scoring radiomic features shed light on possible correlations between MRI and tumor histology.