A generalized multiple classifier system for improving computer-aided classification of breast masses in mammography

Monkey king evolution (MKE)-GA-SVM model for subtype classification of breast cancer

Objective

Recently, numerous research studies have concentrated on employing hybrid metaheuristic approaches for the analysis and diagnosis of breast cancer which motivated us to devise a computer-driven diagnostic tool that could aid in improving the precision of clinical decision-making.

Methods

In the present study, an integrated metaheuristic machine learning approach-based predictive model was developed that can classify breast cancer into subgroups using clinicopathological data acquired from tertiary care hospitals or oncological institutes.

Results

Monkey king evolution (MKE) was utilized to refine the hyperparameters of the support vector machine to achieve optimal settings, and genetic algorithm (GA) was used to choose the pertinent clinical and pathological attributes involved in classification before being applied to the support vector machine (SVM) classifier for prediction. A comparison was conducted between the results of the integrated MKE-GA-SVM model and those derived from conventional feature selection and hyperparameter tuning models such as GA-SVM, grid search-SVM, and SVM-recursive feature elimination (RFE). The effectiveness of the results was evaluated by applying the 10-fold cross-validation technique to the three multicentre datasets across all models. The integrated machine learning (ML) model achieved classification accuracies of 91.4%, 86.6%, and 75.5% across three clinicopathological breast cancer datasets, outperforming the existing models. The generated model performance was also assessed with notable metrics, namely F1-score, precision-recall curve, area under the ROC curve, mean square error and logarithmic loss.

Conclusion

Thus, the newly developed bio-inspired integrated metaheuristic model may be deployed as a surrogate diagnostic tool that allows clinicians to offer patients with better therapeutic outcomes.

Firefly-SVM predictive model for breast cancer subgroup classification with clinicopathological parameters

Background

Breast cancer is a highly predominant destructive disease among women characterised with varied tumour biology, molecular subgroups and diverse clinicopathological specifications. The potentiality of machine learning to transform complex medical data into meaningful knowledge has led to its application in breast cancer detection and prognostic evaluation.

Objective

The emergence of data-driven diagnostic model for assisting clinicians in diagnostic decision making has gained an increasing curiosity in breast cancer identification and analysis. This motivated us to develop a breast cancer data-driven model for subtype classification more accurately.

Method

In this article, we proposed a firefly-support vector machine (SVM) breast cancer predictive model that uses clinicopathological and demographic data gathered from various tertiary care cancer hospitals or oncological centres to distinguish between patients with triple-negative breast cancer (TNBC) and non-triple-negative breast cancer (non-TNBC).

Results

The results of the firefly-support vector machine (firefly-SVM) predictive model were distinguished from the traditional grid search-support vector machine (Grid-SVM) model, particle swarm optimisation-support vector machine (PSO-SVM) and genetic algorithm-support vector machine (GA-SVM) hybrid models through hyperparameter tuning. The findings show that the recommended firefly-SVM classification model outperformed other existing models in terms of prediction accuracy (93.4%, 86.6%, 69.6%) for automated SVM parameter selection. The effectiveness of the prediction model was also evaluated using well-known metrics, such as the F1-score, mean square error, area under the ROC curve, logarithmic loss and precision-recall curve.

Conclusion

Firefly-SVM predictive model may be treated as an alternate tool for breast cancer subgroup classification that would benefit the clinicians for managing the patient with proper treatment and diagnostic outcome.

Machine learning techniques for breast cancer computer aided diagnosis using different image modalities: A systematic review

BACKGROUND AND OBJECTIVE

The high incidence of breast cancer in women has increased significantly in the recent years. Physician experience of diagnosing and detecting breast cancer can be assisted by using some computerized features extraction and classification algorithms. This paper presents the conduction and results of a systematic review (SR) that aims to investigate the state of the art regarding the computer aided diagnosis/detection (CAD) systems for breast cancer.

METHODS

The SR was conducted using a comprehensive selection of scientific databases as reference sources, allowing access to diverse publications in the field. The scientific databases used are Springer Link (SL), Science Direct (SD), IEEE Xplore Digital Library, and PubMed. Inclusion and exclusion criteria were defined and applied to each retrieved work to select those of interest. From 320 studies retrieved, 154 studies were included. However, the scope of this research is limited to scientific and academic works and excludes commercial interests.

RESULTS

This survey provides a general analysis of the current status of CAD systems according to the used image modalities and the machine learning based classifiers. Potential research studies have been discussed to create a more objective and efficient CAD systems.

Performance of machine learning methods in diagnosing Parkinson's disease based on dysphonia measures

Parkinson's disease (PD) is a widespread degenerative syndrome that affects the nervous system. Its early appearing symptoms include tremor, rigidity, and vocal impairment (dysphonia). Consequently, speech indicators are important in the identification of PD based on dysphonic signs. In this regard, computer-aided-diagnosis systems based on machine learning can be useful in assisting clinicians in identifying PD patients. In this work, we evaluate the performance of machine learning based techniques for PD diagnosis based on dysphonia symptoms. Several machine learning techniques were considered and trained with a set of twenty-two voice disorder measurements to classify healthy and PD patients. These machine learning methods included linear discriminant analysis (LDA), k nearest-neighbors (k-NN), naïve Bayes (NB), regression trees (RT), radial basis function neural networks (RBFNN), support vector machine (SVM), and Mahalanobis distance classifier. We evaluated the performance of these methods by means of a tenfold cross validation protocol. Experimental results show that the SVM classifier achieved higher average performance than all other classifiers in terms of overall accuracy, G-mean, and area under the curve of the receiver operating characteristic plot. The SVM classifier achieved higher performance measures than the majority of the other classifiers also in terms of sensitivity, specificity, and F-measure statistics. The LDA, k-NN and RT achieved the highest average precision. The RBFNN method yielded the highest F-measure.; however, it performed poorly in terms of other performance metrics. Finally, t tests were performed to evaluate statistical significance of the results, confirming that the SVM outperformed most of the other classifiers on the majority of performance measures. SVM is a promising method for identifying PD patients based on classification of dysphonia measurements.