Table 3.
Research questions.
Index Questions Descriptions | ||
---|---|---|
TQ1 | What classification is employed to apply ML technology to BC diagnosis? | The systematic classification of applying ML approaches in the BC diagnosis process can include:
|
TQ2 | What evaluation criteria are usually used to evaluate the performance of ML techniques in the process of BC diagnosis? | The studied articles consider different criteria to evaluate the performance of ML techniques in the BC diagnosis process. The most important criteria to evaluate are Accuracy, Sensitivity, Specificity, Recall, F-score, and Area under the ROC Curve (AUC). Applying these criteria allows researchers to cite their results and accurately evaluate their research. This question is answered in section. 6. |
TQ3 | What clinical parameters are usually used to diagnose BC using ML techniques? | The studied articles consider various clinical parameters in diagnosing BC using ML techniques. The most clinical parameters are the Benign and malignant, Tumor grade, Tumor size, Gender, Age, and Number of tumors. Using these clinical parameters, researchers can evaluate the results of their BC diagnosis research. The answer to this inquiry is provided in Section 6. |
TQ4 | What ML techniques are usually used in the BC diagnosis process? | The articles studied use a variety of ML techniques to diagnose BC. ML techniques frequently used include CNN, SVM, Random Forest, Decision Tree, Logistic Regression, DNN, ANN, Naive Bayes, K-Nearest Neighbor, DBN, DAL, and K-means. Researchers always strive to select the best ML technique by applying different techniques. This question is answered in section. 6. |
TQ5 | What methods are typically used to diagnose BC clinically? | The articles studied use various clinical diagnosis methods in this study. Therefore, the clinical diagnosis methods data were gathered using various clinical techniques. Some of the critical methods include MRI, Mammography, Histopathological, Clinical pathology, Ultrasound, Microwave Tomography, and Radar-base. Thus, researchers continuously try to use methods that produce accurate results with the least possible consequences. This question is answered in section. 6. |
TQ6 | What are the future research directions and open perspectives for using ML technology in the BC diagnosis process? | This question aims to describe different orientations that can assist researchers in clarifying their focus areas and applying them to their research. This question is answered in Sect. 7. |