Table A1.

Comparison of the earlier proposed works.

S.No.	Reference	Year	Summary	Novelty
1	Gupta et al. [78]	2022	A combined model with InceptionResNetV2 and Random Forest Tree was proposed for brain tumor classification. The model achieved 99% and 98% accuracy for the suggested tumor classification and detection models, respectively.	Two-model fusion
2	Haq E et al. [73]	2022	CNN and ML are combined to improve the accuracy of tumor segmentation and classification. The suggested technique attained the maximum classification accuracy of 98.3% between gliomas, meningiomas, and pituitary tumors.	Extracted features of CNN and ML models are fused
3	Srinivas et al. [74]	2022	Three CNN models are used in transfer learning mode for brain tumor classification: VGG16, ResNet50, and Inception-v3. The VGG16 has the best accuracy of 96% in classifying tumors as benign or malignant.	Compare three CNN models in transfer learning mode for brain tumor classification
4	Almalki et al. [75]	2022	Classified tumors using a linear machine learning classifiers (MLCs) model and a DL model. The proposed CNN with several layers (19, 22, and 25) is used to train the multiple MLCs in transfer learning to extract deep features. The accuracy of the CNN-SVM fused model was higher than that of previous MLC models. The fused model provided the highest accuracy (98%).	CNN-SVM mode fused for classification
5	Kibriya et al. [76]	2022	Suggested a new deep feature fusion-based multiclass brain tumor classification framework. Deep CNN features were extracted from transfer learning architectures such as AlexNet, GoogleNet, and ResNet18, and fused to create a single feature vector. SVM and KNN models are used as a classifier on this feature vector. The fused feature vector outperforms the individual vectors and system, achieving 99.7% highest accuracy.	Features of three CNNs are combined in a single feature vector
6	Gurunathan et al. [77]	2022	Suggested a CNN Deep net classifier for detecting brain tumors and classifying them into low and high grades. The suggested technique claims segmentation and classification accuracy of 99.4% and 99.5%, respectively.	Proposed a CNN Deep net classifier