Table 1.

A list of related studies that used histopathological images along with their limitations.

Article	Segmentation	Features	Classifier	Accuracy	Medulloblastoma (MB) Class	Limitations
[16]	N/A	TICA 1 Wavelet analysis 2-layered convolutional neural networks (CNN)	Soft-max	99.7%	Anaplastic and Non- anaplastic	Use only one type of feature extraction either textural features or spatial features extracted from CNN. Very small dataset (10 images only)
[28]	N/A	TICA	Soft-max	97%	Anaplastic and Non- anaplastic	Depends on only texture-based feature extractors. Very small dataset (10 images only)
[12]	N/A	Haar Wavelet Transform	k-NN 2	87%	Anaplastic and Non- anaplastic	Use only one type of feature extraction (textural features) to build their computer-aided diagnosis (CADx) Very small dataset (10 images only)
[29]	N/A	Haar Haralick Laws textural features	RF 3	91%	Anaplastic and Non- anaplastic	Use an only individual type of feature extraction to build their CADx Very small dataset (6 images only)
[30]	N/A	16-layered CNN 2-Layered CNN	softmax	76.6% 89.8%	Anaplastic and Non- anaplastic	Depends on only spatial deep learning-based feature extractors. Very small dataset (10 images only)
[7]	K-means clustering	HOG 4 GLCM 5 GLRM 6 Tamura Color Feature LBP 7 Morphological Principal component analysis (PCA)	SVM	84.9%	Classic Desmoplastic Nodular Large Cell (anaplastic)	Depends only on conventional handcrafted features. They used only individual feature set to perform the classification task.
[31]	K-means clustering	HOG GLCM rGLRM Tamura Color Feature LBP Morphological MANOVA 8	SVM	65.2%	Classic Desmoplastic Nodular Large Cell (anaplastic)	Depends only on conventional handcrafted features. Used only individual feature set to perform the classification task.
[13]	K-means clustering	Different combinations of fused features including: HOG GLCM GLRM Tamura LBP PCA	SVM	96.7%	Classic Desmoplastic Nodular Large Cell (anaplastic)	Depends only on conventional handcrafted features.
[32]	N/A	AlexNet	Softmax	79.3%	Classic Desmoplastic Nodular Large Cell (anaplastic)	Depends only on spatial deep learning (DL) features.
		VGG-16 10		65.4%
[32]	N/A	AlexNet DL features	SVM	93.21%	Classic Desmoplastic Nodular Large Cell (anaplastic)	Depends only on spatial DL features. Use Individual DL for the classification task.
		VGG-16 DL features		93.38%

¹ TICA: Topographic independent component analysis, ² k-NN: k-nearest neighbors,³ RF: Random Forest, ⁴ HOG: Histogram of oriented gradients, ⁵ GLCM: grey level covariance matrix, ⁶ GLRM: grey level run matrix, ⁷ LBP: local binary pattern, ⁸ MANOVA: Multivariate analysis of variance, and ¹⁰ VGG: Visual Geometry Group.