Investigating the Role of Image Fusion in Brain Tumor Classification Models Based on Machine Learning Algorithm for Personalized Medicine

Abstract

Image fusion can be performed on images either in spatial domain or frequency domain methods. Frequency domain methods will be most preferred because these methods can improve the quality of edges in an image. In image fusion, the resultant fused images will be more informative than individual input images, thus more suitable for classification problems. Artificial intelligence (AI) algorithms play a significant role in improving patient’s treatment in the health care industry and thus improving personalized medicine. This research work analyses the role of image fusion in an improved brain tumour classification model, and this novel fusion-based cancer classification model can be used for personalized medicine more effectively. Image fusion can improve the quality of resultant images and thus improve the result of classifiers. Instead of using individual input images, the high-quality fused images will provide better classification results. Initially, the contrast limited adaptive histogram equalization technique preprocess input images such as MRI and SPECT images. Benign and malignant class brain tumor images are applied with discrete cosine transform-based fusion method to obtain fused images. AI algorithms such as support vector machine classifier, KNN classifier, and decision tree classifiers are tested with features obtained from fused images and compared with the result obtained from individual input images. Performances of classifiers are measured using the parameters accuracy, precision, recall, specificity, and F 1 score. SVM classifier provided the maximum accuracy of 96.8%, precision of 95%, recall of 94%, specificity of 93%, F 1 score of 91%, and performed better than KNN and decision tree classifiers when extracted features from fused images are used. The proposed method results are compared with existing methods and provide satisfactory results.

Prediction of the Age and Gender Based on Human Face Images Based on Deep Learning Algorithm

“…Feature selection is then proceeded by the classification process which classifies the face image in terms of age and gender. Age is categorized into 8 classes as (0-2), (4-6), (8)(9)(10)(11)(12), (15)(16)(17)(18)(19)(20), (25)(26)(27)(28)(29)(30)(31)(32), (38)(39)(40)(41)(42)(43), (48)(49)(50)(51)(52)(53), and (60-100), and gender is categorized as male/female. Figure 7 shows the result of classification.…”

“…The initial deep learning technology utilized in a ML algorithm [41][42][43] was the deep neural network (DNN) [44,45]. However, DNN has an overfitting problem and takes much too long to train.…”

High-Performance Method for Brain Tumor Feature Extraction in MRI Using Complex Network

“…Those features were obtained from segmented tumors using U-Net and then fed to machine learning methods for tumor classification. The obtained experimental results [9] 95.00 Ansari et al [10] 98.91 Li et al [11] 88.00 Alves et al [12] 76.60 Kang et al [13] 98.50 Jena et al [14] 94.25 Nanmaran et al [15] 96.8 Susanto et al [16] 98.65 Aamir et al [17] 98.98 This study 99.69 kNN Alves et al [12] 80.60 Kang et al [13] 98.50 Jena et al [14] 87.88 Nanmaran et al [15] 91 GLCM 5 features 95.00 Ansari et al [10] GLCM and DWT 12 features 98.91 Li et al [11] Gabor transform, texture, and DWT 80 best-ranked features 88.00 Alves et al [12] Genetic algorithm, GLCM, GLRL, and DWT Five best-ranked features 82.70 Kang et al [13] CNN Top-3 deep features 98.50 Jena et al [14] Genetic algorithm, GLCM, GLRL, and DWT 471 features 97 Han et al [22] Complex network + wavelet transform 4 features 93.06 Nanmaran et al [15] Discrete cosine transform 6 features 96.8 Susanto et al [16] GLCM and DWT Applied Bionics and Biomechanics…”

Human Centered Decision-Making for COVID-19 Testing Center Location Selection: Tamil Nadu—A Case Study

“…Li et al [33] created a TOPSIS model for determining the site of a logistics center based on five criteria: traffic, communication, candidate land area, candidate land value, and freight transportation. Nanmaran et al [34] proposed a model that combined the analytic network process (ANP), TOPSIS, and DEMATEL approaches to determine the location of an international distribution center based on criteria such as location resistance, extension transportation, port rate, one-stop service, import and export volume, convenience, transshipment time, port operation system, information abilities, and port and warehouse facilitation. Using this method, decision-makers cannot select candidate locations at the same time [35][36][37].…”