The breast cancer detection has received a great attention in the histopathology image classification. In this paper, a thresholded wavelet transformation with deep transfer learning has devised for breast cancer detection. The breast histopathology images are enhanced using thresholded wavelet transformation. Then, a fusion based deep transfer learning has employed to perform binary classification (benign/malignant) of breast histopathology images. The proposed fusion model has evaluated on Breast Cancer Histopathological (BreakHis) dataset and achieved 97.09% on 40X magnified images of the dataset. Further, the proposed model outperforms existing state-of-the-art models and pre-trained models in vital metrics.

Environmental Sound Classification (ESC) is an area of high interest in audio technologies, such as urban sound monitoring and intelligent surveillance solutions. The variability and dynamics of environmental sounds make it challenging to classify them using traditional models. In this work, ESC is addressed using Mel-spectrogram images and a deep twodimensional convolutional neural network (2D CNN). Initially, the input sound recordings are converted into Mel-spectrogram images. Then, a deep CNN model with three convolution blocks is designed for ten-class environmental sound classification. We have utilized batch normalization and dropout techniques to improve training stability and prevent overfitting. Further, SpecAugment and label smoothing are used for robustness and improving generalization of the proposed model. The proposed approach is evaluated on the UrbanSound8K dataset and achieves a classification accuracy of 95.19% after 50 training epochs. The experimental results demonstrate that the proposed model outperforms the existing model on the UrbanSound8K dataset.

Transformer models have been greatly improved recently, which has led to their extensive use in the process of captions for images, responding to questions regarding images, and the processing of both images and words. The current models use extensive computational power, time, and many configurable parameters for handling inquiries about remote working and detecting images. In this paper, we introduced an efficient and rapid visual question answering (VQA) system for satellite images using the Vision-and-Language Transformer framework (ViLT). Compared to the past models of remote sensing VQA, our approach improves the efficiency of the training process while maintaining robust image understanding. Our model utilizes a ViLT backbone consisting of immobilized transformer components. A customized block using dense layers with ReLU6 activation has been developed for categorization and the generation of answers from low-resolution images. This approach makes it easier to adapt to VQA tasks that involve several tunable parameters. Testing of the conventional remote sensing VQA (RSVQA) dataset shows that our system performs well with an overall accuracy of 86.26%. Moreover, this proposed model has lower operational expenses than the alternatives that require comprehensive retraining.

Cardiovascular diseases (CVDs) are the leading causes of global mortality and require an early and precise diagnosis. This work presents an automated multi-class classifier for diagnosing cardiac disease from electrocardiogram (ECG) images through image processing and machine-learning techniques. The proposed framework consists of three steps, including pre-processing, feature extraction, and ensemble learning. Initially, the ECG image undergoes a comprehensive pre-processing pipeline that includes lead segmentation, grayscale conversion, Gaussian filtering, and Otsu thresholding. The contour-based features are extracted and then reduced by PCA to preserve discriminative information. Finally, multiple machine-learning models, including K-nearest neighbors (KNNs), Random Forest and support vector machines (SVMs), are ensembled using voting and stacking classifiers to improve the performance of the proposed framework. The proposed ensemble model is evaluated on a public dataset that consists of ECG images that are categorized into four classes: normal, abnormal, myocardial infarction (MI), and history of MI. The proposed ensemble model attained the highest classification accuracy of 98.06% and outperformed the existing pre-trained and state-of-the-art models.

The impact of rapid urbanisation, industrialisation, lack of awareness and lifestyle changes has resulted in an increase in the prevalence of diabetes and its complications such as diabetic macular edema (DME). It is the most common cause of blindness, characterised by an abnormal rise in the level of fluid in the macula. It affects the keenest vision in severe cases. In this paper, we propose an automatic grading of DME to help ophthalmologists diagnose the condition timely and early. The proposed system consists of an ensemble deep neural network using a residual convolution block followed by global average pooling and a dense layer. The proposed ensemble model consists of two models, namely model 1 and model 2. The proposed model has been evaluated on the publicly available benchmark ISBI IDRiD dataset. Our proposed model outperforms its competitive models in the IDRiD competition with an accuracy of 0.94. Moreover, it achieves a ROC-AUC of 0.98. Gradient-weighted Class Activation Mapping is also constructed to ensure that our proposed model is accurate.

Osteosarcoma is a highly aggressive primary bone malignancy whose diagnosis and treatment planning heavily rely on histopathology slide-level analysis. Manual analysis of histopathology slides is very time-consuming and also requires skilled pathologists to examine them. Recent advances in deep learning have demonstrated outstanding performance in digital pathology. However, most Convolutional Neural Network (CNN) approaches rely heavily on global pooling, which is unable to capture diagnostically critical areas. This study proposes an attention pooling-based patch-level classification for osteosarcoma histopathology images. The proposed model consists of two modules: The first is a hierarchical feature extractor that uses ConvNeXt tiny as its backbone for feature extraction, and the second is an attention pooling model that assigns adaptive weights to emphasize spatial importance within the deep feature maps. It allows the model to concentrate on diagnostically relevant areas. The experimental results demonstrate that the attention pooling-based feature flattening utilized in the proposed framework improves the performance. This proposed model is evaluated on an osteosarcoma tumor assessment dataset. The proposed model outperforms the existing models with an accuracy of 92.31%.

High performance computing (HPC) based visual question and answering is the contemporary research in visual question answering. This study introduces the Image Patch-Question feature fusion approach for HPC-based visual question and answering. The proposed framework divides the input image into four patches. Subsequently, each image patch, accompanied by the input question, is processed using a vision-and-language transformer utilizing an HPC cluster to produce a feature vector. Finally, the answer will be produced with a fusion-based voting algorithm that employs four feature vectors obtained from the HPC clusters. The proposed model has been assessed using the low-resolution Visual Question Answering for Remote Sensing (RSVQA) dataset. Our model demonstrates higher performance compared to the base models with 86.4% accuracy.

The classification of brain tumor images is the prevalent task in computer-aided brain tumor diagnosis. Recently, three-class classification has become a superlative task in brain tumor type classification. The existing models are fine-tuned for a single dataset, and hence, they may exhibit displeasing results on other datasets. Thus, there is a need for a generalized model that can produce superior performance on multiple datasets. In this paper, we have presented a generalized model that produces similar results on two datasets. We have proposed an EfficientNet and Mixed Convolution Network model to perform a three-class brain tumor type classification. We have devised a mixed convolution network to enhance the feature vector extracted from pre-trained EfficientNet. The proposed network consists of two blocks, namely, separable convolution and residual convolution. We have utilized a Gaussian dropout layer before the softmax layer to avoid model overfitting. In our experiments, two publicly available datasets (BTDS and CPM) are considered for the evaluation of the proposed model. The BTDS dataset has been segregated into three tumor types: Meningioma, Glioma, and Pituitary. The CPM dataset has been divided into three glioma subtypes: Glioblastoma, Oligodendroglioma, and Astrocytoma. We have achieved an accuracy of 98.04% and 96.00% on BTDS and CPM datasets, respectively. The proposed model outperforms existing pre-trained models and state-of-the-art models in vital metrics.

The development of Generative Artificial Intelligence (Generative AI) presents significant ethical and privacy issues that need to be carefully considered as it opens up previously unheard-of opportunities for creativity, automation, and problem-solving. This chapter explores the convergence of technology, society, and privacy as it digs into the complex world of ethical issues surrounding the use of generative AI. A wide range of concerns falls under the category of ethical considerations, such as the potential for social injustices to be amplified, biases present in training data, and the moral implications of content produced by artificial intelligence. This chapter explores how biases in training datasets might unintentionally reinforce societal preconceptions and provide discriminating results in AI-generated applications. Additionally, it looks at how organizations and developers can help reduce prejudice and guarantee fairness in generative AI systems. The chapter also explores the implications AI-generated content has for society, from text and images to full stories. The ethical implications of producing and sharing content that conflates real and artificial intelligence bring up issues with inaccuracy, manipulation, and potentially harmful applications. With a focus on protecting individual privacy, the chapter examines the fine line that must be drawn between fostering innovation and ensuring the welfare of society. Regarding privacy, the chapter discusses the risks posed by generative artificial intelligence, especially about safeguarding personal information. The chapter examines the possibility of privacy being exploited by creating synthetic content and highlights the necessity of strong security measures to prevent nefarious actors from compromising personal data. The chapter also encourages a proactive approach to addressing these issues by outlining developing best practices and current frameworks for ethical AI development. To create a responsible AI ecosystem that puts individual privacy first, it offers methods for incorporating explainability, accountability, and transparency into Generative AI systems. This chapter seeks to contribute to the ongoing discussion on the direction of AI development by critically analyzing the ethical and privacy issues surrounding generative AI. To fully realize the transformational potential of generative AI while avoiding unforeseen repercussions, it emphasizes the significance of an ethical foundation and strong privacy safeguards.

The brain tumor grade classification is one of the prevalent tasks in brain tumor image classification. The existing models have employed transfer learning and are unable to preserve semantic features. Moreover, the results are reported on small datasets with pre-trained models. Thus, there is a need for an optimized model that can exhibit superior performance on larger datasets. We have proposed an efficientnet and coupled convolution network for the grade classification of brain magnetic resonance images. The feature extraction is performed using a pre-trained EfficientNetB0. Then, we have proposed a coupled convolution network for feature enhancement. Finally, enhanced features are classified using a fully connected dense network. We have utilized a global average pooling and dropout layers to avoid model overfitting. We have evaluated the proposed model on the REMBRANDT dataset and have achieved 96.95% accuracy. The proposed model outperforms existing pre-trained models and state-of-the-art models in vital metrics.

Brain tumor grade detection is one of the perpetual tasks in brain image classification. Deep learning models are the most successful for multi-class classification which are trained for non-medical image classification. Thus, there is a need for re-training and feature enhancement for better performance in medical image classification. In this paper, we have proposed a residual multi-head attention network to uplift the retraining process with polished feature extraction. The proposed model consists of three parts including a pre-trained EfficientNetB4, a residual multi-head attention network, and a dense network. The residual multi-head attention network utilizes the attention block with three convolution layers for better tumor detection. The residual connection used in the network avoids the vanishing gradient problem. We have extracted a two-class (low-grade/high-grade) dataset from REMBRANDT repository. The proposed model has attained an accuracy of 96.39% and outperforms its competing models in vital metrics.

Grade classification is a challenging task in brain tumor image classification. Contemporary models employ transfer learning technique to attain better performance. The existing models ignored the semantic features of a tumor during classification decisions. Moreover, contemporary research requires an optimized model to exhibit better performance on larger datasets. Thus, we propose an EfficientNet and multi-path convolution with a multi-head attention network for the grade classification. We used a pre-trained EfficientNetB4 in the feature extraction phase. Then, a multi-path convolution with multi-head attention network performs a feature enhancement task. Finally, features obtained from the above step are classified using a fully connected double dense network. We utilize TCIA repository datasets to generate a three-class (normal/low-grade/high-grade) classification dataset. Our model achieves 98.35% accuracy and 97.32% Jaccard coefficient. The proposed model achieves superior performance than its competing models in all key metrics. Further, we achieve similar performance on a noisy dataset.

Brain cancer is one of the deadliest hazards in the world and hence tumor classification became a dominant task in brain tumor diagnosis. There is a wide range of brain tumors, and each tumor exhibits distinct properties like location, shape, size, and texture. Thus, multi-class brain magnetic resonance (MR) image classification became a trivial task. In this article, we have proposed a seven-layer convolutional neural network to address three-class brain MR image classification. We have employed separable convolution to optimize computation time. The proposed separable convolution based neural network model exhibits accuracy of 97.52% on a publicly available dataset consists of 3064 images. The proposed model has analyzed with the help of four key parameters. Our proposed model exhibits superior performance than existing methods in key parameters. Further, our model takes less training time due to sparse network consists of seven layers.

A newly emerged coronavirus disease affects the social and economical life of the world. This virus mainly infects the respiratory system and spreads with airborne communication. Several countries witness the serious consequences of the COVID-19 pandemic. Early detection of COVID-19 infection is the critical step to survive a patient from death. The chest radiography examination is the fast and cost-effective way for COVID-19 detection. Several researchers have been motivated to automate COVID-19 detection and diagnosis process using chest x-ray images. However, existing models employ deep networks and are suffering from high training time. This work presents transfer learning and residual separable convolution block for COVID-19 detection. The proposed model utilizes pre-trained MobileNet for binary image classification. The proposed residual separable convolution block has improved the performance of basic MobileNet. Two publicly available datasets COVID5K, and COVIDRD have considered for the evaluation of the proposed model. Our proposed model exhibits superior performance than existing state-of-art and pre-trained models with 99% accuracy on both datasets. We have achieved similar performance on noisy datasets. Moreover, the proposed model outperforms existing pre-trained models with less training time and competitive performance than basic MobileNet. Further, our model is suitable for mobile applications as it uses fewer parameters and lesser training time

Severity classification is the most pivotal task in Alzheimer's disease diagnosis. Detection of brain structural changes from brain MR images is crucial for Alzheimer's classification. In this paper, we have proposed a transfer learning and residual separable convolution network for the classification of Alzheimer's. The proposed network includes three separable convolution layers with two average pooling layers. An upsampling has been performed to regain its spatial resolution for the residual connection. The main intuition of separable convolution is to optimize parameters with depth-wise convolution. Similarly, the residual connection has been used to reduce the vanishing gradient problem. Finally, a three-layer fully connected dense network has been used for the four-class Alzheimer's classification. Kaggle dataset has been utilized for the experiments to report results. We have achieved an accuracy of 97.32% on the dataset with five-fold cross-validation. Our model has reported an improvement of 1% in jaccard similarity and outperforms the competing models in all vital metrics.

Three-class brain tumor classification becomes a contemporary research task due to the distinct characteristics of tumors. The existing proposals employ deep neural networks for the three-class classification. However, achieving high accuracy is still an endless challenge in brain image classification. We have proposed a deep dense inception residual network for three-class brain tumor classification. We have customized the output layer of Inception ResNet v2 with a deep dense network and a softmax layer. The deep dense network has improved the classification accuracy of the proposed model. The proposed model has been evaluated using key performance metrics on a publicly available brain tumor image dataset having 3064 images. Our proposed model outperforms the existing model with a mean accuracy of 99.69%. Further, similar performance has been obtained on noisy data.

A rapid increase in brain tumor cases mandates researchers for the automation of brain tumor detection and diagnosis. Multi-tumor brain image classification became a contemporary research task due to the diverse characteristics of tumors. Recently, deep neural networks are commonly used for medical image classification to assist neurologists. Vanishing gradient problem and overfitting are the demerits of the deep networks. In this paper, we have proposed a deep network model that uses ResNet-50 and global average pooling to resolve the vanishing gradient and overfitting problems. To evaluate the efficiency of the proposed model simulation has been carried out using a three-tumor brain magnetic resonance image dataset consisting of 3064 images. Key performance metrics have used to analyze the performance of the proposed model and its competitive models. We have achieved a mean accuracy of 97.08% and 97.48% with data augmentation and without data augmentation, respectively. Our proposed model outperforms existing models in classification accuracy.

Brain tumor image classification is one of the predominant tasks of brain image processing. The three-class brain tumor classification becomes a trivial task for researchers as each tumor exhibit distinct characteristics. Existing classification models use deep neural networks and suffer from high computational cost. We have proposed an eight-layer average-pooling convolutional neural network to address three-class brain tumor classification. The proposed model uses three convolution blocks along with a dense layer and a softmax layer. We have utilized N-adam optimizer with a sparse-categorical cross-entropy loss function to improve the learning rate. The proposed model has been evaluated using a dataset consists of 3064 brain tumor magnetic resonance images. The proposed model outperforms state-of-the-art models with 97.42% accuracy and takes lesser computation time than its competitive models.

Coronavirus disease is the latest epidemic that forced an international health emergency. It spreads mainly from person to person through airborne transmission. Community transmission has raised the number of cases over the world. Many countries have imposed compulsory face mask policies in public areas as a preventive action. Manual observation of the face mask in crowded places is a tedious task. Thus, researchers have motivated for the automation of face mask detection system. In this paper, we have presented a MobileNet with a global pooling block for face mask detection. The proposed model employs a global pooling layer to perform a flatten of the feature vector. A fully connected dense layer associated with the softmax layer has been utilized for classification. Our proposed model outperforms existing models on two publicly available face mask datasets in terms of vital performance metrics.

Brain tumour segmentation evolved as the dominant task in brain image processing. Most of the contemporary research proposals devise deep neural networks and sparse representation to address this issue. These methods inherently suffer from high computational cost and additional memory requirements. Thus, optimization of the computational cost became a challenging task for the contemporary research. This paper discusses an optimized U-Net model with post-processing for fast brain tumour segmentation. The proposed model includes two phases: training and testing. Training phase computes weights for optimized U-Net and an adaptive threshold value. In the testing phase, a trained U-Net model predicts a rough tumour segment. Adaptive thresholding grabs the final tumour with improved segmentation results. We have considered a brain tumour dataset of 3064 images with three types of brain tumours for evaluation. Our proposed model exhibits superior results than the existing models in terms of recall and dice similarity metrics. It exhibits competitive performance in accuracy and precision. Moreover, the proposed model outperforms its competitive models in training time.

Evolution of digital Health-care Information System established Medical Image Security as the new contemporary research area. Most of the researchers used either image watermarking or image encryption to address medical image security. However, very few proposals focused on both issues. This paper has implemented a Fast Medial Image Security algorithm for color images that uses both watermarking and encryption of each color channel. The proposed method starts with embedding of a smoothened key image (K) and patient information over the original image (I) to generate a watermarked image (W). Then, each color channel of the watermarked image (W) is encrypted separately to produce an encrypted image (E) using the same smoothened key image (K). This image can be transmitted over the public network and the original image (I) can be achieved using decryption algorithm followed by de-watermarking using the same key image (K) at the receiver. Qualitative and quantitative results of the proposed method show good performance when compared with the existing method with high Mean, PSNR and Entropy.

Brain tumor extraction has increased its potential in brain image processing. Extraction of tumor region from brain MRI image is still a trivial task. However, most of the contemporary proposals has devised sparse representation and convolutional neural networks to address this issue. These methods suffer from high computational cost and additional memory requirements. Threshold-based techniques are efficient for fast brain tumor extraction. In this paper, we have proposed an adaptive threshold selection approach to address it. The proposed method has implemented Adaptive Threshold Selection Network (ATSN) with two phases: training and testing with a common pre-processing step. In training phase, pre-processed train images and their ground truth images are used to achieve an adaptive threshold. Testing phase extracts tumor segment from the pre-processed test image using thresholding. We considered brain tumor dataset of 2295 images with three types of brain diseases: meningioma, pituitary and glioma tumor. Performance of proposed method has been evaluated using five essential measures: dice similarity, jaccard coefficient, accuracy, sensitivity, and specificity. Proposed method achieved superior results in terms of specified measures except accuracy and sensitivity while comparing with its competing methods.

Login will be the critical step for most of the web users for the user authentication. As hackers utilizing a variety of techniques to steal the passwords, it is recommended to offer secured transmission of passwords to the servers from the client system. Current web developers are using common traditional hashing techniques for securing passwords over the network. However, hashing techniques are vulnerable to several attacks like brute-force attack, dictionary attack, and birthday attack. This paper implemented password security with an easy and robust transposition based encryption technique using an extended ADFGVX cipher. The proposed cipher uses two 7 × 6 Polybius squares for encryption to accommodate common characters of the password along with the random key for encryption. Finally, this study considered four metrics for performance evaluation and compared the proposed method with the state-of-Art techniques. It is found that the proposed method performed excellently regarding complexity for cracking password and satisfactorily concerning execution time.