Early cancer identification is regarded as a challenging problem in cancer prevention for the healthcare community. In addition, ensuring privacy-preserving healthcare data becomes more difficult with the growing demand for sharing these data. This study proposes a novel privacy-preserving non-invasive cancer detection method using Deep Learning (DL). Initially, the clinical data is collected over the Internet via wireless channels for diagnostic purposes. It is paramount to secure personal clinical data against eavesdropping by unauthorized users that may exploit it for personalized interests. Therefore, the collected data is encrypted before transmission over the channel to prevent data theft. Various security measures, including correlation, entropy, contrast, structural content, and energy, are used to assess the proposed encryption method's efficiency. In this paper, we proposed using the Convolutional Neural Network (CNN)-based model and Magnetic Resonance Imaging (MRI) with different techniques, including transfer learning, fine-tuning, and K-fold analysis cancer detection. Extensive experiments are carried out to evaluate the performance of the proposed DL techniques with regard to traditional machine learning approaches such as Decision Tree (DT), Naive Bayes (NB), Random Forest (RF), and Support Vector Machine (SVM). Results show that the CNN-based model has achieved an accuracy of 98.9% and outperforms conventional ML algorithms. Further experiments demonstrate the efficiency of both encryption schemes, achieving entropy, contrast, and energy of 7.9999, 10.9687, and 0.0151, respectively.

In this paper, a noise-resistant image encryption scheme is proposed. We have used a cubic-logistic map, Discrete Wavelet Transform (DWT), and bit-plane extraction method to encrypt the medical images at the bit-level rather than pixel-level. The proposed work is divided into three sections; In the first and the last section, the image is encrypted in the spatial domain. While the middle section of the proposed algorithm is devoted to the frequency domain encryption in which DWT is incorporated. As the frequency domain encryption section is a sandwich between the two spatial domain encryption sections, we called it a ”sandwich encryption.” The proposed algorithm is lossless because it can decrypt the exact pixel values of an image. Along with this, we have also gauge the proposed scheme's performance using statistical analysis such as entropy, correlation, and contrast. The entropy values of the cipher images generated from the proposed encryption scheme are more remarkable than 7.99, while correlation values are very close to zero. Furthermore, the number of pixel change rate (NPCR) and unified average change intensity (UACI) for the proposed encryption scheme is higher than 99.4% and 33, respectively. We have also tested the proposed algorithm by performing attacks such as cropping and noise attacks on enciphered images, and we found that the proposed algorithm can decrypt the plaintext image with little loss of information, but the content of the original image is visible.

With recent advancements in multimedia technologies, the security of digital data has become a critical issue. To overcome the vulnerabilities of current security protocols, researchers tend to focus their efforts on modifying existing protocols. Over the last few decades, though, several proposed encryption algorithms have been proven insecure, leading to major threats against important data. Using the most appropriate encryption algorithm is a very important means of protection against such attacks, but which algorithm is most appropriate in any particular situation will also be dependent on what sort of data is being secured. However, testing potential cryptosystems one by one to find the best option can take up an important processing time. For a fast and accurate selection of appropriate encryption algorithms, we propose a security level detection approach for image encryption algorithms by incorporating a support vector machine (SVM). In this work, we also create a dataset using standard encryption security parameters, such as entropy, contrast, homogeneity, peak signal to noise ratio, mean square error, energy, and correlation. These parameters are taken as features extracted from different cipher images. Dataset labels are divided into three categories based on their security level: strong, acceptable, and weak. To evaluate the performance of our proposed model, we have performed different analyses (f1-score, recall, precision, and accuracy), and our results demonstrate the effectiveness of this SVM-supported system.

The advancement in wireless communication has encouraged the process of data transferring through the Internet. The process of data sharing via the Internet is prone to several attacks. The sensitive information can be protected from hackers with the help of a process called Encryption. Owing to the increase in cyber-attacks, encryption has become a vital component of modern-day communication. In this article, an image encryption algorithm is suggested using dynamic substitution and chaotic systems. The suggested scheme is based upon the chaotic logistic map, chaotic sine maps and the dynamical substitution boxes (S-boxes). In the proposed scheme, the S-box selection is according to the generated sequence by deploying the chaotic sine map. To evaluate the robustness and security of the proposed encryption scheme, different security analysis like correlation analysis, information entropy, energy, histogram investigation, and mean square error are performed. The keyspace and entropy values of the enciphered images generated through the proposed encryption scheme are over 2 ²⁷⁸ and 7.99 respectively. Moreover, the correlation values are closer to zero after comparison with the other existing schemes. The unified average change intensity (UACI) and the number of pixel change rate (NPCR) for the suggested scheme are greater than 33, 99.50% respectively. The simulation outcomes and the balancing with state-of-the-art algorithms justify the security and efficiency of the suggested scheme.

A little over a year after the official announcement from the WHO, the COVID-19 pandemic has led to dramatic consequences globally. Today, millions of doses of vaccines have already been administered in several countries. However, the positive effect of these vaccines will probably be seen later than expected. In these circumstances, the rapid diagnosis of COVID-19 still remains the only way to slow the spread of this virus. However, it is difficult to predict whether a person is infected or not by COVID-19 while relying only on apparent symptoms. In this context, we propose to use machine learning (ML) algorithms in order to diagnose COVID-19 infected patients more effectively. The proposed diagnosis method takes into consideration several symptoms, such as flu symptoms, throat pain, immunity status, diarrhea, voice type, body temperature, joint pain, dry cough, vomiting, breathing problems, headache, and chest pain. Based on these symptoms that are modelled as ML features, our proposed method is able to predict the probability of contamination with the COVID-19 virus. This method is evaluated using different experimental analysis metrics such as accuracy, precision, recall, and F1-score. The obtained experimental results have shown that the proposed method can predict the presence of COVID-19 with over 97% accuracy.