Osteoporosis is a widespread metabolic bone disorder characterized by reduced bone mineral density and increased fracture risk, particularly in older adults and postmenopausal women. Early and accurate diagnosis is essential to prevent severe complications. This study proposes a novel osteoporosis detection framework using knee X-ray data, integrating a mathematical feature extraction technique based on the Whirling Triangle. This method captures structural and spatial patterns through geometric relationships such as angles and triangle areas, enhancing feature representation by reducing noise and dimensionality. The extracted features are fused and processed through a hybrid classification architecture comprising an autoencoder and attention-based three Multi-Layer Perceptrons (MLPs) with varying structures. This ensemble network categorizes images into normal, osteopenic, or osteoporotic classes. Extensive experiments with different dataset splits were conducted to validate the model’s generalizability. Comparative analysis with several frontline technique of Deep Learning and Transfer Learning models demonstrates the accuracy and robustness of the proposed approach.

Accurate recognition of uncommon neurological disorders such as rare subtypes of stroke remains a challenging critical problem because of the lack of annotated data and complex multimodal imaging inputs. This is a new type of nested self-supervised learning (SSL) algorithm that is designed to make use of multimodal imaging data included in the Multimodal Stroke Image Dataset, which contains both CT and MRI scans in pairs. The nested SSL approach adopts multiple synergistic tasks such as cross modal reconstruction, contrastive learning and modality specific feature extraction to learn robust modality invariant features without the need of large amount of labels. By capturing the spatial, anatomical features, and contextual features of the input space in a hierarchical manner through volumetric data, the model solves problems of data scarcity and domain-aware problems in the detection of rare stroke subtypes. After pretraining, the framework is trained for classification and localization of rare stroke conditions and proved to be sensitive and diagnostic when compared with single modality or supervised baseline. Quantitative assessment through Dice scores, AUC-ROC and sensitivity scores verify the usefulness of the multi-task SSL based method for boosting the rare disease detection. This work highlights the potential of nested SSL architectures for the advancement of multi-fundamental medical image analysis and opens the path towards further personalization and even earlier diagnosis for medical imaging in the clinical set up for neuro imaging.

A standard Video Surveillance System (VSS) is a well-equipped network infrastructure that enables surveillance cameras to record incidents from deployed locations. Subsequently, the surveillance cameras capture and store the video data permanently in designated repositories for future access. It is necessary to take essential security measures to store surveillance video data securely for such network infrastructure and allow only legitimate users to access the stored video data from repositories. Several researchers have proposed some secure VSSs. However, these models often have limitations, including the high-volume storage of surveillance data and the high communication/computation overhead during transmission. Furthermore, the VSS is vulnerable to various security threats. In this paper, we present a cloud-based VSS utilizing ASCON cryptography to support a secure and scalable storage framework for surveillance video data. The lightweight ASCON cryptography is used to achieve secure authentication and access of both live and archived surveillance data, with reduced communication/computation cost. The suggested security framework provides essential security measures and validates its resilience from attacks using the Scyther tool. The security analysis of the proposed LPSF-DA-VSS scheme is further verified using the Random Oracle Model (ROM), which ensures that the scheme is provably secure. Finally, the Raspberry Pi testbed is used to exhibit the possible practical realization of the LPSF-DA-VSS scheme. Overall, the LPSF-DA-VSS scheme supports enhanced security features and is provably secure on the cloud framework due to its wide acceptability.

In the digital communication era, protecting the confidentiality of visual data like images is done through suitable chaotic map-based pseudo sequences rather than using any standard block ciphering algorithm. The chaotic map-based image cryptosystem supports faster and more desirable image-enciphering, even in real-time scenarios. An effective chaotic map is desirable for image encryption to ensure the sensitivity property and dynamic behaviour with different seed values. Many existing chaotic map-based cryptosystems have shown poor sensitivity to the initial condition and lower dynamic behaviour with key generation algorithms. The suggested encryption process for image data exploits bit plane level content shuffling in Intra and Inter-level with a suitable 3D chaotic system to enhance security and performance. The overall iterations are significantly fewer than standard practice, even with supporting desirable cipher image formation. Further, the cipher image reveals the negligible linear relationship between the neighbouring pixels, and the change in the neighbouring pixel of the plain image reduces the chances of statistical attacks. In the experimental process, the natural grayscale and medical imaging in DICOM formats are considered during an enciphering time, where keyspace analysis reveals a large key space of 2312, providing strong resistance against brute-force attacks. Statistical tests show the encrypted images achieve ideal entropy values of approximately 7.999 and negligible inter-pixel correlation values of less than 0.005 in all directions. The proposed scheme also performs robustly against exhaustive attacks, noise, occlusion, and chosen plaintext attack, including differential attack analysis where NPCR is more significant than 99.6%, UACI is approximately 33.95%. The chosen plaintext attack analysis exhibits the cryptosystem’s resistance to the cryptanalysis attack. The results are comparable to some related works, indicating standard performance across multiple security metrics, and are found effective in secure image communication even with a small key size.

Lossless compression is preferred for the compact representation of medical images. In general, single compression algorithms are not that effective compared to the hybrid compression algorithm where several compression algorithms are collectively used properly to realize better compression. In this paper, the Burrows–Wheeler Transform (BWT) is applied at a block level for each bit plane of the medical images. Thus, this process identifies highly correlated binary streams, and subsequently, the obtained binary streams are further encoded through the modified Run Length Encoding (RLE). The probability of RLE is calculated for each bit plane and used subsequently to perform Huffman coding on each bit plane individually. The DICOM category of medical images and some images that are represented by 8-bit planes are used for the validation of the above approach. It is found that the present approach yields significantly better compression than that obtained with the standalone lossless compression algorithm that is Huffman coding and the conventional BWT-based lossless image compression algorithm.

In online healthcare systems, patient data confidentiality and integrity are the primary concerns. After COVID-19, we have seen that the usability of the online healthcare system has increased significantly. However, in this scenario, there is a need to protect the security of the patient data. Since a patient's data might become vulnerable when unauthorized changes are made, there is a need for a strategy where if any alteration is made, it can be easily detected. This paper provides a blockchain framework to validate the patient's test report. Initially, the patient data is captured by the Medical Diagnostic Center (MDC), and subsequently, lossless compression is applied to the patient data for proper bandwidth utilization. Here, the MDC applied a partial encryption process on medical data for uploading into the InterPlanatary File System (IPFS). The encryption is carried out using Advanced Encryption Standard (AES) in Galois Counter Mode (GCM). The practitioners can retrieve the secure medical test data through the suggested blockchain framework, and subsequently, they can check the integrity of the data before considering their treatment process. Overall, the system preserves the security of the patient data, and legitimate users can access the data securely. The suggested mechanism has been implemented using a private Ethereum blockchain. Overall, the proposed framework is suitable for preserving the security of the patient data.

With the voluminous increase in the amount of data in this era, there is a dire need for a suitable medium to store them, for which DNA is coming up as an emerging technology. However, this data should not only be stored efficiently so as to reduce the cost of storage, but also be protected against malicious use. In this paper, we propose a DNA-based lossless compression-encryption algorithm for a highly efficient and secure storage of data. Huffman coding proves to be one of the efficient compression algorithms, which is first used to compress the data. This data is then converted into a sequence of DNA nucleotides using DNA Encoding Rules. These obtained sequences are then secured with the help of DNA-based Advanced Encryption Standard (AES) using a DNA-based symmetric key. The proposed algorithm works efficiently for text and image data and proves to give a better compression ratio than the existing algorithms. The encrypted data is secure against various cryptography attacks, and the original data can only be recovered by the knowledge of correct initial vectors and symmetric key.

In the current era, the deployment of video surveillance systems (VSSs) is becoming a common practice in society for properly maintaining laws and order in public and private places. The surveillance video data (SD) generated from such infrastructure is sensitive and must address the concerns about its security. To ensure the security of SD, there is a need for a mechanism that provides authenticity, confidentiality, and integrity, all at the same time while being lightweight. ASCON delivers all these capabilities. This paper presents a novel approach to enhance the security of the smart city surveillance system using ASCON and Blockchain. Due to its immutable storage property, the Blockchain detects hardware attacks by validating the cameras and local servers. In this paper, SD is stored in the cloud server due to limited local server storage, which can become a primary target for attackers. Therefore, this paper stores the encrypted SD into the cloud server. Formal security analysis using the Scyther tool and the Random Oracle Model (ROM) validates the system’s robustness. Additionally, this paper evaluates and contrasts the proposed scheme’s computational and transmission overheads with the other existing schemes. The results show that the proposed system significantly enhances the security of SD, offering a resilient solution for smart city surveillance infrastructures.

The availability of obscene content across diverse domains poses a threat to user well-being, especially children, as well as adults at the workplace. Detecting obscene content across image domains is vital, whereas a single classifier approach lacks the ability to represent the domain differences semantically. To tackle this issue, we propose a domain-aware obscenity classification framework in this paper. It trains the domain-aware classifier with shared head that aligns the semantic boundary between obscene and non-obscene content across domains. We present GRASP-former, a novel feature representation architecture for obscene image classification, where the random–global sparse attention builds a lightweight global context by attending to a small set of learnable global tokens and randomly sampled tokens. We fuse it with depthwise local convolution, which further refines the obscenity features to distinguish visual ambiguities present in obscene and non-obscene classes. We evaluate the proposed architecture with standard performance metrics using samples from the NPDI and NSFW datasets.

In this paper, we developed a Blockchain-based User Authentication Data-Sharing (BC-UADS) framework. In BC-UADS, several hospital servers form a consortium blockchain network to maintain the transparency, immutability, and authenticity of the patient's Electronic Healthcare (Record EHR) medical data. The BC-UADS framework allows doctors to share or retrieve a patient's EHR metadata from the blockchain network. Since, the metadata is stored on a blockchain platform, it is more secure and trusted for real-time applications to utilize the medical data. The data-sharing protocol of the BC-UADS framework is implemented based on Proof-of-Reputation consensus algorithm in a blockchain network. The BC-UADS framework is analyzed in the AVISPA (Automated Validation of Internet Security Protocols and Applications) tool, demonstrating that it is secure against active and passive attacks. Besides, the BC-UADS framework is provably secure in the random oracle model based on the hardness assumption of Elliptic Curve-based Computational Diffie-Hellman (ECCDH) problem. The mutual authentication property of the BC-UADS framework is analyzed in the BAN (Burrows-Abadi-Needham) logic model. We have computed the communication, execution, and storage costs of the BC-UADS framework in different security levels: 80-bit, 112-bit, 128-bit, 192-bit, and 256-bit using PBC library. The proposed BC-UADS framework is compared with the state-of-the-art schemes.

To evaluate the diagnostic accuracy of artificial intelligence-based algorithms in identifying neck of femur fracture on a plain radiograph. Systematic review and meta-analysis. PubMed, Web of science, Scopus, IEEE, and the Science direct databases were searched from inception to 30 July 2023. Eligible article types were descriptive, analytical, or trial studies published in the English language providing data on the utility of artificial intelligence (AI) based algorithms in the detection of the neck of the femur (NOF) fracture on plain X-ray. The prespecified primary outcome was to calculate the sensitivity, specificity, accuracy, Youden index, and positive and negative likelihood ratios. Two teams of reviewers (each consisting of two members) extracted the data from available information in each study. The risk of bias was assessed using a mix of the CLAIM (the Checklist for AI in Medical Imaging) and QUADAS-2 (A Revised Tool for the Quality Assessment of Diagnostic Accuracy Studies) criteria. Of the 437 articles retrieved, five were eligible for inclusion, and the pooled sensitivity of AIs in diagnosing the fracture NOF was 85

The video surveillance system (VSS) is an integral part of the modern society. The surveillance camera captures the video of its designated areas and subsequently stores the video data in the server for future needs. In this situation, the VSS should send the captured video to the server securely; later, according to the demand, legitimate users can access the video data securely. This chapter suggests a secure VSS where the camera transmits the surveillance video data (SVD) to the server once both entities are mutually authenticated and generates the session key. Further, a legitimate user can access the SVD from the server once both parties are mutually authenticated and ensure a session key for secure data transmission. This chapter proposes a lightweight authenticated session key agreement for VSS based on the ASCON, strengthening access control and safeguarding private data. We call the proposed scheme LASKV. The LASKV scheme is lightweight and upholds security against various cryptographic attacks. In this regard, the informal security analysis and the Scyther simulation endorse the security of the LASKV scheme. Overall, the LASKV scheme effectively stores and accesses the SVD safely within an organization.

Recently, the image retrieval process appears to be a challenging task to filter out a huge volume of objectionable images from retrieval. It is very easy for people of all ages to obtain such obscene images with just a few clicks on the internet, which could harm the adolescent. To prevent such serious social problems, this paper introduces a content-based image filtering (CBIF) framework in an image retrieval system. The proposed CBIF approach is split into two components. In the first component, the skin region of exposed human body parts is extracted based on a color segmentation method, and an efficient feature vector generation process is introduced based on frequent skin color pixels and scale-invariant structural features. A novel texture feature representation is presented in the retrieval and filtering process obtained by block-paired local binary pattern (BP-LBP). The final component introduces a novel image filtering framework for the retrieved images found from the first component based on a majority of voting method on three different machine learning classifiers backed by SVM, MLP, and CNN. The retrieval efficiency is analyzed with standard performance indicators such as precision, recall, and F-score on two large datasets: NPDI and NSFW.

Recently, DNA encoding has shown its potential to store the vital information of the image in the form of nucleotides, namely A, C, T , and G , with the entire sequence following run-length and GC-constraint. As a result, the encoded DNA planes contain unique nucleotide strings, giving more salient image information using less storage. In this paper, the advantages of DNA encoding have been inherited to uplift the retrieval accuracy of the content-based image retrieval (CBIR) system. Initially, the most significant bit-plane-based DNA encoding scheme has been suggested to generate DNA planes from a given image. The generated DNA planes of the image efficiently capture the salient visual information in a compact form. Subsequently, the encoded DNA planes have been utilized for nucleotide patterns-based feature extraction and image retrieval. Simultaneously, the translated and amplified encoded DNA planes have also been deployed on different deep learning architectures like ResNet-50, VGG-16, VGG-19, and Inception V3 to perform classification-based image retrieval. The performance of the proposed system has been evaluated using two corals, an object, and a medical image dataset. All these datasets contain 28,200 images belonging to 134 different classes. The experimental results confirm that the proposed scheme achieves perceptible improvements compared with other state-of-the-art methods.

In an era marked by the contrast between information and disinformation, the ability to differentiate between authentic and manipulated images holds immense importance for both security professionals and the scientific community. Copy-move forgery is widely practiced thus, sprang up as a prevalent form of image manipulation among different types of forgeries. In this counterfeiting process, a region of an image is copied and pasted into different parts of the same image to hide or replicate the same objects. As copy-move forgery is hard to detect and localize, a swift and efficacious detection scheme based on keypoint detection is introduced. Especially the localization of forged areas becomes more difficult when the forged image is subjected to different post-processing attacks and geometrical attacks. In this paper, a robust, translation-invariant, and efficient copy-move forgery detection technique has been introduced. To achieve this goal, we developed an AKAZE-driven keypoint-based forgery detection technique. AKAZE is applied to the LL sub-band of the SWT-transformed image to extract translation invariant features, rather than extracting them directly from the original image. We then use the DBSCAN clustering algorithm and a uniform quantizer on each cluster to form group pairs based on their feature descriptor values. To mitigate false positives, keypoint pairs are separated by a distance greater than a predefined shift vector distance. This process forms a collection of keypoints within each cluster by leveraging their similarities in feature descriptors. Our clustering-based similarity-matching algorithm effectively locates the forged region. To assess the proposed scheme we deploy it on different datasets with post-processing attacks ranging from blurring, color reduction, contrast adjustment, brightness change, and noise addition. Even our method successfully withstands geometrical manipulations like rotation, skewing, and different affine transform attacks. Visual outcomes, numerical results, and comparative analysis show that the proposed model accurately detects the forged area with fewer false positives and is more computationally efficient than other methods.

DNA is a macromolecule that carries the genetic information of nearly all living things on the planet. They not only determine the characteristics and behavior of an organism but also pass the essential features to the next generation, ensuring that “like begets like." Because of their same genetic structure, organisms of the same species appear identical. Inspired by this property, a novel DNA-based scheme for class-based image retrieval has been proposed. The algorithm imitates the flow of genetic information, which is initially stored in the DNA and is transcripted and translated to RNA and amino-acid sequences, respectively, using genetic coding. Since similar images would generate a similar sequence, ensuring the preservation of salient features of the images required for retrieval. Thus, these amino-acid sequences are then deployed on a DNA-inspired ResNet-50 CNN architecture for performing image classification-based image retrieval. The proposed scheme has been extensively tested on six different datasets to demonstrate its performance. Comparative results reveal that the proposed scheme outperforms competing state-of-the-art Content-based Image Retrieval techniques in terms of retrieval performance.

The trend for deploying Video Surveillance Systems (VSSs) in public places has become common practice to maintain effective law and order in modern civilization. Further, data access control and the proper management of surveillance data with valid users are desirable for the safety and security of the communities. This paper aims to develop practical solutions to protect VSSs against evolving threats and challenges. This paper proposes a Two-Factor Mutual Authentication and Session Key Agreement usable in VSS (2F-MASK-VSS) environments for real-time data storage and access. In 2F-MASK-VSS, lightweight cryptographic tools, viz. hash function and symmetric key encryption, are used to maintain the desirable security features. In 2F-MASK-VSS, a surveillance camera captures real-time data and sends them securely to a central server for storage through the established session key agreement among valid concerns. Moreover, 2F-MASK-VSS can protect access control among valid users. The security strength of 2F-MASK-VSS has been proven by formal and informal analysis. The BAN logic model, AVISPA and Scyther tools validate the attack-resilience of 2F-MASK-VSS. Furthermore, the security analysis in the random oracle model shows that 2F-MASK-VSS is provably secure. In addition, 2F-MASK-VSS has been implemented using the Raspberry PI testbed to demonstrate its practical implementation.