The digitisation of agriculture, integral to Agriculture 4.0, has brought significant benefits while simultaneously escalating cybersecurity risks. With the rapid adoption of smart farming technologies and infrastructure, the agricultural sector has become an attractive target for cyberattacks. This paper presents a systematic literature review that assesses the applicability of existing cyber threat intelligence (CTI) techniques within smart farming infrastructures (SFIs). We develop a comprehensive taxonomy of CTI techniques and sources, specifically tailored to the SFI context, addressing the unique cyber threat challenges in this domain. A crucial finding of our review is the identified need for a virtual Chief Information Security Officer (vCISO) in smart agriculture. While the concept of a vCISO is not yet established in the agricultural sector, our study highlights its potential significance. The implementation of a vCISO could play a pivotal role in enhancing cybersecurity measures by offering strategic guidance, developing robust security protocols, and facilitating real-time threat analysis and response strategies. This approach is critical for safeguarding the food supply chain against the evolving landscape of cyber threats. Our research underscores the importance of integrating a vCISO framework into smart farming practices as a vital step towards strengthening cybersecurity. This is essential for protecting the agriculture sector in the era of digital transformation, ensuring the resilience and sustainability of the food supply chain against emerging cyber risks.

The diagnosis of thoracic diseases, many of which are easily treatable, is mainly done with the use of chest X-rays and other chest imaging techniques and making sense of these require expert radiologists, who aren’t always accessible. Using Deep Neural Networks, patterns corresponding to different thoracic diseases can be detected from these chest X-rays with the aid of machines. In this study, we have proposed such a model which can detect the presence of 14 different thoracic diseases from a chest X-ray alone. Our novel dense convolutional neural network model works in 2 stages, first training on images from disease-ridden patients alone, and then training the entire network on the whole dataset which includes X-rays from both healthy and unhealthy patients. Given a chest X-ray alone, our model can give accurate predictions, with an AUROC mean score of 82.9% competing with the existing state-of-the-art models in this field.

Alzheimer’s disease is a neurological disease that affects the healthy cells of the brain and results in people having long-term memory loss, thinking problems, disorientation, behavioral inconsistencies and finally death. When the disease gets detected, the pathological load is already high, and there is no coming back from there. This neurodegenerative disease consists of three general stages, which we classified in this research and that includes very mild (early stage), mild (middle stage) and finally, the moderate stage (late-stage). Using transfer learning, we implemented five existing efficient and recent Convolutional Neural Networks (CNN) models such as VGG19, Inception- ResNetv2, ResNet152v2, EfficientNetB5 and EfficientNetB6, and another custom one of our own. Later, we ensembled thrice with multiple combinations of the models to enhance our outcome. This led us to achieve our proposed model, which is a weighted average ensemble of all the six classifiers mentioned above and this novel approach gave us an accuracy of 96%, which was quite satisfactory compared to any other existing models.

Mental stress is the main well-being problem worldwide today. It is responsible for most of all mental-brain diseases. It does not need any specific reason to happen. It can be experienced from a very little incident to a huge incident. The consequence of it depends on how people handle it. Depression and anxiety are the results of mental stress, and they are the major challenges in today’s world. Depression and anxiety are the leading causes of suicide. Most of the time suicidal patients hide their true feelings and fail to communicate their psychiatric problems to physicians. The specific issues that need to be addressed are finding an easy, reliable and realistic way to diagnose mental stress to keep it from becoming a serious and irreversible condition. The primary prevention of mental stress utilizing machine learning algorithms based on reward and punishment processing is important to avoid mental diseases. Several techniques have been used to detect mental stress, and very few papers have tried to detect a patients’ comorbidity condition. However, literature shows that there are still chances of further improvement in this field. The traditional methods of detecting mental stress involve a statistical questionnaire approach with some shortcomings as results based on the traditional approach can be biased, which is not possible if Electroencephalogram (EEG) signals are used. Therefore, in this paper, we proposed a method to evaluate the EEG signals on thirty-two individuals for identifying comorbid patients using nine machine learning classifiers based on reward and punishment processing. The performance of our method is also shown to be better than some existing methods.

In recent years, advancement in the realm of machine learning has introduced a feature known as Deepfake pictures, which allows users to substitute a genuine face with a fake one that seems real. As a result, distinguishing between authentic and fraudulent pictures has become difficult. There have been several cases in recent years where Deepfake pictures have been used to defame famous leaders and even regular people. Furthermore, cases have been documented in which Deepfake yet realistic pictures were used to promote political discontent, blackmail, spread fake news, and even carry out false terrorism attacks. The objective of our model is to differentiate between real and Deepfake images so that the above mentioned situations can be avoided. This project represents a deep CNN model with 13000 images divided in two segments that are: Training and Testing. The dataset was prepared using necessary image augmentation techniques. A total of 2 categories are considered (real image category and fake image category). Our suggested model was successful in achieving 98.77% accuracy. The model shows promising results in the case of detecting real and DeepFake images than all the other models used before.

The neurodegenerative Alzheimer's Disease is the most widely recognized cause of ‘Dementia’ and was allegedly the 7th highest cause of death globally. Yet, there is still no conclusive test for distinguishing Alzheimer's disease. Our proposed model eliminates these challenges in a significant manner. The technique is fit for investigating and analyzing different classes in a single setting and requires significantly less previous apprehension. Several handcrafted or predefined machine learning and deep learning models have been imple-mented in this field of study. Our proposed multi-classification model is primarily implemented based on the Open Access Series of Imaging Studies (OASIS) data and suggests an 18-layer architecture. We have implemented a unique preprocessing approach using all three anatomical planes of the MRI scans in a single sequential model, which was also evaluated afterwards. The research also explores a comparative study among multiple and binary classes in terms of performance and efficiency. Pre-defined models such as Inception V3and VGG19 have also been brought to comparison to measure the model's reliability. Our multiclass setting shows an accuracy of over 80%, which is higher than most of the existing multi-classification models in this dataset. Moreover, the in-depth comparative study using binary classification shows a significant accuracy of over 92%, which ensures the all-around efficacy of the model.

Alzheimer's Disease (AD) is a neurological condition in which the decline of brain cells causes memory loss and cognitive decline. Various Neuroimaging techniques have been developed to diagnose AD; among those, Magnetic Resonance Imaging (MRI) is one of the most prominent ones. Historically, expert radiologists were solely responsible for making decisions of a patient's AD situation by manually analyzing brain MR images. However, the recent progress in medical image analysis using deep learning especially has automated this task significantly. Although the state-of-the-art architectures have achieved human-level performance in classifying AD images from Normal Control (NC), they often require predefined Regions of interest as a basis for feature extraction. This condition not only requires specialized domain knowledge of the human brain but also makes the overall design complicated. In this paper, we designed a 14 layer Neural network architecture that can facilitate AD diagnosis without being dependent on any neurological assumption. The network was tested over ADNI-1, a benchmark MRI dataset for AD research, and found an accuracy of 87.06 %$(\mathbf{AUC}=\mathbf{0. 9 3}.)$

Epileptic seizure is a neurological disorder that is prevalent in both males and females of all age ranges. Detection of epileptic seizure serves as an important role for epileptic patients as it allows the initiation of systems to prevent injuries and limiting the possibilities of risk by providing targeted therapy by anticipating their onset prior to presentation. Electroencephalography (EEG) plays an important role in seizure detection and is one of the most well-known techniques for determining stages of epilepsy. Since, EEG is a non-stationary signal it can be quite difficult to differentiate amongst seizure activity and normal neural activity. In this paper we have proposed an epilepsy detection method based on five different feature extraction methods and followed by that the original domain of the extracted features were transformed using Discrete Wavelet Transform (DWT) and three different classifiers- Decision Tree, Random Forest and KNN to classify into seizure and non-seizure stages. Results demonstrated in this paper have outperformed the existing state-of-the-art methods with 97.22%, 100% and 83.33% for 2 class classification and 91.67%, 91.67% and 80.56% for 4 class classification for the aforementioned classification techniques accordingly.

Using a number of Neural Network Models, we attempt to explore and explain the prediction of Alzheimer's in patients in various stages of the disease, using MRI imaging data. Alzheimer's disease(AD) often described as dementia is one of the major neurological dysfunctionalities among humans and does not yet have a proven detection system; unless the final stage symptoms of AD starts to be seen. It is observed that multimodal biological, imaging and other available neuropsychological data can ensure a high percentage of separation among (AD) patients from cognitively normal elders. However, they cannot surely predict or detect early enough that patients with mild cognitive impairment (MCI) can get converted into Alzheimer's disease dementia in the future. But the research done till date shows a high probable detection rate in which they used the pattern classifier built on various longitudinal data. So in this paper we experimented with the existing Neural Network models to detect Alzheimer's disease in its early stage by classification techniques; and will be using a recent hybrid dataset in the process to have four separate classification in total. And also explored the exact region for which that specific classification occurs for the patients, looking at the T1 weighted MRI scans from a hybrid dataset from Kaggle [1] using the LIME based XAI(Explainable Artificial Intelligence) framework. For the Convolution Neural Network Models we are using Resnet50, VGG16 and Inception v3 and received 82.56%, 86.82%, 82.04% of categorical accuracy respectively.

Functional Magnetic Resonance Imaging(fMRI) is one of the best neuroimaging techniques which helps to understand the activity of the human brain. With the help of recent advancement in the field of machine learning algorithms in terms of pattern recognition, now it is possible to extract in-depth information about brain activity by analyzing fMRI data. In this paper, we have shown the analysis of the data of a particular part of the human brain called Ventral Temporal Cortex. The dataset contains the fMRI data of the subjects while viewing grey-scale image different categories of objects such as cat, chair, etc. We have applied the machine learning algorithms on the extracted feature set from fMRI data to classify the objects that the subject is viewing. Here, we have emphasized on hyper-parameter tuning for the classifiers. Among the classifiers, we have found that the performance of Support Vector Machine (i.e., 96.92%) and k-nearest neighbor classifier(i.e., 96.90%) is quite persistent and have better accuracy. The further application of this research may motivate to develop brain-computer interface (BCI) based solutions.

Reliable detection of cognitive load would benefit the design of intelligent assistive navigation aids for the visually impaired (VIP). Ten participants with various degrees of sight loss navigated in unfamiliar indoor and outdoor environments, while their electroencephalogram (EEG) and electrodermal activity (EDA) signals were being recorded. In this study, the cognitive load of the tasks was assessed in real time based on a modification of the well‐established event‐related (de)synchronization (ERD/ERS) index. We present an in‐depth analysis of the environments that mostly challenge people from certain categories of sight loss and we present an automatic classification of the perceived difficulty in each time instance, inferred from their biosignals. Given the limited size of our sample, our findings suggest that there are significant differences across the environments for the various categories of sight loss. Moreover, we exploit cross‐modal relations predicting the cognitive load in real time inferring on features extracted from the EDA. Such possibility paves the way for the design on less invasive, wearable assistive devices that take into consideration the well‐being of the VIP.

Electroencephalogram (EEG) has a great potential for diagnosis and treatment of brain disorders like epileptic seizure. Feature extraction and classification of EEG signals is the crucial task to detect the stages of ictal and interictal signals for treatment and precaution of epileptic patients. However, existing seizure and non‐seizure feature extraction techniques are not good enough for the classification of ictal and interictal EEG signals considering the non‐abruptness phenomena and inconsistency in different brain locations. In this study, the authors present a new approach for feature extraction and classification by exploiting temporal correlation within EEG signals for better seizure detection as any abruptness in the temporal correlation within a signal represents the transition of a phenomenon. In the proposed methods, they divide an EEG signal into a number of epochs and arrange them into two‐dimensional matrix and then apply different transformation/decomposition to extract a number of statistical features. These features are then used as an input into LS‐SVM to classify them. Experimental results show that the proposed methods outperform the existing state‐of‐the‐art method for better classification in terms of sensitivity, specificity and accuracy of ictal and interictal period of epilepsy for benchmark datasets and different brain locations.