Deep fake detection and classification using error-level analysis and deep learning.

Due to the wide availability of easy-to-access content on social media, along with the advanced tools and inexpensive computing infrastructure, has made it very easy for people to produce deep fakes that can cause to spread disinformation and hoaxes. This rapid advancement can cause panic and chaos as anyone can easily create propaganda using these technologies. Hence, a robust system to differentiate between real and fake content has become crucial in this age of social media. This paper proposes an automated method to classify deep fake images by employing Deep Learning and Machine Learning based methodologies. Traditional Machine Learning (ML) based systems employing handcrafted feature extraction fail to capture more complex patterns that are poorly understood or easily represented using simple features. These systems cannot generalize well to unseen data. Moreover, these systems are sensitive to noise or variations in the data, which can reduce their performance. Hence, these problems can limit their usefulness in real-world applications where the data constantly evolves. The proposed framework initially performs an Error Level Analysis of the image to determine if the image has been modified. This image is then supplied to Convolutional Neural Networks for deep feature extraction. The resultant feature vectors are then classified via Support Vector Machines and K-Nearest Neighbors by performing hyper-parameter optimization. The proposed method achieved the highest accuracy of 89.5% via Residual Network and K-Nearest Neighbor. The results prove the efficiency and robustness of the proposed technique; hence, it can be used to detect deep fake images and reduce the potential threat of slander and propaganda.

Neurological manifestations in SARS-CoV-2 infection: A single-center cross-sectional study in Malaysia.

Background: Neurological involvement associated with SARS-CoV-2 infection has been reported from different regions of the world. However, data from South East Asia are scarce. We described the neurological manifestations and their associated factors among the hospitalized COVID-19 patients from an academic tertiary hospital in Malaysia. Methods: A cross-sectional observational study of hospitalized COVID-19 patients was conducted. The neurological manifestations were divided into the self-reported central nervous system (CNS) symptoms, stroke associated symptoms, symptoms of encephalitis or encephalopathy and specific neurological complications. Multiple logistic regression was performed using demographic and clinical variables to determine the factors associated with outcome. Results: Of 156 hospitalized COVID-19 patients with mean age of 55.88 ± 6.11 (SD) years, 23.7% developed neurological complications, which included stroke, encephalitis and encephalopathy. Patients with neurological complications were more likely to have diabetes mellitus (p = 0.033), symptoms of stroke [limb weakness (p < 0.001), slurred speech (p < 0.001)]; and encephalitis or encephalopathy [confusion (p < 0.001), forgetfulness (p = 0.006) and seizure (p = 0.019)]. Unvaccinated patients had a 4.25-fold increased risk of having neurological complications (adjusted OR = 4.25; 95% CI: 1.02, 17.71, p = 0.047). Anosmia and dysgeusia were less associated with neurological complications (adjusted OR = 0.22; 95% CI: 0.05, 0.96, p = 0.044). The odds of neurological complications were increased by 18% in patients with leukocytosis (adjusted OR = 1.18, 95% CI: 1.003, p = 0.0460). Conclusions: Stroke, encephalitis and encephalopathy were the common neurological complications from our study. Diabetes mellitus, presence of symptoms of stroke, symptoms of encephalitis or encephalopathy, leukocytosis, and being unvaccinated against COVID-19 were the associated risk factors of developing neurological complications.

Ultrasonographic Evaluation of Skin Toxicity Following Radiotherapy of Breast Cancer: A Systematic Review.

The present review aimed to systematically review skin toxicity changes following breast cancer radiotherapy (RT) using ultrasound (US). PubMed and Scopus databases were searched according to PRISMA guidelines. The characteristics of the selected studies, measured parameters, US skin findings, and their association with clinical assessments were extracted. Seventeen studies were included with a median sample size of 29 (range 11-166). There were significant US skin changes in the irradiated skin compared to the nonirradiated skin or baseline measurements. The most observed change is skin thickening secondary to radiation-induced oedema, except one study found skin thinning after pure postmastectomy RT. However, eight studies reported skin thickening predated RT attributed to axillary surgery. Four studies used US radiofrequency (RF) signals and found a decrease in the hypodermis's Pearson correlation coefficient (PCC). Three studies reported decreased dermal echogenicity and poor visibility of the dermis-subcutaneous fat boundary (statistically analysed by one report). The present review revealed significant ultrasonographic skin toxicity changes in the irradiated skin most commonly skin thickening. However, further studies with large cohorts, appropriate US protocol, and baseline evaluation are needed. Measuring other US skin parameters and statistically evaluating the degree of the association with clinical assessments are also encouraged.

Otalgia and Facial Nerve Palsy: Common Symptoms Revealing the Uncommon Pathology of Middle Ear Lymphoma.

Lymphoma of the middle ear and mastoid is uncommon and is rarely diagnosed early. The clinical presentation occurs due to consequences of extensive progressive disease. It can manifest as benign middle ear pathologies such as otitis media; other presentations include facial nerve palsy, hearing loss, and vestibular dysfunction. We treated a case of a 38-year-old male who presented with extranodal involvement of diffuse large B-cell lymphoma (DLBCL) of the middle ear, mastoid, and temporalis muscle, which mimicked an acute otitis media complicated with facial nerve palsy and hearing loss. He has underlying mediastinal and cervical DLBCL diagnosed 20 months before the current presentation. He underwent cortical mastoidectomy and chemotherapy. Despite treatment, he succumbed to the disease. We discuss the clinical significance of middle ear lymphoma by reviewing similar cases in the literature. To conclude, refractory middle ear disease should alert the surgeon of a more sinister underlying pathology in a patient with malignancy.

Dragonfly Algorithm and Its Hybrids: A Survey on Performance, Objectives and Applications.

Swarm intelligence is a discipline which makes use of a number of agents for solving optimization problems by producing low cost, fast and robust solutions. The dragonfly algorithm (DA), a recently proposed swarm intelligence algorithm, is inspired by the dynamic and static swarming behaviors of dragonflies, and it has been found to have a higher performance in comparison to other swarm intelligence and evolutionary algorithms in numerous applications. There are only a few surveys about the dragonfly algorithm, and we have found that they are limited in certain aspects. Hence, in this paper, we present a more comprehensive survey about DA, its applications in various domains, and its performance as compared to other swarm intelligence algorithms. We also analyze the hybrids of DA, the methods they employ to enhance the original DA, their performance as compared to the original DA, and their limitations. Moreover, we categorize the hybrids of DA according to the type of problem that they have been applied to, their objectives, and the methods that they utilize.

Gastrocnemius acrometastasis from muscle-invasive urothelial bladder carcinoma: A case report.

Urinary bladder carcinoma is a common malignancy worldwide. The metastatic disease to distant organs including lung, liver, and bone is well established. However, metastasis to below-knee-level; also known as acrometastasis is a rare occurrence and occurs approximately 0.1% of all bone metastases. It is standard of care to obtain a contrast enhanced computed tomography scan of the chest, abdomen, and pelvis for pretreatment planning, primary staging, and post treatment disease surveillance. This makes the occurrence of acrometastasis harder to detect and may only manifest clinically in advance disease. We report a case of 55 years old gentleman treated as muscle-invasive bladder urothelial carcinoma, presented with chronic left knee pain, and imaging demonstrating tumor in the left knee region. Histopathologic study shows features of metastatic disease from urinary bladder carcinoma to the left gastrocnemius muscle. The attending physician should raise the suspicion of metastatic disease if the patient with known malignancy presented with new soft tissue lesion elsewhere in the body.

Intraventricular tuberculosis abscess in an immunocompromised patient: clinical vignette.

Tuberculosis is caused by Mycobacterium tuberculosis. Tuberculosis of the central nervous system is common and manifestations include meningeal and intraparenchymal diseases. However, intraventricular tuberculous abscess is a rare manifestation of intracranial tuberculous infection. We present a case of an immunocompromised female patient with high-grade fever and signs of meningism. The computed tomography and magnetic resonance imaging (MRI) of the brain showed hydrocephalus with rim-enhancing lesion in the right lateral ventricle. The MRI demonstrated a hypointense signal on T1-weighted imaging, hyperintense signal on T2-weighted imaging, and mild restricted diffusion in diffusion-weighted imaging. She underwent emergency external ventricular drainage and frank pus was drained. Diagnosis of tuberculosis was made via polymerase chain reaction analysis and culture. Understanding the intracranial manifestation of neurotuberculosis is imperative to arrive at the diagnosis correctly and ensure prompt treatment.

Epidemiology and disease burden of tuberculosis in south of Tunisia over a 22-year period: Current trends and future projections.

BACKGROUND: Tuberculosis (TB) is a public health problem worldwide. Characterizing its trends over time is a useful tool for decision-makers to assess the efficiency of TB control programs. We aimed to give an update on the current chronological trends of TB in Southern Tunisia from 1995 to 2016 and to estimate future trajectories of TB epidemic by 2030. METHODS: We retrospectively collected data of all notified TB new cases by the Center of Tuberculosis Control between 1995 and 2016 in South of Tunisia. Joinpoint Regression Analysis was performed to analyze chronological trends and annual percentage changes (APC) were estimated. RESULTS: In the past 22 years, a total of 2771 cases of TB were notified in Southern Tunisia. The annual incidence rate of TB was 13.91/100,000 population/year. There was a rise in all forms of TB incidence (APC = 1.63) and in extrapulmonary tuberculosis (EPTB) (APC = 2.04). The incidence of TB increased in children and adult females between 1995 and 2016 (APC = 4.48 and 2.37, respectively). The annual number of TB declined in urban districts between 2004 and 2016 (APC = -2.85). Lymph node TB cases increased (APC = 4.58), while annual number of urogenital TB decreased between 1995 and 2016 (APC = -3.38). Projected incidence rates would increase to 18.13 and 11.8/100,000 population in 2030 for global TB and EPTB, respectively. CONCLUSIONS: Our study highlighted a rise in all forms of TB and among high-risk groups, notably children, females and lymph node TB patients in the last two decades and up to the next one.

A fuzzy logic control in adjustable autonomy of a multi-agent system for an automated elderly movement monitoring application.

Autonomous agents are being widely used in many systems, such as ambient assisted-living systems, to perform tasks on behalf of humans. However, these systems usually operate in complex environments that entail uncertain, highly dynamic, or irregular workload. In such environments, autonomous agents tend to make decisions that lead to undesirable outcomes. In this paper, we propose a fuzzy-logic-based adjustable autonomy (FLAA) model to manage the autonomy of multi-agent systems that are operating in complex environments. This model aims to facilitate the autonomy management of agents and help them make competent autonomous decisions. The FLAA model employs fuzzy logic to quantitatively measure and distribute autonomy among several agents based on their performance. We implement and test this model in the Automated Elderly Movements Monitoring (AEMM-Care) system, which uses agents to monitor the daily movement activities of elderly users and perform fall detection and prevention tasks in a complex environment. The test results show that the FLAA model improves the accuracy and performance of these agents in detecting and preventing falls.

A review of norms and normative multiagent systems.

Norms and normative multiagent systems have become the subjects of interest for many researchers. Such interest is caused by the need for agents to exploit the norms in enhancing their performance in a community. The term norm is used to characterize the behaviours of community members. The concept of normative multiagent systems is used to facilitate collaboration and coordination among social groups of agents. Many researches have been conducted on norms that investigate the fundamental concepts, definitions, classification, and types of norms and normative multiagent systems including normative architectures and normative processes. However, very few researches have been found to comprehensively study and analyze the literature in advancing the current state of norms and normative multiagent systems. Consequently, this paper attempts to present the current state of research on norms and normative multiagent systems and propose a norm's life cycle model based on the review of the literature. Subsequently, this paper highlights the significant areas for future work.

Effect of temporal relationships in associative rule mining for web log data.

The advent of web-based applications and services has created such diverse and voluminous web log data stored in web servers, proxy servers, client machines, or organizational databases. This paper attempts to investigate the effect of temporal attribute in relational rule mining for web log data. We incorporated the characteristics of time in the rule mining process and analysed the effect of various temporal parameters. The rules generated from temporal relational rule mining are then compared against the rules generated from the classical rule mining approach such as the Apriori and FP-Growth algorithms. The results showed that by incorporating the temporal attribute via time, the number of rules generated is subsequently smaller but is comparable in terms of quality.

Naive Bayes-guided bat algorithm for feature selection.

When the amount of data and information is said to double in every 20 months or so, feature selection has become highly important and beneficial. Further improvements in feature selection will positively affect a wide array of applications in fields such as pattern recognition, machine learning, or signal processing. Bio-inspired method called Bat Algorithm hybridized with a Naive Bayes classifier has been presented in this work. The performance of the proposed feature selection algorithm was investigated using twelve benchmark datasets from different domains and was compared to three other well-known feature selection algorithms. Discussion focused on four perspectives: number of features, classification accuracy, stability, and feature generalization. The results showed that BANB significantly outperformed other algorithms in selecting lower number of features, hence removing irrelevant, redundant, or noisy features while maintaining the classification accuracy. BANB is also proven to be more stable than other methods and is capable of producing more general feature subsets.