Video Files Research Articles

Postprocessing Method for Enhanced Left Ventricular Thrombus Detection in Echocardiography

Background and Objectives: The incidence of left ventricular thrombus has decreased in recent years due to advancements in reperfusion strategies for acute myocardial infarction and the use of medications to reduce ventricular remodeling. However, the accurate detection of thrombus remains crucial. Echocardiography is a primary diagnostic tool for thrombus detection, but in cases where the apex of the left ventricle is not clearly visualized, contrast injection is often required for diagnosis. We developed a postprocessing Left Ventricular Thrombus Detection Method (LVTDM) to enhance image details in the region of interest, enabling diagnosis without additional contrast injection. A purpose of our study is the evaluation of Left Ventricular Thrombus Detection Method. Materials and Methods: We analyzed echocardiography video files from 29 patients with apical wall motion abnormalities using LVTDM to identify the presence or absence of thrombus in the left ventricular apex. The results were verified with diagnoses obtained from the same echocardiography examinations following contrast injection. Our method demonstrated a sensitivity of 100% and a specificity of 83%, with a negative predictive value of 100% for ruling out thrombus. There was a strong correlation in thrombus detection/ruling out between LVTDM and contrast echocardiography. The Left Ventricular Thrombus Detection Method can be integrated into routine echocardiography examinations to effectively rule out thrombus when the apex is not clearly visualized. The implementation of this method has the potential to reduce the need for contrast injection by approximately half for detecting left ventricular thrombus.

CoFFEE: a codec-based forensic feature extraction and evaluation software for H.264 videos

The forensic analysis of digital videos is becoming increasingly relevant to deal with forensic cases, propaganda, and fake news. The research community has developed numerous forensic tools to address various challenges, such as integrity verification, manipulation detection, and source characterization. Each tool exploits characteristic traces to reconstruct the video life-cycle. Among these traces, a significant source of information is provided by the specific way in which the video has been encoded. While several tools are available to analyze codec-related information for images, a similar approach has been overlooked for videos, since video codecs are extremely complex and involve the analysis of a huge amount of data. In this paper, we present a new tool designed for extracting and parsing a plethora of video compression information from H.264 encoded files, including macroblocks structure, prediction residuals, and motion vectors. We demonstrate how the extracted features can be effectively exploited to address various forensic tasks, such as social network identification, source characterization, and double compression detection. We provide a detailed description of the developed software, which is released free of charge to enable its use by the research community to create new tools for forensic analysis of video files.

Demonstrating Brownian motion and estimating Avogadro’s number using simple tools: a practical guide for students

Abstract Brownian motion provides crucial observational evidence for the kinetic molecular theory by demonstrating the random perpetual movement of small particles. This phenomenon underscores the existence and behaviour of atoms and molecules, fundamentally supporting our current understanding of molecular kinetic theory. Remarkably, in 1905 Albert Einstein was able to predict this motion exactly purely from a theoretical point of view and in 1908 Jean Perrin provided experimental proof. This provided experimental proof of the existence of atoms and molecules. This paper outlines a method to demonstrate and record Brownian motion using a drop of homogenised low fat milk, a microscope and a mobile phone. The captured video files can be analysed with Physics Tracker software, allowing for the estimation of Avogadro’s number.

Enhancing Speech-to-Text and Translation Capabilities for Developing Arabic Learning Games: Integration of Whisper OpenAI Model and Google API Translate

This study tackles language barriers in computer-mediated communication by developing an application that integrates OpenAI’s Whisper ASR model and Google Translate machine translation to enable real-time, continuous speech transcription and translation and the processing of video and audio files. The application was developed using the Experimental method, incorporating standards for testing and evaluation. The integration expanded language coverage to 133 languages and improved translation accuracy. Efficiency was enhanced through the use of greedy parameters and the Faster Whisper model. Usability evaluations, based on questionnaires, revealed that the application is efficient, effective, and user-friendly, though minor issues in user satisfaction were noted. Overall, the Speech Translate application shows potential in facilitating transcription and translation for video content, especially for language learners and individuals with disabilities. Additionally, this study introduces an Arabic learning game incorporating an Artificial Neural Network using the CNN algorithm. Focusing on the “Speaking” skill, the game applies to voice and image extraction techniques, achieving a high accuracy rate of 95.52%. This game offers an engaging and interactive method for learning Arabic, a language often considered challenging. The incorporation of Artificial Neural Network technology enhances the effectiveness of the learning game, providing users with a unique and innovative language learning experience. By combining voice and image extraction techniques, the game offers a comprehensive approach to enjoyably improving Arabic speaking skills.

Image Steganography Approaches and Their Detection Strategies: A Survey

Steganography is the art and science of hidden (or covered) communication. In digital steganography, the bits of image, video, audio and text files are tweaked to represent the information to hide. This article covers the current methods for hiding information in images, alongside steganalysis methods that aim to detect the presence of steganography. By reviewing 456 references, this article discusses the different approaches that can be taken toward steganography and its much less widely studied counterpart. Currently in research older steganography approaches are more widely used than newer methods even though these show greater potential. New methods do have flaws; therefore, more research is needed to make these practically applicable. For steganalysis one of the greatest challenges is the generalisability. Often one scheme can detect the presence of one specific hiding method. More research is needed to combine current schemes and/or create new generalisable schemes. To allow readers to compare results between different papers in our work, performance indications of all steganalysis methods are outlined and a comparison of performance is included. This comparison is given using ‘topological sorting’ graphs, which compares detection results from all papers (as stated in the papers themselves) on different steganographic schemes.

Descemet Membrane Endothelial Keratoplasty Graft Preparation Using the Liquid Bubble Technique with Subtrabecular Hydrodissection: A Retrospective Real-Life Study

Background: Techniques for the preparation of grafts for Descemet membrane endothelial keratoplasty (DMEK) can be classified into those that involve the manual dissection of the Descemet membrane (DM) and those that use an injection of a liquid or a gas to achieve a separation of the DM from the posterior corneal stroma. The purpose of this study was to evaluate the efficiency of the liquid bubble technique. Methods: The success rate of the technique was calculated retrospectively using the operating reports. Video files for each graft preparation were retrieved and the time, number of injections, and number of injections sites required for the hydrodissection were measured. The number of cases in which a manual dissection of the Descemet membrane was necessary was recorded. Information on donor age and graft preservation time were retrieved from the eye bank file. Results: In 58 cases, the success rate was 98.3%. In the 28 procedures where a video was available, the median time for hydrodissection was 4.4 min. The median number of injection sites was 2, with a median number of injections of 3.5. Manual dissection as a rescue technique was performed in 25% of cases, with one case resulting in graft tears. The mean graft diameter was 7.6 mm. The mean donor age was 66 years, and the mean graft storage time was 22 days. Conclusions: The liquid bubble technique can be a fast and valuable choice for DMEK graft preparation, especially in centers where the tissue is prepared in the operating theater.

Video source identification using machine learning: A case study of 16 instant messaging applications

In recent years, there has been a notable increase in the prevalence of cybercrimes related to video content, including the distribution of illegal videos and the sharing of copyrighted material. This has led to the growing importance of identifying the source of video files to trace the owner of the files involved in the incident or identify the distributor. Previous research has concentrated on revealing the device (brand and/or model) that “originally” created a video file. This has been achieved by analysing the pattern noise generated by the image sensor in the camera, the storage structural features of the file, and the metadata patterns. However, due to the widespread use of mobile environments, instant messaging applications (IMAs) such as Telegram and Wire have been utilized to share illegal videos, which can result in the loss of information from the original file due to re-encoding at the application level, depending on the transmission settings. Consequently, it is necessary to extend the scope of existing research to identify the various applications that are capable of re-encoding video files in transit. Furthermore, it is essential to determine whether there are features that can be leveraged to distinguish them from the source identification perspective. In this paper, we propose a machine learning-based methodology for classifying the source application by extracting various features stored in the storage format and internal metadata of video files. To conduct this study, we analyzed 16 IMAs that are widely used in mobile environments and generated a total of 1974 sample videos, taking into account both the transmission options and encoding settings offered by each IMA. The training and testing results on this dataset indicate that the ExtraTrees model achieved an identification accuracy of approximately 99.96 %. Furthermore, we developed a proof-of-concept tool based on the proposed method, which extracts the suggested features from videos and queries a pre-trained model. This tool is released as open-source software for the community.

Video file example of HearScope video otoscopy.

Video file example of HearScope video otoscopy.

Video file example of Cupris TYM video otoscopy.

Video file example of Cupris TYM video otoscopy.

Optimizing the Transition: Strategies for Migrating On-Premise Storage to the Cloud

In recent years, Cloud Computing has had great relevance and has aroused interest in the Telecommunications industry, due to the reduction of costs in hardware acquisition, maintenance and operation personnel. One of the great advantages of cloud computing is the automation of processes that makes systems more robust and secure by limiting the margin of human and operational errors. It is of utmost importance to consider the large storage capacity management of video files that can provide solutions to media companies or related industries, this is key to think about a cost benefit analysis that can generate the migration and rental of these services in the cloud, with more reliable and accessible systems, this will help us to study hybrid technologies that offer on premise solutions. In addition to the above, when talking about companies that seek to digitally transform themselves to achieve agility and resilience, as well as to support future applications, it is necessary to mention their current infrastructures that have the capacity to host new technologies that require high performance, security and an adequate digital interconnection that allows the correct access to the acquired applications. In this research, an analysis of the existing process models for migration from on-premise storage to the cloud has been made with the objective of identifying the most important challenges, from planning, execution, factors to consider when assessing the complexity of applications and data before migrating to the cloud.

Automated analysis of a novel object recognition test in mice using image processing and machine learning

The novel object recognition test (NORT) is one of the most commonly employed behavioral tests in experimental animals designed to evaluate an animal's interest in and recognition of novelty. However, manual procedures, which rely on researchers’ observations, prevent high throughput analysis. In this study, we developed an automated analysis method for NORT utilizing machine learning-assisted exploratory behavior detection. We recorded the exploratory behavior of the mice using a video camera. The coordinates of the mouse nose and tail base in recorded video files were detected using a pre-trained machine learning model, DeepLabCut. Each video was then segmented into frame images, which were categorized into "exploratory,” or "non-exploratory" frames based on manual observation. Mouse feature vectors were calculated as vectors from the nose to the vertices of the object and were utilized for SVM training. The trained SVM effectively detected exploratory behaviors, showing a strong correlation with human observer assessments. Upon application to NORT, the duration of mouse exploratory behavior towards objects predicted by the SVM exhibited a significant correlation with the assessments made by human observers. The novelty discrimination index derived from the SVM predictions also aligned well with that from human observations.

Automated MRI Video Analysis for Pediatric Neuro-Oncology: An Experimental Approach

Over the past year, there has been a significant rise in interest in the application of open-source artificial intelligence models (OSAIM) in the field of medicine. An increasing number of studies focus on evaluating the capabilities of these models in image analysis, including magnetic resonance imaging (MRI). This study aimed to investigate whether two of the most popular open-source AI models, namely ChatGPT 4o and Gemini Pro, can analyze MRI video sequences with single-phase contrast in sagittal and frontal projections, depicting a posterior fossa tumor corresponding to a medulloblastoma in a child. The study utilized video files from single-phase contrast-enhanced head MRI in two planes (frontal and sagittal) of a child diagnosed with a posterior fossa tumor, type medulloblastoma, confirmed by histopathological examination. Each model was separately provided with the video file, first in the sagittal plane, analyzing three different sets of commands from the most general to the most specific. The same procedure was applied to the video file in the frontal plane. The Gemini Pro model did not conduct a detailed analysis of the pathological change but correctly identified the content of the video file, indicating it was a brain MRI, and suggested that a specialist in the field should perform the evaluation. Conversely, ChatGPT 4o conducted image analysis but failed to recognize that the content was MRI. The attempts to detect the lesion were random and varied depending on the plane. These models could not accurately identify the video content or indicate the area of the neoplastic change, even after applying detailed queries. The results suggest that despite their widespread use in various fields, these models require further improvements and specialized training to effectively support medical diagnostics.

Beyond PhacoTrainer: Deep Learning for Enhanced Trabecular Meshwork Detection in MIGS Videos.

The purpose of this study was to develop deep learning models for surgical video analysis, capable of identifying minimally invasive glaucoma surgery (MIGS) and locating the trabecular meshwork (TM). For classification of surgical steps, we had 313 video files (265 for cataract surgery and 48 for MIGS procedures), and for TM segmentation, we had 1743 frames (1110 for TM and 633 for no TM). We used transfer learning to update a classification model pretrained to recognize standard cataract surgical steps, enabling it to also identify MIGS procedures. For TM localization, we developed three different models: U-Net, Y-Net, and Cascaded. Segmentation accuracy for TM was measured by calculating the average pixel error between the predicted and ground truth TM locations. Using transfer learning, we developed a model which achieved 87% accuracy for MIGS frame classification, with area under the receiver operating characteristic curve (AUROC) of 0.99. This model maintained a 79% accuracy for identifying 14 standard cataract surgery steps. The overall micro-averaged AUROC was 0.98. The U-Net model excelled in TM segmentation with an Intersection over union (IoU) score of 0.9988 and an average pixel error of 1.47. Building on prior work developing computer vision models for cataract surgical video, we developed models that recognize MIGS procedures and precisely localize the TM with superior performance. Our work demonstrates the potential of transfer learning for extending our computer vision models to new surgeries without the need for extensive additional data collection. Computer vision models in surgical videos can underpin the development of systems offering automated feedback for trainees, improving surgical training and patient care.

A new approach to video summary generation

This paper describes a video summary approach for an actual video file that provides a rapid way to assess its information. This is a difficult task to identify relevant or informative sections. When the original video demands knowledge of its content, we build a video summary via extracting in-depth characteristics of each original video. Video summarization deals with the generation of a condensed version of the original video by including meaningful frames or segments while eliminating redundant information provides a concise and accurate representation of the original video clips by providing the most representative summary with increased interest through the application of a matching function between the current and prior frames. The primary goal of this work is to give a strategy for recording video clips using security cameras and converting them into a short video to provide a clear analysis of the video by reducing excess and extracting the key video contents into the real video. In a lengthy video clip, viewers may not have enough time to see the full thing. The viewer may be interested in seeing a specific topic under the label "Important". It may be composed of pieces that may be shared on social media and communicate the semantics of long videos, particularly video snippets that must be submitted to Internet. To browse, will be needed a network with adequate bandwidth. That is, non-affiliated video streaming, indexing, and retrieval of video content makes it valuable to viewers.

Quality Assessment of TikTok as a Source of Information About Mitral Valve Regurgitation in China: Cross-Sectional Study.

In China, mitral valve regurgitation (MR) is the most common cardiovascular valve disease. However, patients in China typically experience a high incidence of this condition, coupled with a low level of health knowledge and a relatively low rate of surgical treatment. TikTok hosts a vast amount of content related to diseases and health knowledge, providing viewers with access to relevant information. However, there has been no investigation or evaluation of the quality of videos specifically addressing MR. This study aims to assess the quality of videos about MR on TikTok in China. A cross-sectional study was conducted on the Chinese version of TikTok on September 9, 2023. The top 100 videos on MR were included and evaluated using quantitative scoring tools such as the modified DISCERN (mDISCERN), the Journal of the American Medical Association (JAMA) benchmark criteria, the Global Quality Score (GQS), and the Patient Education Materials Assessment Tool for Audio-Visual Content (PEMAT-A/V). Correlation and stepwise regression analyses were performed to examine the relationships between video quality and various characteristics. We obtained 88 valid video files, of which most (n=81, 92%) were uploaded by certified physicians, primarily cardiac surgeons, and cardiologists. News agencies/organizations and physicians had higher GQS scores compared with individuals (news agencies/organizations vs individuals, P=.001; physicians vs individuals, P=.03). Additionally, news agencies/organizations had higher PEMAT understandability scores than individuals (P=.01). Videos focused on disease knowledge scored higher in GQS (P<.001), PEMAT understandability (P<.001), and PEMAT actionability (P<.001) compared with videos covering surgical cases. PEMAT actionability scores were higher for outpatient cases compared with surgical cases (P<.001). Additionally, videos focused on surgical techniques had lower PEMAT actionability scores than those about disease knowledge (P=.04). The strongest correlations observed were between thumbs up and comments (r=0.92, P<.001), thumbs up and favorites (r=0.89, P<.001), thumbs up and shares (r=0.87, P<.001), comments and favorites (r=0.81, P<.001), comments and shares (r=0.87, P<.001), and favorites and shares (r=0.83, P<.001). Stepwise regression analysis identified "length (P<.001)," "content (P<.001)," and "physicians (P=.004)" as significant predictors of GQS. The final model (model 3) explained 50.1% of the variance in GQSs. The predictive equation for GQS is as follows: GQS = 3.230 - 0.294 × content - 0.274 × physicians + 0.005 × length. This model was statistically significant (P=.004) and showed no issues with multicollinearity or autocorrelation. Our study reveals that while most MR-related videos on TikTok were uploaded by certified physicians, ensuring professional and scientific content, the overall quality scores were suboptimal. Despite the educational value of these videos, the guidance provided was often insufficient. The predictive equation for GQS developed from our analysis offers valuable insights but should be applied with caution beyond the study context. It suggests that creators should focus on improving both the content and presentation of their videos to enhance the quality of health information shared on social media.

The Effect of In-person and Online Training on Learning Jump Rope Skills in Girl Students

Background: In situations with physical limitations, combining online and in-person methods with a greater focus on practical exercises, feedback, and maintaining motivation and commitment to learning may be the best approach. Objectives: The aim of the present study was to compare the effectiveness of two educational methods— in-person training and online training— on learning jump rope skills. Methods: Forty fourth-grade girl students were randomly assigned to two groups: In-person training and online training. Initially, participants took a pre-test that involved performing the basic single jump technique (BSJT) for 30 seconds. The online class was conducted on Iran's official educational platform, a messaging service that allows for the transmission of audio, video, and text files, as well as live calls. The in-person group received all instructions with the same content in person. Immediately after completing eight 45-minute instructional sessions, a post-test similar to the pre-test was administered. One week after the acquisition sessions, a retention test for 10 other jump rope techniques was conducted within a 30-second time frame, and the average score was calculated for each participant. Results: The results showed a significant main effect of the group, with the overall performance of the in-person group being better than that of the online group. Participants' scores in BSJT improved from the pre-test to the post-test. A significant difference was observed between the in-person and online groups in the retention test. Conclusions: In-person training appears to be generally more effective than online training for improving jump rope skills. While online training can be effective for mastering and consolidating previously learned skills, in-person training is preferable for learning and mastering more complex techniques.

Multi-Directional Long-Term Recurrent Convolutional Network for Road Situation Recognition.

Understanding road conditions is essential for implementing effective road safety measures and driving solutions. Road situations encompass the day-to-day conditions of roads, including the presence of vehicles and pedestrians. Surveillance cameras strategically placed along streets have been instrumental in monitoring road situations and providing valuable information on pedestrians, moving vehicles, and objects within road environments. However, these video data and information are stored in large volumes, making analysis tedious and time-consuming. Deep learning models are increasingly utilized to monitor vehicles and identify and evaluate road and driving comfort situations. However, the current neural network model requires the recognition of situations using time-series video data. In this paper, we introduced a multi-directional detection model for road situations to uphold high accuracy. Deep learning methods often integrate long short-term memory (LSTM) into long-term recurrent network architectures. This approach effectively combines recurrent neural networks to capture temporal dependencies and convolutional neural networks (CNNs) to extract features from extensive video data. In our proposed method, we form a multi-directional long-term recurrent convolutional network approach with two groups equipped with CNN and two layers of LSTM. Additionally, we compare road situation recognition using convolutional neural networks, long short-term networks, and long-term recurrent convolutional networks. The paper presents a method for detecting and recognizing multi-directional road contexts using a modified LRCN. After balancing the dataset through data augmentation, the number of video files increased, resulting in our model achieving 91% accuracy, a significant improvement from the original dataset.

Measurement of the on-road wind environment in channelled roadways

A road test was undertaken to characterize the complex wind environment experienced by vehicles in channelled roadways. Channelled roadways consist of highways that are bordered by forests, hills, or other terrain features. A test vehicle was outfitted with three fast-response pressure probes in front of the vehicle located at 1 m, 2 m, and 3 m above the ground. Measurements were acquired on three highway routes in Canada with varying degrees of surrounding terrain, roadside forest cover, and in different wind conditions. Reference wind conditions were established using a network of federally-managed weather stations. The study identified that the effects of terrain and roadside forests can cause changes in wind direction of more than 50∘ to 90∘ relative to nearby reference weather stations. Directional channelling of winds was observed in forested and complex terrain, where changes in the wind direction mimicked the changes in highway alignment. The on-road video files were analysed using a machine-learning algorithm to quantitatively assess road-side forests and terrain and to identify fully-channelled sections of highway. Generalized channelled-flow trends were identified by considering the wind conditions in the channelled sections of all three routes. A linear relationship between the flow direction in the channel to that of the reference conditions was observed when the reference wind direction and the highway was within ±50∘. As the reference wind direction approached perpendicularity to the channelled segments, greater variations in the on-road wind direction were observed due to the unsteady flow within the channel. On-road wind speeds on each route were a maximum when the reference wind direction was approximately aligned with the highway and decreased to approximately 20% of the reference wind speed when the reference wind was perpendicular to the channel.

Multi-level authentication for security in cloud using improved quantum key distribution

ABSTRACT Cloud computing is an on-demand virtual-based technology to develop, configure, and modify applications online through the internet. It enables the users to handle various operations such as storage, back-up, and recovery of data, data analysis, delivery of software applications, implementation of new services and applications, hosting websites and blogs, and streaming of audio and video files. Thereby, it provides us many benefits although it is backlashed due to problems related to cloud security like data leakage, data loss, cyber attacks, etc. To address the security concerns, researchers have developed a variety of authentication mechanisms. This means that the authentication procedure used in the suggested method is multi-levelled. As a result, a better QKD method is offered to strengthen cloud security against different types of security risks. Key generation for enhanced QKD is based on the ABE public key cryptography approach. Here, an approach named CPABE is used in improved QKD. The Improved QKD scored the reduced KCA attack ratings of 0.3193, this is superior to CMMLA (0.7915), CPABE (0.8916), AES (0.5277), Blowfish (0.6144), and ECC (0.4287), accordingly. Finally, this multi-level authentication using an improved QKD approach is analysed under various measures and validates the enhancement over the state-of-the-art models

Editorial

The Alpha Centauri Scientific and Technological Research Journal is a peer-reviewed journal produced internationally, it is published QUARTERLY in online version ISSN: 2709-4502. Alpha Centauri publishes scientific articles specializing in Applied Social Sciences and Social Sciences in Information Technologies with great impact at an international level. We are currently trending in more than 24 countries disseminating Open Access scientific knowledge. IMAGES: Still images, audio recordings, video and related computer files for non-commercial use. NASA content (images, audio, video, and computer files used in the reproduction of three-dimensional models, such as texture maps and polygonal data in any format).