Claim Missing Document
Check
Articles

Found 4 Documents
Search

Comparation between Fibroscan and Hepatus for Detecting NAFLD in Patients with Metabolic Dysregulation Sun, Jie-hui; Zhang, Ping-ping; Wang, Kun; Xu, Miao; Sun, Jing; Li, Li
The Indonesian Journal of Gastroenterology, Hepatology, and Digestive Endoscopy Vol 25, No 1 (2024): VOLUME 25, NUMBER 1, April, 2024
Publisher : The Indonesian Society for Digestive Endoscopy

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24871/251202411-20

Abstract

Background The clinical application of the innovative instantaneous elasticity and fat attenuation measurement technology, Hepatus, is currently in the evaluation stage. This study aimed to compare the detection performance of Fibroscan and Hepatus in patients with metabolic dysregulation who are "at-risk" for non-alcoholic fatty liver disease (NAFLD).Methods Between January 2021 and April 2021, 149 patients were enrolled in this study. Clinical data were collected, and all patients underwent both Fibroscan and Hepatus assessments to determine liver stiffness measurement (LSM) and attenuation parameters. The correlation between the results obtained from the two transient elastography (TE) devices was analyzed. Receiver operating characteristic curves (ROC) were constructed to compare the diagnostic value of Fibroscan and Hepatus for Hepatic Steatosis Index (HSI)-based NAFLD.Results The detection success rate of Hepatus (100.0%) was higher than that of Fibroscan (96.0%). LSM (r = 0.663, P0.05) and attenuation parameters (r = 0.778, P0.05) obtained by Fibroscan and Hepatus were significantly correlated. Hepatus tended to produce a higher LSM (Hepatus vs. Fibroscan: 6.04 vs. 5.66 kPa, P=0.016) but a lower attenuation parameter than Fibroscan (Hepatus vs. Fibroscan: 264 vs. 277 dB/m, P0.001). The area under the ROC curve for detecting HSI-based NAFLD was 0.811 for Fibroscan and 0.832 for Hepatus.Conclusion Measurements obtained by Fibroscan and Hepatus are strongly correlated, and the diagnostic value of the two TE devices is comparable in detecting HSI-based NAFLD. Hepatus offer a potential TE alternative in NAFLD examination.
An Effective Investigation of Genetic Disorder Disease Using Deep Learning Methodology Vidhya, B.; Shivakumar, B. L.; Maidin, Siti Sarah; Sun, Jing
Journal of Applied Data Sciences Vol 5, No 3: SEPTEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i3.370

Abstract

This study evaluates the performance of four neural network models—Artificial Neural Network (ANN), ANN optimized with Artificial Bee Colony (ANN-ABC), Multilayer Feedforward Neural Network (MLFNN), and Forest Deep Neural Network (FDNN)—across different iteration levels to assess their effectiveness in predictive tasks. The evaluation metrics include accuracy, precision, Area Under the Curve (AUC) values, and error rates. Results indicate that FDNN consistently outperforms the other models, achieving the highest accuracy of 99%, precision of 98%, and AUC of 99 after 250 iterations, while maintaining the lowest error rate of 2.8%. MLFNN also shows strong performance, particularly at higher iterations, with notable improvements in accuracy and precision, but does not surpass FDNN. ANN-ABC offers some improvements over the standard ANN, yet falls short compared to FDNN and MLFNN. The standard ANN model, though improving with iterations, ranks lowest in all metrics. These findings highlight FDNN's robustness and reliability, making it the most effective model for high-precision predictive tasks, while MLFNN remains a strong alternative. The study underscores the importance of model selection based on performance metrics to achieve optimal predictive accuracy and reliability. 
Empirical Study of the Correlation between Social Media Content and Health Issues among College Students Using Machine Learning Hemalatha, M.; Maidin, Siti Sarah; Sun, Jing
Journal of Applied Data Sciences Vol 5, No 4: DECEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i4.365

Abstract

This study analyzes the effect of social media content on college student addiction using data science techniques. It aims to examine the correlation between different types of social media content and addictive behavior in college students. The research methodology used is non-probability sampling with a sample size of 587 college students in Tamil Nadu, India. The study uses statistical tools such as correlation analysis, regression analysis, one-way ANOVA, and Friedman ranking test to analyze the data collected. The findings suggest that the factors influencing social media addiction are positively correlated with the health issues faced by college students. The study indicates that demographic variables such as age, gender, year in college, and place of living may play a role in shaping an individual's perception of social media addiction. The results of the study can inform the development of interventions and prevention strategies to reduce social media addiction among college students. The study recommends a multi-pronged approach to address the root causes of addiction and provide students with the tools and resources they need to manage their social media use and promote their physical and mental health.
An Adaptive Cuckoo Search Algorithm with Deep Learning for Addressing Cyber Security Problem Jeyaboopathiraja, J.; Mariajohn, Princess; Maidin, Siti Sarah; Sun, Jing
Journal of Applied Data Sciences Vol 5, No 4: DECEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i4.366

Abstract

IoT (Internet of Things) offers continued services to organizations by connecting systems, application and services using the medium of internet. They also leave themselves open to threats including virus attacks and software thefts where the risks of losing crucial information are high. These threats harm both the business’ finances and reputation. This work offers a combined Deep Learning strategy using Artificial Neural Networks that can assist in detecting illegal software and malware tainted files. The proposed cyber security architecture uses data mining techniques to forecast cyber-attacks and prepare Internet of Things for suitable countermeasures. This framework uses two phases namely detections and predictions. This paper proposes Adaptive Cuckoo Search Optimization-based Algorithms for cloud network routes. Adaptive Cuckoo Search Algorithm are a bio-inspired protocol based on cuckoo birds’ characteristics. Artificial Neural Networks classify assaults on cloud environments. The major goal of this work is to separate malicious servers from legitimate servers that are impacted by Denial of Service and Distributed Denial of Service assaults and thus safeguard server data and ensuring they are sent to legitimate servers. The outcome from this research proposed scheme shows better performances for protecting systems from cyber-attacks in terms of values for accuracy, Precision, Recall and F1-Measure when compared to existing algorithms.