Malaysian Journal of Computer Science

A HYBRID MODEL FOR CLASSIFICATION OF TUBERCULOSIS CHEST X-RAYS IMAGES

2024-08-16T08:06:42+08:00

Tuberculosis (TB), a grave infectious disease affecting millions globally, is often diagnosed using chest X-rays. For accurate diagnosis, especially for detecting early stage, medical practitioners require the assistance of advanced technologies. In contrast to existing models, which focus largely on TB detection in the images, the proposed work aims to classify the images affecting TB such that an appropriate method can be chosen for accurate chest TB detection in chest X-ray images. Thus, we aim to combine the powerful features of the VGG16 architecture with a convolutional neural network (CNN) for classification purposes. Drawing inspiration from VGG16, known for its effective method of capturing essential image information, we aim to modify VGG16 for feature extraction to identify signs of tuberculosis (TB) in images. For the classification task, we employ a CNN to categorize images impacted by TB. Our proposed technique is evaluated on a standard dataset, demonstrating its superiority over current leading methods in accuracy, recall, and precision.

FINE VESSEL SEGMENTATION WITH REFINEMENT GATE IN DEEP LEARNING ARCHITECTURES

2024-10-09T09:43:08+08:00

Automated vessel segmentation is essential in diagnosing eye-related disorders and monitoring progressive retinal diseases. State-of-the-art methods have achieved excellent results in this field, but very few have considered the post-processing of feature maps. As a result, there is often a lack of small and fine vessels or discontinuities in segmented vessels. To address this issue, this study introduces a novel post-processing technique called the refinement gate, which works with a deep learning model during training. The refinement gate enhances contextual information to extract important features from feature maps better. The proposed technique is applied with U-net architecture and placed after every convolution block in the encoder path. Visual and statistical comparisons demonstrate the robustness of the proposed method using three publicly available datasets, namely: the DRIVE DB, the STARE DB, and CHASE_DB1 datasets, showing significant improvements to segment weak and tiny vessels. The reported results confirm the potential of the model to be used as a segmentation tool in the medical field. This study is the first to propose such a gating mechanism without additional trainable parameters or standalone networks as in other literature.

AN ENHANCED META-CLASSIFIER APPROACH FOR ALCOHOL ADDICTION PREDICTION

2024-11-14T20:15:32+08:00

Chronic alcohol consumption poses significant public health challenges globally. In underserved regions, the lack of AI-based interventions for alcohol addiction highlights a critical gap in the healthcare system, particularly regarding the early detection of alcohol abuse. Henceforth, this research aims to raise awareness of alcohol use disorder and proposes a novel AI-powered solution designed with an improved classification algorithm to address this deficiency, with a primary focus on a cutting-edge prediction model. This research shifts the current reactive approach in alcohol addiction intervention to proactive approach by employing an enhanced meta-classification algorithm (EMC) that focuses on improving the interpretability, efficiency, and accuracy of predictions. The proposed EMC ultimately provides a robust tool for healthcare professionals and patients which fosters more effective and personalized intervention strategies for alcohol addiction recovery. The results demonstrate a remarkable 10.13% improvement in balanced accuracy and a 9.72% enhancement in the area under the curve compared to traditional ensemble and state-of-the-art methods. Thus, findings from this study will assist medical practitioners and policymakers in developing evidence-based strategies to combat alcoholism and enhance public health outcomes. By deriving insights from real-world case study, the outcome of this research represents a pioneering effort to betterment of healthcare in underserved regions, offering a low-cost, scalable solution for early detection, and has the potential to significantly improve outcomes in marginalized communities.

HUMAN ACTIVITY RECOGNITION BASED ON DEVICE-FREE WI-FI SENSING: A COMPREHENSIVE REVIEW

2024-11-14T20:40:44+08:00

Within the context of the current era’s research and advancement in technology, in recent years, the Internet of Things (IoT) has been consistent in the development of applications in several fields like smart cities, smart homes, smart grids, smart agriculture, and so on. Most of the existing research in human activity recognition relies on vision-based, wearable devices, object-tagged, and sensor-based approaches. Despite the superior performance of these approaches, a number of issues have arisen related to the invasion of privacy, light dependency, cost effectiveness, and feasibility. Moreover, these approaches require domain knowledge of different tasks, which may make them complicated for practical deployment. Wi-Fi technology, on the other hand, offers robust possibilities in indoor and outdoor environments for recognizing applications and, combined with a few significant features, makes it a far more attractive option compared to other sensing technologies. Hence, the device-free Wi-Fi-sensing approach is more practical in the smart home environment as it does not require the targeted human to have any device for day-to-day activities. This article’s contributions can be summarized as, first, providing primary knowledge on a wireless LAN, the Wi-Fi sensing model; second, sharing the findings of the comprehensive survey scrutinizing the latest developments in human activity or gesture recognition systems based on device-free Wi-Fi sensing systems; third, sharing the analysis of the limitations and key research challenges that need to be addressed in order to expand the device-free Wi-Fi sensing system; and lastly, a discussion and future directions of existing device-free Wi-Fi sensing techniques.