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

Thu, 01 Aug 2024 00:00:00 +0800

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

Thu, 01 Aug 2024 00:00:00 +0800

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.

Malaysian Journal of Computer Science

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

FINE VESSEL SEGMENTATION WITH REFINEMENT GATE IN DEEP LEARNING ARCHITECTURES