International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 04 | Apr 2024
p-ISSN: 2395-0072
www.irjet.net
Performance Analysis of Efficient Storage and Retrieval of Medical Images Using Deep Learning Techniques Subhajit Das, R. R. Sedamkar ME Student, Dept. of Computer Engineering, Thakur College of Engineering and Technology, Maharashtra, India Professor, Dept. of Computer Engineering, Thakur College of Engineering and Technology, Maharashtra, India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The utilization of medical imaging has seen an
manual processing would be time-consuming, labourintensive, and highly inefficient.The storage and retrieval of medical images using deep learning involves employing advanced neural network architectures to efficiently store, organize, and retrieve vast amounts of medical imaging data. Deep learning models, such as convolutional neural networks (CNNs), are trained on labeled image datasets to extract meaningful features and patterns from medical images. Additionally, deep learning techniques facilitate image compression, reducing storage requirements while maintaining diagnostic quality. By integrating deep learning algorithms into storage and retrieval systems, healthcare professionals can access and analyze medical images swiftly, aiding in diagnosis, treatment planning, and research endeavors.
exponential increase in recent years, leading to a substantial volume of digital medical images being generated and stored. The effective management of this massive data is crucial for timely diagnosis and treatment. This project focuses on leveraging deep learning techniques to address the challenges of efficient storage and retrieval of these medical images. The project aims to develop a system that employs deep learning algorithms to optimize the storage and retrieval process of medical images. This involves the utilization of convolutional neural networks (CNNs) and advanced image processing techniques to analyse, compress, and categorize medical images. The need for efficient storage and quick retrieval of medical images is paramount in healthcare settings. Traditional storage methods often face challenges in scalability, accessibility, and speed. By implementing deep learning algorithms, this project seeks to alleviate these challenges, facilitating rapid access to pertinent medical image data while conserving storage space without compromising diagnostic quality. Furthermore, the system aims to improve patient care by enabling quick access to previous diagnostic results and facilitating collaboration among healthcare providers. Its objective includes implementing user-friendly interfaces for seamless interaction, ensuring robust data storage and retrieval mechanisms, and integrating advanced machine learning techniques for accurate image analysis. Ultimately, the system's goal is to contribute to the advancement of medical imaging practices, ultimately leading to better patient outcomes and healthcare delivery.
2.LITERATURE SURVEY “Deep neural networks automatically detect osteoporotic vertebral fractures on CT scans,” the article reads. Osteoporotic vertebral fractures (OVFs) are common in the elderly and are associated with significant personal and socioeconomic suffering. Early diagnosis and treatment of OVF is important to prevent further damage and morbidity. However, OVFs are often underdiagnosed and underreported on computed tomography (CT) because they can be asymptomatic in their early stages. In this article, we propose and evaluate an electronic device that can detect OVFs in chest, abdomen and pelvis CT examinations at the level of radiologists. Our OVF detection system uses deep convolutional neural networks (CNN) to extract radiographic images from each slice of the CT scan. This extraction process was performed from the collected samples to perform the final analysis of the entire CT scan. In this work, we investigate different clustering methods, including the use of short-term temporal (LSTM) networks. We trained and evaluated our system on 1,432 CT scans, including 10,546 sagittal twodimensional (2D) images. Our system achieved 89.2% accuracy and 90.8% F1 score when we evaluated 129 CT scan indexes, which were semi-structured and created as a reference model in many respects. The results of our system were based on the performance of radiologists working on this test set in a real clinical setting. We hope that the proposed method will aid and improve the diagnosis of OVF in the clinic by pre-screening the CT
Key Words: (Size 10 & Bold) Key word1, Key word2, Key word3, etc (Minimum 5 to 8 key words)…
1. INTRODUCTION As the biomedical field advances, various sectors increasingly utilize biomedical signals captured through imaging for multiple purposes. This encompasses the generation of a substantial volume of images. These images play a pivotal role in supporting healthcare professionals by swiftly and accurately localizing anatomical lesions and gauging disease progression. However, the sheer volume of image data poses a challenge, especially given the limited number of experienced physicians. Consequently, relying solely on
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 2247