International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 11 Issue: 05 | May 2024
www.irjet.net
p-ISSN: 2395-0072
Pneumonia Detection using Machine Learning Sahil Sawant1, Prof. Mario Pinto2 1Student, Department of Information Technology and Engineering, Goa College of Engineering, Farmagudi, Goa,
India
2Assistant Professor, Department of Information Technology and Engineering, Goa College of Engineering,
Farmagudi, Goa, India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract- Pneumonia, a prevalent and potentially life-
affected cases, ensuring comprehensive model training and evaluation.
threatening respiratory infection, poses significant challenges in timely and accurate diagnosis. Leveraging advancements in machine learning (ML) techniques, this research aims to develop an efficient and reliable pneumonia detection system. Through the analysis of chest X-ray images, a convolutional neural network (CNN) model is trained to differentiate between pneumonia-infected and healthy lung images. The dataset used for training and validation comprises a diverse set of chest X-ray images collected from various sources. The proposed model demonstrates promising results, achieving high accuracy and sensitivity in pneumonia detection. Moreover, interpretability techniques are employed to elucidate the decision-making process of the CNN model, enhancing its clinical relevance and trustworthiness. The developed system holds considerable potential for aiding healthcare professionals in prompt and accurate pneumonia diagnosis, thereby facilitating timely intervention and improving patient outcomes.
The methodology involves extracting relevant image features from chest X-rays, encompassing texture, intensity, and shape descriptors. To address challenges related to dataset variability and generalization, feature engineering and selection techniques are employed. The research aims to showcase the potential of machine learning models in pneumonia detection, with a focus on practical applicability in clinical settings. The ultimate goal is to provide healthcare professionals with a reliable tool that aids in the early identification of pneumonia cases, contributing to improved patient outcomes. This project contributes to the broader field of medical diagnostics, highlighting the promising role of machine learning in enhancing pneumonia detection capabilities. As technology continues to intersect with healthcare, the findings from this study aim to pave the way for the integration of machine learning models into real-world clinical workflows, supporting healthcare professionals in making timely and accurate diagnostic decisions for pneumonia.
Key Words: Pneumonia detection, Convolutional Neural Networks, Deep Learning, Chest X-ray images
1.INTRODUCTION Pneumonia is a leading cause of morbidity and mortality worldwide, particularly among children and the elderly. Early detection and prompt treatment of pneumonia are crucial for preventing complications and improving patient outcomes. Chest X-ray imaging is commonly used for the diagnosis of pneumonia due to its accessibility and effectiveness in detecting abnormalities in lung tissue. However, the interpretation of chest X-ray images can be challenging and time-consuming, requiring expertise from radiologists. Automated methods based on deep learning techniques offer a promising solution to streamline the diagnosis process and improve the efficiency and accuracy of pneumonia detection.
2. Related Works The author produced and showcase a merged DL model for identifying Pneumonia patients from CXR. In the proposed model, three distinct models are trained on the CXR dataset. The first of them is a bespoke CNN model. Xception and EfficientNetB4 are the two other models. Several data augmentation and preprocessing strategies are utilized, along with hyperparameter tuning. A composite model is generated by giving different trained models weights based on their accuracy and recall rates.Several performance metrics are improved compared to the prior art, thanks to the suggested approach. [1]
The project focuses on developing a machine learningbased system for the early and accurate detection of pneumonia from chest X-ray images. Pneumonia, a prevalent respiratory infection, requires timely diagnosis for effective medical intervention. The study utilizes a diverse dataset containing both normal and pneumonia-
© 2024, IRJET
|
Impact Factor value: 8.226
In this paper ensemble of 3 CNN model was used Deep transfer learning is used to deal with the data shortage Dataset is collected which is than pre-processed Image enhancement is done utilizing threshold LBP (Local binary pattern) feature extraction done Data augmentation to form a new data point and split into two datasets. [2]
|
ISO 9001:2008 Certified Journal
|
Page 275