International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 11 Issue: 11 | Nov 2024
www.irjet.net
p-ISSN: 2395-0072
Early Detection of Anemia : A well-developed system using machine learning model Belnekar Janhavi1, Dube Anuj Soumendra2, Kothari Moksha3 1,2,3 Student,
Information Technology, Thakur College of Engineering and Technology Mumbai, India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract
By combining the power of machine learning, image analysis and mobile technology, we present a viable solution that has the potential to revolutionize anemia screening and diagnosis, particularly in resourceconstrained settings. This research holds the promise of significantly improving health care access and outcomes for individuals affected by anemia and offers a ray of hope in the ongoing global battle against this pervasive health problem.
Anemia, characterized by a deficiency of red blood cells or hemoglobin, remains a global public health concern, particularly in resource-limited regions where access to advanced diagnostic tools is limited. The essence of this work lies in the comprehensive evaluation and analysis of ML models like Convolutional Neural Networks, Logistic Regression, and Gaussian Blur algorithm on publicly available dataset of 710 images of the conjunctiva for pallor analysis. This endeavor aims to furnish the ongoing efforts to improve anemia detection and healthcare access, especially in underserved communities. The implications of this work extend to early intervention and prevention of anemia-related health complications.
II.
In recent research, computerized algorithms, particularly machine learning, have shown high accuracy in estimating hemoglobin (Hb) levels and diagnosing medical conditions like anemia # [1]. Various algorithms, including support vector machines (SVM), k-nearest neighbors (k-NN), Bayesian networks, artificial neural networks (ANN), and decision tree classifiers, have been employed for classification. Studies using conjunctiva images of the eyes have demonstrated that the choice of algorithm depends on the specific problem domain # [2]. Non-invasive techniques utilizing clinical symptoms based on fingernails and palm images, as well as conjunctiva images, have shown promise.Different studies have explored the use of machine learning techniques such as ANN, SVM, decision tree, Naïve Bayes, K-NN, and rule-based approaches for anemia detection.The conjunctiva of the eye has been a focus, with SVM achieving 78.90% accuracy # [3], LS-SVM reaching 85% precision, and 98.96% accuracy. Deep learning approaches, including a three-tier deep convolutional fused network, achieved accurate anemia severity prediction.
Keywords—pallor analysis, ML models, healthcare access I.
INTRODUCTION
This research paper presents a task-critical implementation of Convolutional Neural Networks (CNN), Logistic Regression and Gaussian Blur algorithm in computer vision. CNNs are a core deep learning architecture widely used in computer vision. They are designed to automatically and adaptively learn hierarchical features from data, making them highly effective for tasks such as image classification, object detection, and image segmentation. Deep learning techniques have consistently demonstrated significant progress and efficiency in the field of semantic segmentation.Anemia, moving nearly 33% of the global state, presents a important fitness challenge, particularly with youngsters and pregnant daughters. Iron imperfection stands as a chief cause, contributing to 42% of emptiness cases. The prompt discovery of anemia is critical to check the risk of irrevocable organ damage. This study intends a creative non-obtrusive approach, utilizing machine intelligence algorithms to resolve images of palms, fingernails, and conjunctiva. The determined search out overcome support limitations guide unoriginal chlorosis detection orders.
© 2024, IRJET
|
Impact Factor value: 8.315
LITERATURE SURVEY
By utilizing non-invasive methods, healthcare professionals can gain new insights into understanding individual emotional states, leveraging this information for personalized treatment plans. Moreover, the utilization of advanced technologies like the three-tier deep convolutional fused network highlights the immense power of deep learning approaches in predicting the severity of anemia, potentially revolutionizing the early detection and management of this condition. Together, these advancements hold promise for improving diagnosis and treatment outcomes in various medical fields.
|
ISO 9001:2008 Certified Journal
|
Page 87