Predicting Anemia in Patients Through Clinical and Hematological Data Using Machine Learning Algorithms

Abstract

Anemia is a prevalent condition affecting millions of individuals worldwide. Notably, Peru has witnessed a concerning increase in anemia cases. This study aims to delves into an extensive analysis of the role of machine learning within healthcare, particularly in the early identification and management of anemia using maching learning. The proposed Methodology consists of five phases: Obtained the dataset, Preprocessing, Model Implementations, Evaluation and Validation. In the proposed method, anemia detection is performed by reviewing relevant studies, encompassing Rodríguez's data analysis methodology, Gómez's exploration of AI applications, Hernández's optimization-based SVM model, Díaz's email classification, and Dávila's algorithms for discerning between healthy and anemic blood samples. The approach involves comprehensive data collection, meticulous preprocessing, feature extraction employing techniques such as cell count and crucial factor identification, and data classification using a range of machine learning algorithms. The results highlight the Artificial Neural Network model as the most optimal, achieving a noteworthy accuracy rate of 90.3\% in correctly identifying anemic blood samples. Finally, this study underscores the pivotal role of machine learning techniques in advancing the diagnosis and treatment of anemia, especially in regions experiencing escalated incidences like Peru, offering promising avenues for addressing this pressing global health concern. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.