Abstract: |
Sickle cell disease (SCD) presents a significant health challenge with diverse clinical manifestations. Early and accurate prediction of the onset and severity of co-morbidities in SCD is vital for improving outcomes. In this study, we employ advanced healthcare informatics, and machine learning techniques to analyze longitudinal blood pathology data. By focusing on crucial hematological parameters, we gain valuable insights into SCD’s pathophysiology. Additionally, incorporating spectroscopic insights into the study unveils molecular details, enriching the understanding of the disease’s complexity and paving the way for more nuanced and targeted interventions. Utilizing this data, we construct predictive models enabling personalized interventions and advancing precision healthcare management. The research revealed that Random Forest outperforms other algorithms, achieving an accuracy of 88%, recall of 82%, and specificity of 92%. This robust evaluation underscores the model’s reliability in predicting both positive and negative instances. These findings offer a promising pathway for enhancing disease prediction, management, and treatment planning, providing invaluable guidance for clinical practice in the context of sickle cell disease. |