Mucormycosis Predictive Analysis Using Machine Learning Techniques

The emergence of mucormycosis, commonly known as black fungus, as a post-COVID-19 complication has led to a surge in morbidity and mortality rates across the globe. In this research paper, we present a comprehensive analysis utilizing machine learning techniques, specifically Support Vector Machine (SVM) and Logistic Regression (LR), to discern predictive patterns and risk factors associated with the incidence and progression of mucormycosis. The study incorporates a dataset comprising clinical records, demographic information, comorbidities, medication history, and laboratory findings of patients diagnosed with mucormycosis post-COVID-19 infection. Feature engineering and selection methods are employed to extract relevant variables, while SVM and LR models are trained and validated using a cross-validation technique to predict the likelihood and severity of mucormycosis infection. Our findings reveal significant predictive capabilities of SVM and LR models in identifying high-risk individuals predisposed to mucormycosis, showcasing notable accuracies and area under the curve (AUC) scores. Furthermore, through the interpretation of feature importance, this study unveils crucial risk factors contributing to the susceptibility and severity of mucormycosis, providing valuable insights for healthcare professionals in early identification and proactive management strategies. This research contributes to the ongoing efforts in understanding and combating the rise of mucormycosis cases post-COVID-19, emphasizing the potential of machine learning models in facilitating timely interventions and improving patient outcomes.