Post-COVID Impact Analysis and Effective Recommendation Solutions Over Risk Prediction Using Hybrid Model

Abstract

The COVID-19 pandemic (2019-2022) resulted in significant global mortality, largely attributed to the virus's unpredictable pathophysiology, rapid disease progression affecting multiple organ systems, and initial lack of effective treatments. This study systematically examines post-COVID-19 complications across major organ systems, including respiratory dysfunction, cardiovascular complications, renal disorders, musculoskeletal pain, gastrointestinal disturbances, neurological sequelae, alopecia, endocrine and metabolic dysregulation, and mental health disorders. The percentage of affected organ systems is demonstrated through clinical scenarios, and evidence-based recommendation systems are proposed to facilitate patient recovery. Disease monitoring is categorized into two approaches: standard hospital-based treatment and individualized home-based care. Unpredicted risk stratification (High or Low) is computed based on significant clinical factors indicating potential organ damage. A hybrid machine learning model combining Long Short-Term Memory (LSTM) and Convolutional Neural Networks (CNN) is employed to assess post-COVID-19 risk with enhanced accuracy. The proposed recommendation systems include AI-based monitoring using wearable sensors, digital health and telemedicine platforms, smart wearable devices, personalized nutrition and dietary management, AI-driven mental health support systems, intelligent rehabilitation and physical therapy programs, and blockchain-enabled AI health records. These integrated systems aim to improve rehabilitation outcomes, enhance patient care quality, and accelerate health recovery by leveraging similar historical patient case data through the hybrid machine learning framework.