The application of artificial intelligence (AI) in the medical field has advanced significantly in recent years, particularly in disease prediction and medical visualization. This study conducted a scoping review of AI-based approaches applied in healthcare, focusing on proposals incorporating multiple technologies within their framework. The findings were categorized into three main approaches: 1) Multimodal data input models; 2) Synthetic medical image generation through generative artificial intelligence, and 3) Multi-disease prediction using machine learning and deep learning. However, in the recent literature (last five years), it has not been found initiatives that integrate both predictive models and generative AI into a complete workflow capable of processing medical records and imaging data to deliver a multi-disease prediction alongside a visual representation of the expected deterioration, particularly in non-communicable diseases with high comorbidity. Based on these findings, a novel architecture is proposed for predicting systemic deterioration through multi-disease forecasting, complemented by the generation of synthetic images that illustrate the expected progression of the patient's condition. This approach aims to raise patient awareness about their health status and enhance doctor-patient communication by providing an intuitive and visually interpretable representation of disease progression.