Generative AI-Driven Frameworks for Streamlining Patient Education and Treatment Logistics in Complex Healthcare Ecosystems

Abstract

Healthcare ecosystems are increasingly compounded with a plethora of patients presenting with an array of diseases and relying on interdisciplinary treatment approaches. Patient scenarios frequently encompass complex comorbidities, and the provision of care must comply with intricate interplays between diagnostic test results and clinical decisions. Patients’ understanding of clinical procedures and treatment logistics is key for successful therapeutic procurement and favorable therapeutic outcomes. As a result, specialized patient education resources have been developed, providing personalized materials to guide patients through diagnostic and therapeutic procedures. Such resources usually include: (i) short training sessions with healthcare professionals; (ii) reading materials describing the medical background; and/or (iii) short educational videos illustrating the treatment process. Nonetheless, patient education materials are commonly oblivious to the underlying medical terminology and complex treatment algorithms, precluding patients from making well-informed choices regarding their therapeutic options. These challenges are propelled by the intricate prescription decisions, which necessitate a comprehensive knowledge of the disease specifics and the therapeutic property of drugs. To this end, MissingLink enables a personalized patient journey within the healthcare facility, aimed at educating patients in heterogeneous healthcare ecosystems with a program consisting of interdisciplinary treatment decisions. MissingLink can help streamline patient therapeutics by dispatching personalized educational materials tailored to the patient’s disease and treatment armamentarium. Inline missing terms, acronyms, drugs, proteins, or complex pathologies are automatically deciphered using explainable artificial intelligence (XAI) components, fostering better patient comprehension. In view of facilitating the comprehensive deciphering of complex educational videos, an Explainable Video Summarization framework is proposed for the concise generation of clinical tests and the underlying decision-making process carried out by clinicians.