JMIR AI

The digital transformation of health care has introduced both opportunities and challenges, particularly in managing and analyzing the vast amounts of unstructured medical data generated daily. There is a need to explore the feasibility of generative solutions in extracting data from medical reports, categorized by specific criteria.

MRI reports are challenging for patients to interpret and may subject patients to unnecessary anxiety. The advent of advanced artificial intelligence (AI) large language models (LLMs), such as GPT-4o, hold promise in translating complex medical information into layman terms.

The mechanisms underlying the mutual relationships between chronic physical illnesses and mental health disorders, which potentially explain their association, remain unclear. Furthermore, how patterns of this comorbidity evolve over time are under-investigated significantly.

Artificial intelligence (AI) is transforming medical imaging, with large language models such as ChatGPT-4 emerging as potential tools for automated image interpretation. While AI-driven radiomics has shown promise in diagnostic imaging, the efficacy of ChatGPT-4 in liver ultrasound analysis remains largely unexamined.

Generative AI (gAI), such as DALL-E 2, are promising tools that can generate novel images or artwork based on text input. However, caution is warranted as these tools generate information based on historical data and are thus at risk of propagating past learned inequities. Women in medicine have routinely been under-represented in academic and clinical medicine and the stereotype of a male physician persists.

There is considerable need to improve and increase the detection and measurement of depression. The use of speech as a digital biomarker of depression represents a considerable opportunity for transforming and accelerating depression identification and treatment; however, research to date has primarily consisted of small-sample feasibility or pilot studies incorporating highly controlled applications and settings. There has been limited examination of the technology in real-world use contexts.

The recent availability of law enforcement and coroner/medical examiner reports for nearly every violent death in the US expands the potential for natural language processing (NLP) research into violence.

Pharmacoepidemiologic studies, which promote rational drug use and improve health outcomes, often require Anatomical Therapeutic Chemical Classification System (ATC) drug classification within real-world data (RWD) sources. Existing classification tools are expensive, brittle, or have restrictive terms of service, and lack context that may inform classification itself.

Alzheimer disease and related dementias (ADRD) are complex disorders with overlapping symptoms and pathologies. Comprehensive records of symptoms in electronic health records (EHRs) are critical for not only reaching an accurate diagnosis but also supporting ongoing research studies and clinical trials. However, these symptoms are frequently obscured within unstructured clinical notes in EHRs, making manual extraction both time-consuming and labor-intensive.

Language, which is a crucial element of human communication, is influenced by the complex interplay between thoughts, emotions, and experiences. Psychiatric disorders have an impact on cognitive and emotional processes, which in turn affect the content and way individuals with these disorders communicate using language. The recent rapid advancements in large language models (LLMs) suggests that leveraging them for quantitative analysis of language usage has the potential to become a useful method for providing objective measures in diagnosing and monitoring psychiatric conditions by analyzing language patterns.

Generative artificial intelligence (AI) is showing great promise as a tool to optimize decision-making across various fields, including medicine. In anesthesiology, accurately calculating maximum safe doses of local anesthetics (LAs) is crucial to prevent complications such as local anesthetic systemic toxicity (LAST). Current methods for determining LA dosage are largely based on empirical guidelines and clinician experience, which can result in significant variability and dosing errors. AI models may offer a solution, by processing multiple parameters simultaneously to suggest adequate LA doses.

Properly configuring modern electronic health records (EHRs) has become increasingly challenging for human operators, failing to fully meet the efficiency and cost-saving potential seen with the digitization of other sectors. The integration of artificial intelligence (AI) offers a promising solution, particularly through a comprehensive governance approach that moves beyond front-end enhancements like user- and patient-facing co-pilots. These co-pilots, although useful, are limited by the underlying EHR configuration, leading to inefficiencies and high maintenance costs. To address this, we propose the concept of an "Elastic EHR," which proactively suggests and validates optimal content and configuration changes, significantly reducing governance costs and enhancing user experience, reducing many of the common frustrations including documentation burden, alert fatigue, system responsiveness, outdated content, and unintuitive design. Our five-tiered model details a structured approach to AI integration within EHRs. Tier I focuses on autonomous database reconfiguration, akin to Oracle Autonomous Database functionalities, to ensure continuous system improvements without direct edits to the production environment. Tier II empowers EHR clients to shape system performance according to predefined strategies and standards, ensuring coordinated and efficient EHR solution builds. Tier III optimizes EHR choice architecture by analyzing user behaviors and suggesting content and configuration changes that minimize clicks and keystrokes, thereby enhancing workflow efficiency. Tier IV maintains the currency of EHR clinical content and decision support by linking content and configuration to updated guidelines and literature, ensuring the EHR remains evidence-based and compliant with evolving standards. Finally, Tier V incorporates context-dependent AI co-pilots to enhance care efficiency, quality, and user experience. Despite the potential benefits, major limitations exist. The market dominance of a few major EHR vendors—Epic Systems, Oracle Health, and MEDITECH—poses a challenge as any enhancements require their cooperation and financial motivation. Furthermore, the diverse and complex nature of healthcare environments demands a flexible yet robust AI system that can adapt to various institutional needs that has not yet been developed, researched, or tested. The Elastic EHR model proposes a five-tiered framework for optimizing EHR systems and user experience with AI. By overcoming the identified limitations through vendor-led, collaborative efforts, AI-enabled EHRs could improve the efficiency, quality, and user experience of healthcare delivery, fully delivering on the promises of digitization within healthcare.