SYNOPSIS
Digital health technology (e.g. smartphone apps and wearable devices) brings challenges and opportunities for Artificial Intelligence (AI).
In England, 189 out of 208 (91%) hospital trusts are using electronic health records (EHRs).
However, most of healthcare data (estimated 80%) is unstructured, which can be made more meaningful by classifying the data and normalising the data with internationally recognised standards, such as SNOMED CT as clinical coding standard.
There are three predominant open data standards for interoperability of EHRs:
OMOP (Observational Medical Outcomes Partnership), openEHR, and HL7 FHIR (Fast Healthcare Interoperability Resources). openEHR is particularly suited to collect clinical care and administration data, FHIR to transfer data (i.e. data exchange between systems and organisations), and OMOP to query patients' records to discover patterns and find insights in the data.
Current AI "cannot reliably deal with facts, perform complex reasoning, or explain its conclusions".
Despite current limitations, AI can foster digital frontiers and the World Health Organization (WHO) "envisions a future where AI serves as a powerful force for innovation, equity, and ethical integrity in healthcare".
WHO recognises that "AI is already playing a role in diagnosis and clinical care, drug development, disease surveillance, outbreak response, and health systems management". AI is not limited to generative AI, such as Large Language Models (LLMs) from neural AI with support for multimodal inputs.
Indeed, symbolic AI and representing diseases as actionable (machine-interpretable) disease knowledge has a long-standing tradition in biomedical research.
Come along to this talk if you are interested in
* A comparison between open-source biomedical LLMs and general-domain LLMs (e.g. DeepSeek, Gemini, Claude, and ChatGPT-4)
* How to lower the technical skills overhead (understanding of AI and programming code) needed to use open-source LLMs for content generation and content analysis of text, images and audio
* Exploring the plausible benefits of neuro-symbolic AI, combining neural AI (to process and extract patterns for health issues from unstructured data) with symbolic AI (explicit representations of background knowledge)
