Exploiting Disorganized Health Records Data: Utilizing Unstructured Health Data for Execution and Action
In the rapidly evolving world of healthcare, advanced technologies such as Artificial Intelligence (AI) and Natural Language Processing (NLP) are revolutionising the way organisations engage with unstructured data. Dr. Tim O'Connell, a practising radiologist and the founder and CEO of emtelligent, a developer of clinical-grade AI software, is at the forefront of this transformation.
Approximately 80% of clinical data remains invisible to conventional analytics due to its unstructured nature. Traditional keyword search tools, commonly used in healthcare, lack contextual understanding and often miss key insights or provide irrelevant results. However, AI and NLP systems can understand the context, semantics, and structure of language, recognising synonyms, interpreting negation, differentiating between historical and current conditions, and extracting relationships between clinical concepts.
The healthcare industry has undergone a significant data transformation, with a focus shifting from electronic health record (EHR) adoption to value-based care, emphasising interoperability, data sharing, and patient access. By 2020, U.S. healthcare data had reached 2,314 exabytes, a 15-fold increase from 2013, due to connected devices and remote monitoring.
Implementing AI in healthcare requires a structured approach to maximise clinical decision-making, predictive analytics, population health management, and other use cases while ensuring patient safety, compliance, and operational efficiency. Here are six best practices aligned with current expert guidance and frameworks:
1. Establish Strategic Alignment and Governance: Align AI goals with patient care and organisational objectives, form cross-functional governance committees, and centralise leadership and accountability.
2. Build Robust Technical and Data Foundations: Assess and upgrade data infrastructure, prioritise compliance with regulations, conduct Privacy Impact Assessments (PIAs), and ensure secure storage to support AI algorithms.
3. Select and Validate Tools Carefully: Choose AI solutions that address specific clinical or operational challenges, evaluate vendors rigorously, and ensure seamless integration with existing systems.
4. Focus on Change Management and User Engagement: Design user-centered workflows, provide targeted training, and promote transparency and communication.
5. Implement Continuous Monitoring and Improvement: Monitor AI performance and outcomes, perform regular audits for bias and fairness, and adapt to evolving needs.
6. Manage Risks and Ensure Ethical Use: Establish clear policies for risk management, maintain human oversight, and balance innovation with patient safety.
By following these best practices, healthcare organisations can harness AI’s full potential while mitigating risks and ensuring sustainable, patient-centered transformation. A 2024 review found that 81% of healthcare systems were using NLP to extract clinical data from EHRs, demonstrating the growing adoption of these technologies in the industry.
References: [1] Healthcare AI Governance Framework, 2021 [2] AI in Healthcare: A Guide for Leaders, 2020 [3] Ethical Guidelines for AI in Healthcare, 2019 [4] AI in Practice: A Guide for Healthcare Professionals, 2022
Dr. Tim O'Connell, an innovator in the field, leverages artificial intelligence and natural language processing to enable his company, emtelligent, to extract meaningful insights from unstructured clinical data. This advancement significantly impacts health and wellness by addressing medical-conditions, where conventional analytics typically fall short. Moreover, the utilization of AI technology in the healthcare industry not only improves clinical decision-making and predictive analytics but also contributes to population health management. The expansion of AI adoption in healthcare, highlighted by the 2024 review that found 81% of healthcare systems utilizing NLP for data extraction from EHRs, signifies the growing influence of science and technology in the sector.
