The role of NHS Big Data and AI technology in unlocking the future of lung and heart disease treatment

The UK stands at the cusp of a new era of medical breakthroughs, fuelled by the increased use of data and the recent plans to open NHS data reserves to improve the development and creation of artificial intelligence (AI) technologies in healthcare.

The Government's recently launched AI Action Plan, focusing on creating a National Data Library, marks a pivotal moment for the use of AI in healthcare. It has the potential to unlock unprecedented opportunities for medical research, drug development, and ultimately new treatment options that will benefit patients.

Removing barriers to AI innovation

For years, the NHS has had a wealth of data – including medical histories, treatment outcomes, and biodata, which has remained largely inaccessible – locked away in silos limiting its potential to drive change and innovation. AI systems rely on extensive, diverse, and high-quality datasets to train algorithms effectively, but restricted access to such data has been a significant obstacle to advancing AI in healthcare.

However, the potential is clear: the NHS sees around 1.3 million people a day, equivalent to the population of Estonia, and the sheer volume of information is staggering. Creating the National Data Library will provide approved organisations with secure and centralised access to anonymised patient records, scans, and other crucial data points.

Opening up the UK's massive repository of longitudinal NHS health data to big tech companies will enable them to train AI models on anonymised patient records for the first time. By leveraging the combined power of big data and AI technology, the NHS aims to enhance its patient care offering and drive a new era of medical research and drug development. Although the innovation potential is unquestionable, we must ensure that data quality and security aren't compromised; complete transparency about how companies use NHS data is crucial.

Effective AI data curation in lung disease

The landscape of complex lung disease diagnosis and treatment has undergone a significant transformation in recent years, driven by advancements in AI and machine learning technologies.

Traditionally, interpreting CT scans in patients with complex lung diseases required the expertise of highly trained radiologists. However, this manual process is time-consuming, prone to human error, and can lead to delays in diagnosis and treatment. It is notable that almost half of patients with idiopathic pulmonary fibrosis, an aggressive condition with no known cure, are misdiagnosed at least once, with the mean time to diagnosis of a year.

Specially created AI platforms designed by clinicians and data experts are now changing this process and allowing treatments and patient outcomes to be more closely and rapidly linked, with deep-learning tech deployed in large lung disease clinical trials. What does this achieve?

1. New insights on clinical outcomes: AI systems swiftly analyse CT scans and clinical data using sophisticated machine learning algorithms, identifying complex patterns within the data that may not be easily identified using conventional data mining techniques.

2. Identifying risks and enabling prioritisation: AI facilitates faster identification of at-risk patients and efficiently prioritises urgent cases. AI automatically overlays and compares multiple CT scans, adjusting for movement or breathing, which helps detect subtle changes in a patient's condition.

3. Better understanding of disease progression: Specialised algorithms can now measure lung volume, airway volume, vascular volume, and fibrosis extent, providing a more nuanced understanding of disease progression than traditional methods and facilitating precision medicine.

4. Continuous learning and early disease detection: By leveraging quality data, AI models can continuously learn and evolve, improving their ability to detect even trivial abnormalities that may represent subclinical, but progressive, early disease.

AI holds the key to improving treatment in IPF

Pulmonary fibrosis, which is responsible for 1% of deaths in the UK and kills nearly as many patients as breast cancer in the US, represents a seriously under-recognised public health problem. The pharmaceutical industry still faces significant hurdles in developing effective treatments for pulmonary fibrosis, primarily due to the lack of precise clinical outcome measures and reliable biomarkers. Traditional pulmonary function tests, which are commonly used to assess disease severity, often lack accuracy. This necessitates larger and more expensive clinical trials to establish drug efficacy definitively. However, AI radically changes how complex diseases like IPF are treated and assessed.

AI can deploy machine learning algorithms to analyse complex CT scan data, enabling precise tracking of lung scarring progression and identification of quantifiable tissue changes. AI's advanced capabilities provide more accurate methods for evaluating treatment efficacy and predicting patient outcomes, offering clinicians powerful tools for personalised treatment strategies.

By leveraging the extensive NHS data repository, AI systems can detect subtle disease patterns and refine treatment approaches, potentially improving care for patients with pulmonary conditions. This is particularly important in conditions such as IPF, where data is sparse and central repositories are needed to power machine learning algorithm development. With access to enough data, the development of novel AI algorithms that facilitate precision medicine in patients with IPF is likely to accelerate. These approaches will likely significantly impact conditions of the chest including lung cancer, cardiovascular disease and pulmonary hypertension.

A future shaped by data and AI

The UK's AI Action Plan represents a bold step towards a clinical future where big data and AI drive healthcare. By unlocking the vast reserves of NHS data and investing in cutting-edge AI technologies, the UK can improve patient outcomes, accelerate drug development, and position the UK as a global leader in medical innovation. The convergence of NHS data and deep learning AI represents a seismic shift in healthcare. By embracing this opportunity, the UK can unlock the insights hidden within its vast data reserves and transform how we approach complex diseases.

About the author

Dr Simon Walsh is a distinguished expert in thoracic radiology and imaging-based biomarker development, with a career spanning 15 years. He currently holds multiple positions, including chief scientific officer at Qureight, as an NHS consultant thoracic radiologist at Royal Brompton and Harefield Foundation Trust, and as a clinician scientist at Imperial College London. Dr Walsh specialises in fibrotic lung disease and has authored over 250 publications. His research focuses on applying deep learning and AI to solve fibrotic lung disease imaging challenges, aiming to improve diagnosis, early detection, and disease progression prediction.