Orlando Agrippa, CEO and Founder of RwHealth, turns the spotlight on AI and cancer, where big data brings big gains for medicine
The words artificial intelligence (AI) and ‘machine learning’ (ML) are likely familiar to anyone in the healthcare sector by now; the phrases are used so regularly that they may perhaps begin to lose their impact, especially when they are not connected to something familiar. Cancer – and cancer screenings – are familiar to us all in a professional or a personal capacity, so the use of AI in this sphere is something we will not only comprehend but likely also benefit from in our lifetimes.
AI and ML are breaking major ground in biomedical research and healthcare. Specifically, in the areas of cancer research and oncology, recent developments could yield a vast amount of potential applications. The possibilities within imaging data and genome-scale experimental studies – only a fraction of what AI is capable of supporting in oncology – are endless.
Owing to the convergence of big data and cancer research, we are now fortunate to see a quantum leap forward in the detection and diagnosis of cancer, as well as more profound methods of classifying new therapeutic targets.
Breast cancer
AI has, over the years, shifted from a concept relegated to the pages of science fiction to a powerful technology that could extend the lives of tens of thousands of patients per year. A team from King’s College London and Guy’s & St. Thomas’ hospitals have recently created an AI that can detect a form of cancer that was undetectable by the previous tests. This one development alone has the ability to detect an estimated 30,000 cases per year, enabling better treatment and improving overall outcomes for patients.
For breast cancer, these technologies have been shown to be as effective as human radiologists in spotting tumours from x-ray images. In another study, this time conducted by the University of Warwick in 2021, the combined efforts of AI tools and human ingenuity correctly identified and outlined the clinical pathways of 79,910 subjects out of a total of 131, 822.
Given the success of this study and many more like it, we could likely see AI integrated into future iterations of national cancer screening programmes. The goal of this technology is not to replace radiologists, but rather to support them in delivering the best possible care and quality of patient outcomes.
Precision therapy
AI’s capacity for precise pattern recognition can be used to unearth clinical information that will improve accuracy in diagnostics and therapy. As such, ML will prove invaluable, with frequent applications for precision therapy.
“AI has, over the years, shifted from a concept relegated to the pages of science fiction to a powerful technology that could extend the lives of tens of thousands of patients per year. A team from King’s College London and Guy’s & St. Thomas’ hospitals have recently created an AI that can detect a form of cancer that was undetectable by the previous tests.”
Working with complex neural networks, diagnostic images and genetic data will help us to predict the probability of treatment outcomes and disease occurrence. In addition, deep learning – particularly in terms of its use in extracting images of microscopic malignant tumours – will change diagnostic imaging analysis forever.
Clinical evidence generated from real-world data contains vast potential for advancing oncology therapies. Thanks to the rise in this technology, we now have the opportunity to explore how we can create unique, bespoke treatments for each patient. By curating treatments targeting each patients’ genetic makeup, AI-driven precision therapies can help create treatment plans that are truly tailored to the individual needs of each patient.
This approach will lead to patients receiving the right medicines for them, rather than exhausting multiple options before finding the right one for them. By taking out the tribulations of trial and error, this methodology not only allows for better outcomes to be achieved but also for the NHS to cut down on substantial costs. Most importantly, these rapid developments will improve patient outcomes and enable the delivery of streamlined and tailored care, resulting in a better patient experience during what is an undoubtedly difficult clinical experience.
Looking forward
While huge progress has been made, challenges certainly remain. Ensuring that AI/ML is trained on diverse datasets will be integral to ensuring their equal accuracy across the population, thus enabling high- quality AI-augmented care for all. These advances must reach the entire population to decrease health inequalities across the UK and ensure this rising tide of innovation raises all ships.
