17 hours ago
Why Generative AI Isn't a Silver Bullet for Healthcare
Home » Smart Sectors » Healthcare » Why Generative AI Isn't a Silver Bullet for Healthcare
Generative AI is revolutionizing healthcare, but challenges remain. Ashley Woodbridge, Lenovo CTO META, explores why it's no silver bullet.
The healthcare sector is on the brink of a significant surge in AI investment, projected to rise by an astounding 169% over the next year – the highest increase among all industries surveyed in EMEA. The uptick follows the promising early results of AI projects, where a remarkable 74% of healthcare AI pilot projects have met or exceeded expectations, with 13% surpassing them. Yet, despite this positive momentum, only 2% of healthcare organisations have actively deployed AI at scale with the majority of initiatives remaining in the pilot or planning stages. This raises the question: how can the industry and its partners move beyond this plateau?
What is Generative AI Doing?
For most people, their experience of Generative AI (GenAI), has been through conversations with chatbots, creating new content or summarising data. These lighter use cases have highlighted numerous challenges that must be addressed before widespread adoption can occur, especially in highly regulated industries like healthcare. One of the primary obstacles identified in the region wide survey is data quality. AI models, including GenAI, are only as effective as the data they are trained on. In healthcare, data often exists in silos, across incompatible systems, and in formats that are difficult to interpret. There are also important patient data privacy considerations to take into account. Despite the complexity involved in sorting and standardising this data, the potential benefits of tackling these issues and making it useable by AI are immense.
For example, tools like Epic's 'Slicerdicer' allow healthcare professionals to query large datasets through conversational interfaces, uncovering important trends in patient outcomes and informing better care. The depth of queries is particularly powerful, allowing healthcare providers to uncover trends amongst patients that share a condition or illness that may have otherwise remained hidden. Taking the technology in a different direction, AI-powered 'ambient digital scribes' are being trialled by NHS doctors in the UK. These systems listen to patient appointments and automatically generate clinical notes, saving hours of administrative work and helping to reduce burnout among medical staff.
However, healthcare providers must tread carefully. Public sentiment towards AI in healthcare remains cautious. Only 28% of people aged over 60 feel comfortable with AI technologies being used in their care, and 75% of consumers overall want to be informed if AI is being used in their healthcare communications. Transparency is critical. For AI to enhance patient experience without undermining it, trust must be maintained.
In countries like the UAE and Saudi Arabia, AI algorithms are enhancing radiology by assisting in the analysis of medical images, enabling quicker and more accurate detection of conditions such as lung cancer. Predictive analytics are being utilized for real-time patient monitoring, allowing healthcare providers to intervene early in critical situations. Generative AI is also making strides in drug discovery, particularly in Qatar, where researchers are modelling molecular interactions to accelerate the development of new therapies. Additionally, AI-powered chatbots and virtual health assistants are streamlining telemedicine services, providing preliminary diagnoses and scheduling appointments, thus improving access to care.
AI's Role in Medical Research
Beyond frontline care, GenAI is turbocharging work in the field of medical research. Earlier this year, a researcher at Imperial College London used an AI tool developed by Google to investigate why certain bacteria are resistant to antibiotics. In just 48 hours, the tool proposed four viable hypotheses, whereas it had taken scientists over a decade to finalise just a single hypothesis manually. The result was so astonishing that the original researcher initially suspected the AI had accessed unpublished work on his personal computer, which was proven not to have been the case.
These breakthroughs are being made possible thanks to the high-performance computing systems behind increasingly powerful AI models. At Lenovo, we are proud to partner with the Broad Institute on genome analysis, helping researchers accelerate one of the most data-intensive tasks in science. The Lenovo Genomics Optimization and Scalability Tool (GOAST) reduces the time needed to analyse a whole human genome from over 100 hours to just 47 minutes.
Other organizations are also harnessing advanced AI and computing tools to push the boundaries of healthcare. Hungarian company 3DHISTECH, for example, uses Lenovo's AMD Threadripper-powered ThinkStation P620 workstations to build detailed 3D virtual models of human and non-human tissue. These models can zoom in to the level of individual chromosomes, enabling new frontiers in digital pathology. Their systems are used by institutions around the world, including Harvard Medical School, Novartis, and the Wuhan Institute of Virology. Notably, a 3DHISTECH system played a pivotal role in diagnosing the first COVID-19 patient in China, demonstrating how AI-enabled technology can impact global health crises.
The Human Element Remains Crucial
While AI offers compelling advantages, it's important to remember that technology alone isn't the answer. The human touch in healthcare remains indispensable. Healthcare providers must ensure transparency in their AI implementations and address data quality issues to fully reap the benefits of AI. By doing so, they can enhance patient care, reduce burnout among medical staff, and drive groundbreaking research, all without sacrificing the trust and comfort of those they serve.
In conclusion, the healthcare industry stands at the brink of an AI-driven revolution. With thoughtful implementation and a focus on maintaining trust, AI has the potential to transform healthcare for the better. The journey won't be without its challenges, but the rewards promise to be well worth the effort.
By Ashley Woodbridge, CTO, Lenovo, META
