Dr Andrew Jones
Head of Digital Transformation, Amazon Web Services (AWS)
Dr Matt Howard
Head of Health Data Sciences, Amazon Web Services (AWS)
Dr Francisco Azuaje
Director of Bioinformatics, Genomics England
Generative artificial intelligence (genAI) is empowering healthcare organisations to better leverage their data to improve the patient experience, address workforce challenges and boost productivity.
By handling routine backroom tasks, genAI frees clinicians from administrative duties, allowing them more time with patients by automatically generating common documents like discharge notes and GP letters. Additionally, genAI mines data from documents, including scientific literature, to aid biomedical research and improve patient outcomes with new therapies.
GenAI for healthcare admin automation
Former GP and Head of Digital Transformation at Amazon Web Services (AWS), Dr Andrew Jones, says that when it comes to improving healthcare workforce productivity, a major challenge is the amount of time clinical staff spend on administration.
A physician with 20 years’ experience in the NHS in both primary and secondary care, he points to scenarios where following a patient consultation, GPs would type into the electronic health record, order tests and then make a referral repeating the same information each time.
“One of the things that genAI is highly efficient at, is transcribing patient-clinician conversations,” says Jones.
Data enhancing patient records
The cloud provider has systems to ease this administrative burden on clinicians. “We have also developed services that take out key entities and important information from these conversations — such as medication, diagnoses and tests — summarise that and put it into the patient record,” he adds. “That is going to save doctors a huge amount of time.”
Aiding in medical document processing
As more patients seek access to their records, AWS is also being asked to design systems that provide summaries of medical documents that are accessible and understandable. Dr Matthew Howard, Head of Health Data Sciences, outlines how a range of customers — from hospitals and clinics to strategic research institutions — are assisted in getting started with these technologies.
As well as supporting with automation of high-volume tasks and extracting information, compliance and ensuring privacy and data security is a critical element, insists Howard. He leads a team of applied data scientists who work with customers, harnessing critical data to develop generative AI-based healthcare solutions.
One of the things that genAI is highly efficient at, is transcribing patient-clinician conversations.
Projects include multimodal cancer patient stratification, application of genAI to automate literature review and development of genAI-based medical information systems. “We provide customers access to a range of genAI models, allowing them to identify the best performing, depending on use,” he adds. We are committed to developing AI responsibly.
By prioritising ethical considerations and taking advantage of the tools, support and healthcare expertise, organisations can harness the full potential of generative AI while ensuring data protection and fairness in decision-making processes.
Assisting biocurators in knowledge curation
The company is working closely with Genomics England, a global leader in enabling genomic medicine and research, on using genAI to accelerate research. Dr Francisco Azuaje, Director of Bioinformatics at Genomics England, says they partner with the company to use genAI tools to help biomedical knowledge management experts (biocurators) process large amounts of information from scientific literature.
He explains: “This knowledge of medical conditions from the scientific literature is needed to help clinical experts establish connections between gene variants and diseases. The knowledge curation process aims to improve and maintain the knowledge base for supporting clinical diagnosis.”
More accurate diagnosis for improved patient outcomes
The AWS services and genAI research application, he continues, enables biomedical experts to efficiently extract and prioritise clinically relevant associations between gene variants and medical conditions from numerous scientific articles, ensuring Genomics England incorporates existing and new findings into its knowledge management pipelines.
Azuaje says: “This application is very important for us because, by integrating these findings into the Genomics England knowledge base, it provides evidence to help inform future decisions on the introduction of new genomic technologies which could make a real difference to the diagnosis and care of patients with rare conditions.”
