Chathuri Daluwatte, Head of AI Diagnostics (Alexion AstraZeneca Rare Disease), AstraZeneca, speaks with Savio Rodrigues, Head of New Business Acquisition at Trianz and CDO Magazine Editorial Board Member, in a video interview about transforming healthcare with AI, AI's diagnosticspotential to improve the diagnostic journey, and health equity for rare disease patients.
AstraZeneca is a global, science-led biopharmaceutical business with its innovative medicines being used by millions of patients worldwide.
Daluwatte has bachelor’s degrees spanning fields of electronics, telecom, and computer science, along with a master’s degree in statistics, and a Ph.D. in biomedical engineering. She has worked at bioMérieux and then in the U.S. Food and Drug Administration (FDA), where she worked in the devices division at the Center for Devices and Radiological Health (CDRH), and later at the Center for Drug Evaluation and Research (CDER).
Her professional experience also includes working at Sanofi, across the value chain and the enterprise Chief Digital Office, leading development, and deployment of AI products at scale globally, while maintaining compliance. A common career thread lies in bringing AI into the real world and driving value, in life sciences.
Coming from the AI perspective, Daluwatte believes that organizations must change their mindsets. Instead of attempting to organize data first, and then think about AI, it should be done the other way around.
She also lays down three crucial reasons behind the need for the changed mindset:
Data organization is a never-ending job. It is constantly evolving, and organizations can never fully say that the data is completely organized for them to start on AI.
Even if the organization has seemingly organized data, it will quickly realize that the data is not organized in a way to support AI.
Most importantly, data organization is a low-hanging fruit, but the real growth value comes from leveraging AI to realize value at a larger scale.
Moving forward, Daluwatte expresses a positive opinion about the healthcare industry being well-equipped to lead AI implementation in a modern and disciplined manner. She highlights the digital transformation in the healthcare sector in the past 15 years and the establishment of global interoperability standards like DICOM for imaging.
Additionally, she stresses the regulation aspect stating that the healthcare industry is ahead of other industries in many ways. Citing an example, she mentions that the U.S. FDA has been actively working to incorporate AI since the passage of the 21st Century Cures Act (Cures Act) in 2016.
In continuation, Daluwatte shares that the tech industry has also started creating sustainable foundations, specifically in the form of interoperability standards. Some examples include Nvidia MONAI, Apple CareKit, and Google’s Open Health Stack, which provide the necessary resources to the healthcare industry.
However, the discussions around accountability are gathering concerns, such as determining liability if an algorithm produces incorrect results. Daluwatte recounts that this aspect makes doctors hesitant towards AI adoption. She says that the risk stratification of algorithms has been there regardless of AI because it has always been tied to use cases and not technology.
She believes that the evolution of technology to provide better user experience will lead to faster adoption of AI and will mitigate healthcare providers’ burnout and staff shortages. Daluwatte asserts that these benefits in life science and healthcare can be studied by other industries to further AI innovations.
Delving further, she opines that there are approximately 7,000 rare diseases that often go unnoticed by healthcare providers. Many a time, the symptoms of these diseases are unspecific, leading to misdiagnosis and wrong therapy.
Sometimes, the therapies are even counter-indicated for the rare disease, says Daluwatte. Unfortunately, this means that patients may not receive the proper diagnosis until their condition has progressed to a severe state, and by then, the treatment may not be effective.
Therefore, Daluwatte recommends integrating expert knowledge around rare diseases with healthcare providers, embedding them into clinicians’ workflows, and leveraging AI to prompt them to consider rare diseases as a possibility. Citing another example, she mentions how cardiac amyloidosis is often misdiagnosed as heart failure.
Daluwatte explains that simply adding AI to the radiology workflow for echocardiograms can aid the early diagnosis of patients, leading to more targeted and effective treatment.
Furthermore, she states that such solutions are FDA-approved and are being validated in clinics. She notes that there are 522 FDA-approved such algorithms, and two-thirds of them are in radiology.
However, Daluwatte confirms that one aspect of AI in healthcare is that the algorithms are built around technologies that focus on MRI and major diseases like heart failure, while rare diseases are often overlooked due to a lack of knowledge.
Another crucial aspect is that while integrating healthcare AI into hospital systems, CTOs and CIOs invest in and implement solutions that help reduce physician burden. Further, from the HealthCare Partners (HCP) side as well, rare diseases fall through the cracks.
In conclusion, Daluwatte states that CDO magazine, with its global editorial board community and many healthcare CDOs, has a unique opportunity to spark conversation and raise awareness around the importance of AI adoption for rare diseases.
CDO Magazine appreciates Chathuri Daluwatte for sharing her invaluable insights with our global community.