A Q&A with Herat Joshi, Data Science – Team Lead, Great River Health Systems, on the transformative potential of AI in healthcare.
The launch of ChatGPT marked a pivotal moment in the evolution of Generative AI, bringing it to the forefront of technological innovation. This shift has reshaped how we interact with AI in everyday life, spanning diverse applications from content generation to enhanced decision-making across industries.
Healthcare, in particular, has been a focal point for AI’s transformative potential, with groundbreaking innovations in diagnostics, personalized medicine, and operational efficiency. As healthcare systems grapple with increasing complexities, AI has emerged as a powerful tool to manage large volumes of data, improve decision-making, and deliver personalized healthcare.
We spoke with Herat Joshi, Data Science – Team Lead, Great River Health Systems, who has been instrumental in driving healthcare technology innovation across Southeast Iowa. His expertise, particularly in implementing interoperable EHR systems, showcases AI’s practical applications in healthcare.
Joshi has played a key role in advancing healthcare technology for globally influential companies like Apollo Hospitals and Siemens Healthineers. At Apollo Hospitals International Ltd, where he served as Sr. Technology Manager, Herat was instrumental in the launch of Apollo Prism, a Personal Health Record (PHR) system used by millions of patients. Apollo Hospitals, ranked among the top 5 in India and 126th globally, is a vast network with 73 hospitals, 200 clinics, 2,300 diagnostic centers, and over 10,000 beds.
At Siemens Healthineers, Herat contributed to the development of the Symbia Intevo, the world’s first xSPECT system, which revolutionized diagnostic imaging by integrating SPECT and CT technologies, significantly improving the precision and clarity needed for detecting complex conditions such as cancer and bone disorders.
In this Interview with Herat Joshi, we explore the future of AI in healthcare, its current applications, challenges and the innovations that will shape the future of the industry.
Q: There’s a lot of buzz about AI in healthcare, but is there a particular topic that you think is being overlooked amidst all the excitement?
I’m really excited about foundation models, which I think are being overlooked in the healthcare space. Models like ChatGPT are trained on vast amounts of data, but we haven’t yet seen foundation models built on high-quality medical data. Imagine training a model on all the imaging data from a health system, combined with EHRs, lab results and genomics. This could lead to AI models that perform exceptionally well on specific medical tasks and improve patient outcomes. This approach hasn’t been fully realized yet – but I believe it could be transformative for AI in healthcare.
Q: How will AI be integrated into healthcare systems in the future, and how will we see it in clinical practice?
While AI is making significant strides in radiology, many current models are focused on narrow clinical problems, such as detecting nodules or brain hemorrhages. These tasks are crucial, but excessive false positives can overwhelm radiologists, leading to more time spent reviewing cases. Going forward, the focus will likely expand to integrating AI into broader clinical workflows – like automating tasks such as clinical notetaking, patient registration and summarizing large volumes of medical data. This will help streamline administrative work, improve care delivery and free up healthcare professionals to focus more on patient care. Broader integration across workflows will be a game-changer in clinical practice.
Q: What challenges do you feel AI has helped tackle and where do you see technology still lacking?
AI has already made significant strides in areas like diagnostics, workflow optimization and predictive analytics. It has helped tackle some key challenges, such as analyzing large datasets to identify patterns that would be difficult for humans to detect and streamlining administrative processes like scheduling or billing. However, one of the biggest challenges still lies in obtaining the right datasets—those that are high-quality, well-curated, and diverse enough to train and evaluate algorithms properly. Since machine learning models depend heavily on the data they’re trained on, the quality of that data directly impacts their performance. Without standardized, comprehensive datasets, AI technologies can struggle with issues like bias or inaccuracies, limiting their effectiveness. So, while AI has helped with many challenges, the need for better data remains a key obstacle to broader adoption and improved outcomes.
Q: What innovations in AI technology are you most excited about?
The integration of AI with the Internet of Things (IoT) is a game-changer for healthcare. Wearable devices equipped with AI sensors can continuously monitor patients’ vitals and send real-time data to healthcare providers. This level of continuous monitoring enables early detection of abnormalities, reducing the need for hospitalizations and improving overall patient outcomes. For example, AI-powered wearables are being used to monitor cardiovascular health, alerting patients and doctors of potential risks before they escalate into serious conditions.
Another exciting innovation is the use of GenAI models in medical imaging. GenAI can enhance images, providing greater clarity and allowing for more accurate diagnoses. This technology is particularly beneficial in fields like radiology, where detailed imagery is crucial for detecting diseases such as cancer.
Q: What steps should healthcare organizations take to stay ahead in the AI revolution?
To stay ahead, healthcare organizations should focus on interoperability, ensuring that AI systems integrate smoothly with existing technologies. Data security and robust governance are also crucial to protect patient privacy and meet regulatory standards. Additionally, forming partnerships with tech companies and academic institutions can accelerate AI innovation through collaboration. Finally, investing in continuous education and training is essential to prepare healthcare professionals to work alongside AI, ensuring they’re equipped with the skills to maximize its potential in healthcare.
Q: What AI advancements do you think will most impact patient outcomes in the next 5-10 years, and are there any emerging technologies that could further transform healthcare?
In the near future, AI will significantly enhance personalized medicine and diagnostics, leading to earlier disease detection. Technologies like quantum computing could further transform healthcare by accelerating data processing for drug discovery, while federated learning will expand AI’s reach by enabling models to learn from decentralized data across multiple health systems, all while maintaining patient privacy.
