The Medical AI Podcast

Episode 41: The path to a medical AI holy grail: medical records analysis and automated treatment selection; with Mika Newton, CEO of XCures

Pat Baird is co-chair of the AI committee for the Association for the Advancement of Medical Instrumentation and the International Medical Device Regulators Forum

Davit Buniatyan is the CEO and Founder of Activeloop, a company advancing the way deep learning teams manage multi-modal datasets. In this episode, Buniatyan discusses how Activeloop's AI-native data format based on tensors is poise to help transform healthcare by enabling the digitization of medical images and streamlining workflows.

Buniatyan shares insights into the primary safety concerns when integrating private company data with large language models, as well as the current challenges around the GPU shortage and how Activeloop is addressing them. He also delves into the rise of unstructured data and the need for industry standards to efficiently process this information for machine learning purposes.

As the leader of a company at the forefront of AI data management, Buniatyan provides valuable perspectives on the future of this rapidly evolving field and how Activeloop is helping teams ship AI products faster without the need to build complex data infrastructure.

In this episode, we explore the pioneering world of digital and computational pathology with Dr. David Klimstra, whose illustrious career spans from his former role as Chair of the Pathology Department at Memorial Sloan Kettering Cancer Center to his current position as Co-Founder and Chief Medical Officer at Paige AI. Dr. Klimstra shares his journey through the evolution of pathology, highlighting the transformative impact of artificial intelligence and digital technologies on the field. We explore how his work at Paige AI is reshaping the landscape of cancer diagnosis and treatment, making it more precise, efficient, and patient-focused.

Episode 37: AI for psychotherapy. With Lindsay McKean, Psychotherapist

The liver is a miracle of physiology, with over 1,000 known functions. This highlights the potential impact of liver dysfunction on various bodily systems. Also, liver disease is a major cause of death globally, and the numbers are on the rise. This adds up to a critical clinical need.

Dr Neil Guha is Professor of Hepatology at the University of Nottingham, an NHS consultant and Research scientist. He recently published an excellent paper in the journal of medical AI on a ML model to rate risk for liver disease.

A year ago, very soon after the release of ChatGPT there was an article in Nature with the curious title… ChatGPT listed as author on research papers: many scientists disapprove

Since then, the use of LLMs in medical research has exploded, including as a coauthor of articles. But how is it being used? Does it matter? Does it undermine the human authors? And does it risk the quality of the research?

With me to discuss this is a brilliant up and coming researcher at the Universal Scientific Education and Research Network, Fiona Morrison.

Anitha Kannan is a founding member and the head of machine learning at Curai Health, an AI-powered virtual clinic on a mission to improve access to care at scale. Anitha spearheads the development of machine learning technology designed to reimagine care delivery and enhance patient outcomes. Prior to her work at Curai, Anitha held senior research positions at Facebook AI Research and Microsoft Research, where she received multiple Gold Star awards. She holds a PhD in machine learning from the University of Toronto and was a Darwin Fellow at the University of Cambridge.

Dr. Dan Heller leads the Cancer Nanomedicine Laboratory at Memorial Sloan Kettering Cancer Center, where he develops advanced diagnostic technologies.

His lab recently made headlines for creating a nanosensor "nose" that can sniff out ovarian cancer sooner than current tests. Using carbon nanotubes and machine learning, this artificial olfaction technology was able to identify ovarian cancer in early stages, making treatment more feasible.

In addition to his pioneering work in the lab, Dr. Heller co-directs the Pat and Ian Cook Doctoral Program in Cancer Engineering at Memorial Sloan Kettering, is a member of the Molecular Pharmacology Program at the Sloan Kettering Institute, and serves as a Professor at the Weill Cornell Graduate School of Medical Sciences.

Episode 32: Medical AI generalism and diabetic retinopathy, with RJ Kedziora, Co-founder of Estenda Solutions

Episode 31: AI for summarizing medical notes, with Vince Hartman, CEO of Abstractive Health

Episode 30: The Real World Deployment of Medical AI, with Louis Culot, Head of Innovation Capabilities for Philips Healthcare

Episode 29: AI for monitoring hospital patients, with Narinder Singh, CEO at LookDeep Health

Can AI systems provide ways of monitoring hospital patients? This innovative idea is being implemented in some American hospitals. With Narinder Singh, CEO at LookDeep Health.

Personalised medicine has been promised as a coming revolution in healthcare. The concept goes all the way back to Hippocrates, around 400 years BC, but it is still very rare in clinical practice. Machine learning has been described as the key technology for personalised medicine - so are we finally about to see this revolution storming the gates? Discussing this with Felix is one of the key figures in this area, Richard Dobson, Professor of Medical Informatics at the Institute of Psychiatry, Psychology & Neuroscience, and Professor of Biomedical and Health Informatics at the Institute of Health Informatics, University College London.

Are there ethical issues with medical AI that are unique? Any major technological development can have huge benefits and huge risks. And medical AI is no different. The benefits of medical AI are talked about endlessly. Much less so the risks. But what are those dangers?

The Declaration of Helsinki is widely regarded as the cornerstone of human ⁠research ethics⁠. This a set of ethical principles regarding ⁠human experimentation⁠ developed originally in 1964 for the medical community by the ⁠World Medical Association⁠. The fundamental principle is respect for the individual, his or her right to self-determination and the right to make ⁠informed decisions⁠ regarding participation in research. But what is the support adn the Declaration of Helsinki? Is it for purpose? Or is it actually highly questionable in some circumstances?

Joining Felix is Professor Erwin Loh, Chief Medical Officer for St Vincent's Health, Australia’s largest not-for-profit health care provider, Professor at Monash, Melbourne and Macquarie Universities and renowned author on medical AI.

Medical AI might seem like a really new field. For example, the first medical AI product approved by the FDA was done so in only 2017. But does medical AI actually have a history that is far longer and richer than we might think? And are there important lessons to be learnt from that history?

Felix discusses with Professor Mark Braunstein, Author of "Health Informatics on FHIR". An author and thought-leader in the field, Dr. Braunstein is a professor of health informatics for the School of Interactive Computing in the College of Computing at the Georgia Institute of Technology where he teaches Introduction to Health Informatics, a popular graduate seminar, both on campus and as part of the OMSCS program. He is involved with HL7 and in various research projects at Georgia Tech, and works with and advises numerous community and industry groups.

Original music by Felix Beacher: https://youtu.be/JS93808r-W4

There are many research groups and companies developing excellent products in medical AI. But some succeed and some fail. What is the difference between the two? Felix discusses this with someone who is a real expert in this - Dr Hugh Harvey Managing Director of Hardian Health. Pure gold for anyone interested in developing a strategy for new medical AI products.

Worldwide, the capacity of health systems is being strangled, by the complex healthcare needs of ageing populations, and the rise in chronic diseases more generally. Radiology departments in particular, are simply unable to meet a massive and rising demand and patients are facing ever longer delays in treatment. Much has been made about the ability of AI systems to interpret medical images as well or better than human radiologists. But are we really moving towards a future in which AI systems have transformed radiology to the point where they are independently processing scans quickly, accurately and en masse? If that transformation is underway, what will be the impact, good and bad? With Fabian Schoeck, Head of Global Product Management for AI Products at Siemens Healthineers.

GE Healthcare is the market leader in medical AI: it has the largest market shares and the largest number of FDA-approved medical AI products. This includes algorithms for MRI image cleaning, detection of collapsed lungs and diagnosis of breast cancer. So where GE Healthcare is going with medical AI really counts. To discuss this I am joined by Simon Rost, Chief Marketing Officer of GE Healthcare.

Medical ultrasound is now old technology – its first use was in 1941 for detecting brain tumors. However, it is also a technology with a great future and AI is a big part of that evolution. Felix is joined by Dr Yoran Hummel president of Us2.ai to discuss current AI-based innovations for ultrasound imaging.

Here we discuss the use of machine learning to predict outcomes of clinical trials. This is potentially a practical and robust way to develop tools for precision medicine and also offers possibilities for radical new approaches to drug development.

Radiology is a classic area of medical AI. AI systems for medical image analysis have several major advantages over human radiologists:

1. They can be trained on millions of scans, far more than a human can be trained on

2. They can give results instantly

3. Many studies indicate that, in many contexts, AI systems are more accurate for disease detection and diagnosis

That’s the promise. But what is the reality? Where is the field at? What are the successes and failures?

I discuss this with Darren Stephens, Senior Vice President of Qure.ai

Artificial General Intelligence is the ability of an AI system to perform any intellectual task performed by humans. The recently released much-hyped Chat GPT has been described as the first glimmering of AGI. It has passed exams for law, biology, medicine even exams for being a wine sommelier. It is a type of AI system known as a large language model. But what does this mean for healthcare? There may be no one in the world who knows more about this than Dr Harvey Castro. Harvey is a New York-based physician, entrepreneur and author of ChatGPT and Healthcare: The Key To New Future of Medicine.

Episode 16: Highlights from the European Congress of Radiology 2023

Every year in the beautiful cultural city of Vienna, there is conference known as ECR, the European Congress of Radiology. This is a chance for the movers and shakers in medical imaging to get together to exchange ideas, make pitches, and launch new products.

In recent years, AI has become more and more prominent at ECR, to the point where it is now quite rare to find an exhibitor not talking about AI.

I wanted to briefly sample some of the most interesting companies, to give an overview of where medical AI is right now. What are they working on? What stage are they at? How is it going?

Here is a series of ten  short interviews, my highlights of ECR 2023. In order, you’ll hear from GE Healthcare, Koios, Siemens Healthineers, Qure.ai, Samsung, Esaote, Agfa, Canon, Carestream and Varex.

Original music by Felix Beacher: https://youtu.be/plg98pnyUew

Today I am joined by Matt McBride, the CEO of Mend, and we discuss the challenges that doctors and patients face when it comes to effective communication, and how AI systems can be used to bridge the gap.

A study by Accenture found that many Americans are unhappy with the communication channels they have with their healthcare providers. This has the potential to contributes to distrust and even ignorance about best care practices by patients and doctors. Matt asks the question: how can we better utilise the technology we have to bridge patient-doctor relationships.

Matt shares insights into the specific ways in which Mend's technology is helping doctors and patients communicate more effectively. Through this journey, we also cover important factors of the ethical considerations around using AI in healthcare, and how Mend is working to ensure that patient privacy and autonomy are always respected. And finally, how might data bias be mitigated for the highest standard of care across a range of subjective issues? 

Matt McBride MBA, is the CEO and Co-Founder of Mend, a health communications company that utilises AI and med tech to streamline patient-doctor information sharing. He has extensive experience in the healthcare technology sector bridging business acceleration and medical optimisation.

Original music by Felix Beacher: https://youtu.be/JS93808r-W4

Janette Phi’s journey into healthcare technology is a personal one, driven by a medical issue her son experienced in his youth. Driven by her own story, she asks how medical normalisation and precision healthcare work together to bring the correct care to correct patient - with the potential to save lives.

Join me, Dr Felix Beacher, and in Episode 14 of our podcast, we welcome Janette Phi, the Chief Business Officer of ThinkCyte. Janette shares her insights on how to grow a company and the challenges she has faced in leading ThinkCyte to success. We dive deep into the world of single-cell treatment, looking at how understanding one cell can help build medical frameworks of the whole body. We ask how doctors weigh the relationship between personal and general treatment, the growing field of precision medicine, and the importance of improving information sharing within the medical community. Join us as we explore new frontiers in using artificial intelligence to fight against cancer and bring well-needed healthcare to those who need it.

Who is Today's Guest?

Janette Phi is Chief Business Officer at ThinkCyte and is responsible for leading global business development and commercial activities. Phi brings over 30 years of technical and business experience with advanced scientific technology companies. She has served as CBO for IntelliCyt (acquired by Sartorius), COO for Xcell Biosciences and CCO for Applied Cells and ONI.

She holds a B.S. in Cell Biology from the University of California, Berkeley and an MBA with an emphasis in Marketing & Finance from Santa Clara University

Original music by Felix Beacher: https://youtu.be/JS93808r-W4

In today’s episode we explore the hardware and components that build the foundations of Med Tech and make AI viable within medical practices.

Join myself, Dr Felix Beacher as I talk to Sub Bhattacharya, Head of Healthcare & Sciences at AMD, as he discusses the hardware perspective of medical AI. We talk about the latest advancements in healthcare technology, including the role of semi-conductors and processors in the technology that makes hand-held diagnostic devices available. Discover how the integration of AI technology is transforming healthcare delivery and research, and gain valuable insights into the future of medical AI.

Whether you're a medical professional or simply curious about the intersection of technology and healthcare, this episode offers a fascinating look at the evolving capabilities and feats of engineering in the world of healthcare technology. Don't miss this expert discussion on the cutting-edge intersection of healthcare and technology.

Sub Bhattacharya has extensive experience as the Head of AMDs healthcare and Sciences department. He brings 20+ years of leadership roles in business development, strategy, marketing, product management to build bridges between healthcare technology and semiconductors.

Original music by Felix Beacher: https://youtu.be/JS93808r-W4

Join me, Dr Felix Beacher, In this episode, we'll delve into the crossroads of AI and health informatics, guided by our expert guest Professor Mark Braunstein, Professor Braunstein shares his expertise in the field of health informatics and reflects on his book “Health Informatics on FHIR: How HL7's New API is Transforming Healthcare”. We explore the role of health informatics in enabling information exchange and standardization in healthcare and address how the landscape is changing through the use of FHIR (Fast Healthcare Interoperability Resources) standards. With the implementation of these standards, healthcare data can be easily shared and analysed, enabling the development of more effective AI applications.

We discuss the challenges and opportunities that arise from using AI in health informatics, including issues of data privacy and security, ethical concerns, and the impact on healthcare professionals and patients. This is not an episode to miss, tune to 30 minutes of cutting-edge content!

An author and thought-leader in the field, Dr. Braunstein is a professor of health informatics for the School of Interactive Computing in the College of Computing at the Georgia Institute of Technology where he teaches Introduction to Health Informatics, a popular graduate seminar, both on campus and as part of the OMSCS program.

He is involved with HL7 and in various research projects at Georgia Tech, and works with and advises numerous community and industry groups.

Original music by Felix Beacher: https://youtu.be/JS93808r-W4

In this episode of the podcast, we sit down with none other than ChatGPT itself for a fascinating conversation about the intersection of AI and healthcare. With a touch of humour, we explore the ways in which GPT-3 has the potential to assist healthcare tasks such as medical documentation, clinical trial analysis, and drug discovery.

We dive into the world of GPT medical and how it is being used for symptom-checking by patient and doctor alike. We also discuss CPT-3’s model use in start-ups like Babylon Health and Informatica. As we delve deeper into the world of healthcare informatics, chatGPT walks us through how NLP models such as BERT, RoBERTa, and ELMo, are affecting the future of healthcare. We also consider the ethical implications of using AI in healthcare, including the potential for dehumanization and the importance of maintaining a human touch in patient care. Tune in to explore the exciting possibilities of GPT-3 in healthcare and beyond.

ChatGPT is an AI language model developed by OpenAI. It uses machine learning to generate human-like responses to text-based prompts, such as questions or statements. Essentially, it is a computer program that can "chat" with humans and respond in a way that resembles human conversation.

Original music by Felix Beacher: https://youtu.be/JS93808r-W4

Join me, Dr Felix Beacher, and this weeks guest, Eric Olson, CEO of Consensus, as we discuss the role of clinical science search engine Consensus. We ask how accessibility is driving innovation in the field of medical research and how these search engines are designed to help researchers find relevant scientific literature quickly and easily. Olson describes consensus as a bridge between platforms like Google Scholar and Chat GPT.

We also explore the development of these search engines and the complex algorithms that make them so effective in organizing and presenting scientific research. How does the platform account for ethical considerations such as author biases, potential misinformation, or conversely correctly displaying amateur research papers that meet search criteria. You'll gain valuable insights into the way technology is shaping the future of medical study and research.

If you're interested in the latest advances in scientific research and the tools that are driving innovation in the field, then tune in to discover more about the clinical science search engine start-up, Consensus.

Eric Olson is a data science professional, with a Master's in Predictive Analytics from Northwestern University. Olson came up with the idea for Consensus after spending years as a die-hard amateur science consumer and wanting access to better evidence-based information at the click of a button. Prior to Consensus, Eric worked at the sports entertainment company DraftKings, where he served as an Analytics Specialist building predictive machine learning models to better understand the user base.

Original music by Felix Beacher: https://youtu.be/JS93808r-W4

Join me, Dr Felix Beacher, for a new episode as I sit down with Madhuri Sebastian, Senior Vice President and Global Head of Enterprise Imaging at Philips. Madhuri is a healthcare technology expert with over 20 years of experience in medical imaging.

Sebastian’s specialism takes us through integrated diagnostics and medical imaging. We discuss how integrated diagnostics lets clinicians build advanced profiles and subsequent care for patients through vast bioinformatic and genomic data. We also cover shared data and interoperability, which is how different healthcare systems and devices communicate with one another. She discusses the challenges that come with this, particularly in the context of bridging fractured data systems that exist within healthcare. Finally, Sebastian also discusses the advancements echocardiograms imagery and cardiology treatments that have been made possible through AI in med tech.

Madhuri Sebastian is the GM and Head of Data Management & Interop Solutions at Philips, based in Boston. Madhuri joined Philips in July 2021, having extensive previous experience as VP of AI Clinical Partnerships. She has partnered with major academic institutions in operational and workflow analytics.

Original music by Felix Beacher: https://youtu.be/JS93808r-W4

In Episode 6 of our podcast, we speak with Eyal Aharon, CEO of MedicVision, about how AI technology is being used to enhance CT and MRI capacities through enhanced imagery and less scan time.

Aharon’s motivations are driven by health care efficiency and safety as he explains how MedicVision's AI technology can reduce exposure to ionizing radiation in CT scans, making the process safer for patients and how AI can be used to reduce MRI scan time, making the process more efficient and less stressful. We explore insights on the potential impact of AI on the healthcare industry, including how it can improve patient outcomes through improved diagnosis and reduce costs. We discuss the global reach of med tech advancements and how MedicVision's technology is being used across the US, Middle East, Japan, China and Europe to improve quality of medical imaging and bring better care to patients. Finally, Aharon walks us through ethical considerations of continuity, phantom testing and how his company gained FDA approval.

Eyal Aharon is an alumnus of Cambridge College Technion Israel Institute of Technology and the CEO of MedicVision, an imaging solutions company at the forefront of AI medical imaging

Original music by Felix Beacher: https://youtu.be/JS93808r-W4