Impulse - Meeting Healthcare Pioneers

One of the greatest fantasies in medicine is to one day be able to offer any patient a therapy perfectly adapted to their profile and needs. A "personalized" medicine. The term was introduced to the public in 1999 through an article in The Wall Street Journal titled "New Era of Personalized Medicine: Targeting Drugs for Each Unique Genetic Profile." The enthusiasm for these next-generation therapies was fueled by advances in DNA sequencing, which hinted at future discoveries in genomics and rapid progress in medicine. Twenty-five years later, it is clear that personalized medicine has become a reality (e.g., for the treatment of cancer or rare diseases), and technological progress continues to bring us closer to this ideal. But while developing new personalized therapies is one approach, identifying the most effective current therapies for each patient is another, equally powerful one. And thanks to the multimodal analysis capabilities enabled by AI, this field is rapidly expanding. Andre Esteva is one of the pioneers in this area, having significantly advanced medical research in this direction before venturing into entrepreneurship to make this technology available broadly. In less than three years, his company Artera has already achieved many things that most healthtech startups take much longer to reach: The launch of the first AI-based predictive test to evaluate treatment response to hormonal therapy in prostate cancer. The inclusion of the test in the National Comprehensive Cancer Network (NCCN) guidelines nationwide. The obtention of federal reimbursement for the test. In this episode, Andre shares his vision of predictive medicine in the AI era, the potential of this technology for complex diseases such as cancer, and the keys to their success and recognition on the international medical scene! Timeline: 00:03:24 - Andre’s background as a researcher and entrepreneur in medical AI 00:06:01 - The origins of Artera and the vision of personalizing therapy for cancer patients 00:09:29 - Why AI is so relevant for predictive and prognostic medical applications 00:11:51 - How the ArteraAI Prostate Cancer test works 00:19:57 - Getting the test included in national guidelines within a year 00:24:16 - Developing further therapy prediction tests for cancer and beyond 00:26:03 - Lessons learned from securing national reimbursement within two years 00:29:22 - Andre’s view on current and future applications of AI in medicine What we also talked about with Andre: Salesforce H&E stain Prostate-specific antigen Chemotherapy Computer vision Large Language Models (LLMs) AI agents AlphaFold NeurIPS Computer Vision and Pattern Recognition Conference Marc Benioff We cited with Andre some of the past episodes of the series: #43 - Rethinking medical prescriptions to save lives - Emmanuel Bilbault - Posos #41 - Making sense of the gut microbiome - Leo Grady - Jona #40 - Cracking biology through AI - Jean-Philippe Vert - Owkin #36 - Turning healthcare preventive with full-body MRI scans - Andrew Lacy - Prenuvo As Andre mentioned during the episode, you can read their landmark paper published in npj Digital Medicine in 2021 to dive deeper into the technology they have developed. For further information about Artera’s publications and all the scientific work they do, take a look at the News & Insights section of their website. You can follow Artera’s activities on LinkedIn and X, and get in touch with Andre via LinkedIn! If you want to give me feedback on the episode or suggest potential guests, contact me over LinkedIn or via email! And if you liked the episode, please share it, subscribe to the podcast, and leave a 5-star review on streaming platforms! 🙏 There’s now a bi-monthly newsletter around the podcast where you will be informed of the latest episodes and kept posted on the latest medical tech progress, subscribe here! Lastly, don’t forget to follow our activities on LinkedIn and our website!

To discover the whole episode type "#44 - Predicting cancer treatment response - Andre Esteva - Artera" on your preferred streaming platform.

According to the WHO, unsafe medication practices and medication errors are a leading cause of avoidable harm in health systems across the globe. In the European Union alone, this translates to nearly 200,000 deaths annually. The magnitude of this problem has led to global initiatives such as "Medication Without Harm," launched in 2017, aiming to reduce medical errors by 50% over five years. Although there is a need for more data on the results of this initiative, we are far from having reached the goal. The key lies in our ability to access medical information, specifically related to prescriptions. Direct access to reliable information, accounting for individual patient characteristics, and able to fit with the tools used by healthcare professionals. A next-generation medical database, incorporating all known information about medications in structured and interoperable bricks capable of supporting caregivers in safer and more effective prescribing practices. It turns out this database exists and has been developed by Pharmacist Emmanuel Bilbault and his team at Posos! Approved less than a year ago by French health authorities, it offers a new perspective on how to approach medical prescriptions in the era of AI and digital health. A way to give back time to every healthcare professional and ensure an unprecedented level of care quality for patients. We talked with Emmanuel about the creation of this new tool - used today by more than 85,000 healthcare professionals in Europe - and what it changes in our fight against medical errors. Timeline: 00:02:56 - What pushed Emmanuel to quit the Pharma industry and found Posos 00:11:11 - The first multilingual, structured, and encoded drug database to streamline medical prescribing 00:16:31 - How Posos is leveraging AI to supercharge its platform 00:22:54 - Keeping the drug database up to date across geographies 00:24:37 - The impact Posos has on healthcare professionals, patients, and health systems 00:29:59 - Toward providing patient access to Posos’ drug database 00:32:40 - Lessons learned from scaling Posos in France and abroad 00:40:47 - Emmanuel’s view on the current healthcare transformation What we also talked about with Emmanuel: Teva Pharmaceuticals IBM’s Watson Health Electronic Health Records SNOMED CT ICD-10-CM ATC MedDRA Named Entity Recognition Optical Character Recognition Word Embedding Machine Learning Pharmagest Ambient AI Nabla We cited with Emmanuel some of the past episodes of the series: #1 - Measuring your arterial age to prevent heart diseases - Ted Baldwin - Imageens As mentioned by Emmanuel during the episode, we recommend taking a look at this page to learn more about the SNOMED CT clinical terminology, as the most comprehensive, multilingual clinical healthcare terminology in the world. You can follow Posos’ activities on LinkedIn, Instagram, and X, and get in touch with Emmanuel via LinkedIn! If you want to give me feedback on the episode or suggest potential guests, contact me over LinkedIn or via email! And if you liked the episode, please share it, subscribe to the podcast, and leave a 5-star review on streaming platforms! 🙏 There’s now a bi-monthly newsletter around the podcast where you will be informed of the latest episodes and kept posted on the latest medical tech progress, subscribe here! Lastly, don’t forget to follow our activities on LinkedIn and our website!

To discover the whole episode type "#43 - Rethinking medical prescriptions to save lives - Emmanuel Bilbault - Posos" on your preferred streaming platform.

What will the future of surgery look like? According to Dr. George Papagiannakis, we must count on virtual simulation. After all, simulators are an integral part of professional training in highly demanding fields like aviation or F1. So why not medicine? The approach might seem futuristic, but George and his team at ORamaVR share a vision that makes a lot of sense. Integrating extended reality in surgery means increasing the volume of hours spent operating. It means preparing each surgery on a virtual patient with properties close to the real case. It means simulating complications and unforeseen situations safely. And it means reducing the risk of errors in the real world. Following his motto "never first time with the real patient", George immerses us in the universe of extended reality and reveals its potential for medicine and surgery. What is extended reality? What makes a good virtual medical simulation? And what is the impact of this technology today and in the future? Many questions that George addresses with passion and enthusiasm, in a conversation reminiscent of Ready Player One! Timeline: 02:49 - Addressing the millions of fatal medical errors worldwide 05:20 - Defining extended reality and underlying technologies 07:31 - Medical applications of extended reality 10:52 - The impact of extended reality on medical training and practice 16:25 - What makes a virtual medical environment feel real 23:32 - How to recreate the feel of touch and dexterity virtually 29:28 - Enabling all healthcare professionals to create virtual environments with AI What we also talked about with George: Prof. Shafi Ahmed Apple Vision Pro Magic Leap Microsoft HoloLens NHS The Anatomy Lesson of Dr. Nicolaes Tulp by Rembrandt IKEA Global illumination Dall-E ChatGPT NotebookLM Frost & Sullivan Meta We cited with George some of the past episodes of the series: #9 - Creating the Gran Turismo of surgery - Raimundo Sierra - Virtamed #12 - Unleashing the power of digital health - Daniel Kraft - Digital.Health #32 - Accelerating radiology with AI - Amine Korchi - Radiologist #36 - Turning healthcare preventive with full-body MRI scans - Andrew Lacy As mentioned by George during the episode, we recommend taking a look at some of the simulations available on the official YouTube channel of ORamaVR, it will bring to life many aspects that we discuss in the episode! You can follow ORamaVR’s activities on LinkedIn, Instagram, and get in touch with George via LinkedIn! If you want to give me feedback on the episode or suggest potential guests, contact me over LinkedIn or via email! And if you liked the episode, please share it, subscribe to the podcast, and leave a 5-star review on streaming platforms! 🙏 There’s now a bi-monthly newsletter around the podcast where you will be informed of the latest episodes and kept posted on the latest medical tech progress, subscribe here! Lastly, don’t forget to follow our activities on LinkedIn and our website!

To discover the whole episode type "#42 - Extending reality to transform surgery - George Papagiannakis - ORamaVR" on your preferred streaming platform.

Our gut contains a world of microorganisms that impact many aspects of our health, well beyond the gastrointestinal system. In fact, more than 30 trillion bacteria, archaea, fungi, and viruses live in the digestive tract of every human, a number that is higher than the number of human cells that make up our body! Each microbiome is unique to an individual, and its genetic heritage represents approximately 100 times that of the human genome. This field is fascinating and science is advancing rapidly, as evidenced by the 2,000 publications generated monthly on the subject. A better understanding of this remarkable ecosystem might bring critical answers to address numerous pathologies and health disorders. This is the mission of Leo Grady and his team at Jona, whose technology enables microbiome analysis based on an at-home test, combined with artificial intelligence to reveal what the current literature says about each person's microbiome. Unveiling the propensity for certain disease risks, along with individual nutritional preferences, Jona can model the body's reaction according to potential changes in the microbiome, and provide valuable insights on behaviors and nutrition that can ultimately improve one’s health. A dive into the heart of the microbiome and its secrets, which today's technology is about to unlock! Timeline: 02:32 - Leo’s journey from his PhD in computer vision to AI applications in medicine 06:47 - The idea behind Jona and understanding the microbiome better 10:33 - What we do know and ignore about the microbiome 13:05 - The Jona test kit and what it can interpret from one’s microbiome picture 20:17 - The volume and complexity of studies around the microbiome 22:11 - The impact of Jona and Leo’s son's story 25:38 - Who is Jona for 29:00 - Leo’s vision for the future of Jona and AI in healthcare What we also talked about with Leo: Neural networks Siemens HeartFlow Paige AI Memorial Sloan Kettering Crohn’s disease Thomas Clozel We cited with Leo some of the past episodes of the series: #32 - Accelerating radiology with AI - Amine Korchi - Radiologist #35 - Making surgery smarter and safer - Björn Von Siemens - Caresyntax #36 - Turning healthcare preventive with full-body MRI scans - Andrew Lacy - Prenuvo #40 - Cracking biology through AI - Jean-Philippe Vert - Owkin As mentioned by Leo during the episode, we recommend looking at the Jona Journal which sheds light on the latest findings on the microbiome fueled by their research platform, and understanding the current literature. Leo also recommends listening to The Mind-Gut Conversation Podcast by Emeran Meyer to find out more, and We Contain Multitudes by Sarah Henstra as a complementary read. You can follow Jona’s activities on LinkedIn, X, Instagram, and YouTube, and get in touch with Leo via LinkedIn. If you want to give me feedback on the episode or suggest potential guests, contact me over LinkedIn or via email! And if you liked the episode, please share it, subscribe to the podcast, and leave a 5-star review on streaming platforms! 🙏 There’s now a bi-monthly newsletter around the podcast where you will be informed of the latest episodes and kept posted on the latest medical tech progress, subscribe here! Lastly, don’t forget to follow our activities on LinkedIn and our website!

To discover the whole episode type "#41 - Making sense of the gut microbiome - Leo Grady - Jona" on your preferred streaming platform.

All experts agree: if there's one sector that AI will transform, it's healthcare. Whether it's understanding pathologies, developing personalized treatments, the doctor-patient relationship, diagnostic methods, or creating tools to facilitate the work of healthcare professionals, everything is affected by the ongoing transformation! This revolution raises many expectations and questions about the real impact these new technologies will have on tomorrow's medicine. One of the leading figures in this movement on the international scene is none other than the French-American company Owkin, which aims to be "the first end-to-end AI biotech company," whose R&D activities are overseen by Jean-Philippe Vert, himself a true pioneer in the field of machine learning and computational biology. The first “AI-first” biotech company of its kind, bringing a breath of fresh air to an industry known for being extremely costly and time-consuming. So how does AI accelerate the development of new therapies? What is its role in identifying new molecules and targets? How can it be used to create new diagnostic tools? Can we leverage patient data without compromising their safety and privacy? And what can we expect from future models and their impact on our understanding of biology? These are all fascinating questions that Jean-Philippe breaks down with pedagogy and realism, shedding light on one of the most exciting fields of medicine in this era! Timeline: 04:21 - Jean-Philippe’s background at the interface of mathematics, computer science, and life sciences 07:10 - Owkin’s mission to become the first end-to-end AI biotech firm 09:48 - Leveraging AI for drug discovery and new treatment modalities 14:35 - Selecting target indications and therapeutic areas 17:13 - Changing the way we run clinical trials through AI 22:03 - Unlocking patient data across hospitals while preserving safety and privacy 25:40 - Developing AI-based diagnostics and driving clinical adoption 34:50 - Towards the first AI foundation model to understand biology What we also talked about with Jean-Philippe Google DeepMind Large Language Models (LLMs) Digital twins External control arms Federated learning Digital pathology MSIntuit CRC Bioptimus Neural networks As mentioned by Jean-Philippe during the episode, we recommend taking a look at Owkin and Bioptimus websites as well as their respective Publications & News section to learn more about their technology and stay posted on their latest activities. If you are a data scientist or machine learning enthusiast, you can tackle one of the many biology-related challenges proposed by Kaggle on their platform. To learn more about Owkin’s portfolio and the diversity of biomarkers and health conditions they cover, take a look at their Pipeline section. You can follow Owkin’s activities on LinkedIn, X, and YouTube, and Bioptimus on LinkedIn and X. To get in touch with Jean-Philippe, contact him over LinkedIn. If you want to give me feedback on the episode or suggest potential guests, contact me over LinkedIn or via email! And if you liked the episode, please share it, subscribe to the podcast, and leave a 5-star review on streaming platforms! 🙏 There’s now a bi-monthly newsletter around the podcast where you will be informed of the latest episodes and kept posted on the latest medical tech progress, subscribe here! Lastly, don’t forget to follow our activities on LinkedIn and our website!

To discover the whole episode type "#40 - Cracking biology through AI - Jean-Philippe Vert - Owkin" on your preferred streaming platform.