Most AI models perform beautifully in the lab but fail when faced with real-world complexity.

In this episode of ByteSight, host Dr. Manasi A-Ratnaparkhe speaks with Dr. Heather Couture, Founder of Pixel Scientia Labs and host of the Impact AI podcast. Together, we explore why AI in digital pathology often fails to generalizeacross scanners, stains, and patient populations and what it really takes to build systems that are robust, explainable, and clinically meaningful.

With over two decades of experience bridging AI research and healthcare practice, Heather shares practical insights on dataset diversity, model validation, and the human factors that shape trustworthy AI.

Learn more about Dr.  Heather Couture.

In this episode of #ByteSight, our host Dr. Manasi A-Ratnaparkhe speaks with Dr. Catalina Lopez-Correa, Chief Global Strategy Officer at Genome Canada. Dr. CatalinaLopez-Correa is a globally recognized physician-scientist whose career spans pandemic preparedness, cancer research, and international health diplomacy. Together, we explore what it means to bring genomics „out of the lab and into the living room“ from pandemic preparedness to cancer care, while tackling the real-world consequences of inequity in genomic databases. Learn more about Dr.  Catalina Lopez-Correa: https://www.linkedin.com/in/clopezcorrea/

In this episode of ByteSight, Prof. Segun Fatumo explains why Africa’s genomic diversity is central to both understanding human evolution and shaping the future of global health. He reminds us that while most genomic research has focused on European populations, Africa remains the most genetically diverse continent, and excluding it means missing discoveries that could benefit everyone.

Follow his work: https://x.com/sfatumo?lang=en

In this episode of ByteSight, we’re joined by Dr. Manuel Corpas, a Senior Lecturer in Genomics at the University of Westminster, Fellow of the Alan Turing Institute, and internationally recognized leader in genomics, bioinformatics, and health data science.

Dr. Corpas shares his mission to make personalized medicine more accessible, especially in underrepresented populations. From leading the Spanish Congress of Genomic Medicine to shaping ethical AI use in healthcare, we explore how his work is driving genomic equity, public engagement, and innovation across borders.

Learn more about his work: manuelcorpas.com

In this episode of ByteSight, host Dr. Manasi A-Ratnaparkhe, co-founder and CEO of PAICON, sits down with with Dr. Rohit Thanki, an AI researcher with hands-on experience across India, Dubai, and Germany, to explore the complex reality of AI in healthcare. From the allure of AI models to the sobering challenges of data silos, regulatory bottlenecks, and lack of global applicability, Dr. Tanki shares deep insights on what it takes to move beyond the buzzwords.

We discuss:

• Key challenges in AI adoption in healthcare, from data access to harmonization

• The need for technical solutions like watermarking for medical data security

• Lessons from global AI implementation – India’s state-driven efforts, Dubai’s fast-paced rollout, and Europe's regulatory focus

• The "Remaining 84%" problem in global AI applicability and the role of demographic diversity

If you're building AI for health, this episode is your reminder that accuracy alone isn’t enough—equity, security, and regulation must be part of the equation.

How do you build trustworthy AI in healthcare? Start with the data.

In the first episode of ByteSight, host Dr. Manasi A-Ratnaparkhe, co-founder and CEO of PAICON, sits down with Dr. Uwe Schalles, a diagnostics veteran with 25+ years of experience at Roche and Ventana, to unpack the full data lifecycle from acquisition and curation to quality control and access for pharma R&D.

Together, they explore:

• Why sourcing high-quality H&E slides is harder than it looks

• How technical diversity (scanners, staining, formats) impacts AI reliability

• Mobile phones in pathology—gimmick or emerging frontier?

• What it takes to push algorithm boundaries and stay within regulatory frameworks

• The systemic nature of medical AI: from fixation protocols to FDA-certified scanners

Whether you're building AI, validating datasets, or shaping the future of digital health, this episode gives you a rare behind-the-scenes look at what really drives performance and trust in medical AI.