#269 Computer Vision Diagnosing Genetic Disorders with Robert Boscacci

Can computers diagnose genetic conditions? Could we train computers to analyze photos of people to help geneticists and other healthcare providers diagnose conditions based on physical features? 

Join the visionary journey as our host Kira Dineen sits across from the insightful Robert Boscacci, the mind behind "The Spring of Sight," and unravels the intertwining of computer vision and genetic diagnosis. Plunge into a dialogue that traverses Rob's transition from the digital effects of television to the cutting-edge of AI, illuminating the intricate process of equipping computers to decode our visual world and its transformative impact on medical science. Our discussion explores the precise identification of genetic disorders, offering a glimpse into a future where technology and genetics coalesce with astonishing clarity. 

The episode ascends further as we scrutinize the nuanced realm of data and bias in teaching machines to see. We dissect the transition from traditional rule-based algorithms to the nuanced world of data-driven AI learning, recognizing the significance of diverse datasets in averting biases—a fundamental aspect when addressing diagnoses across a spectrum of conditions, like the distinctive features of Treacher Collins syndrome. Our conversation turns to the challenges healthcare faces in adopting AI, from stringent validations to the delicate dance of applying AI without overshadowing human expertise. Dive with us into the profound implications of AI in healthcare, where the fusion of human and machine intelligence could redefine the horizons of disease diagnosis and treatment.

In The Spring of Sight, Robert Boscacci draws from his double background in digital film production and machine learning to explore the utility of computer vision with industry experts. In the burgeoning field of computer vision, researchers and entrepreneurs build systems to extract meaning from digital images in unprecedented new ways, and even generate convincing new synthetic images. As advancements in the field carry in a tidal wave of promising use cases, it becomes increasingly clear that we'll have to implement nuanced policies to contend with their double-edged nature.

Robert Boscacci is the former Data Science Lead at Butter Works, a startup that used computer vision to analyze hundreds of thousands of social media videos for clients like Disney+, Netflix, Spotify, and ViacomCBS. He comes from a background in New York City’s film post-production industry, with IMdB credits as a film colorist and dailies technician on Netflix features and episodic productions.

Boscacci is curious about how tech policy can become more proactive, so that communities and organizations can maximize the utility and minimize the harm of the cutting edge. He hopes to encourage readers from all walks of life to join the conversation around tech and policy—so that the folks who normally get steamrolled in the name of progress are more empowered to make their voices heard, and those driving the steamrollers become more aware of themselves.

Boscacci is the caricature of a millennial Brooklyn hipster: Picture him wearing his nearly non-prescription glasses, dismounting his fixie, and tapping to pay to slurp a single-origin espresso on his way to the nonprofit repertory cinema. He un-ironically holds a fishing license from the state of New York and uses it to catch Bluegill in the lake at Prospect Park.

Episode Segments

0:00:43 - Computer Vision for Diagnosing Genetic Conditions 

0:04:25 - Exploring the Concept of Computer Vision 

0:09:55 - Diagnosing Genetic Conditions With Computer Vision

0:12:07 - Transfer Learning for Vision Models

0:20:40 - Computer Vision in Diagnosing Rare Diseases 

0:26:51 - The Importance of Embracing Medical Technologies

Be sure to enter our giveaway to win a signed copy of Rob’s book, “The Spring of Sight” via our X (Twitter), Instagram, Threads, and LinkedIn. Look for the post of me and Rob with the bo