Founder fighting lymphoma turned to AI—and a data-driven mindset—to guide his treatment

Conno Christou, an AI company founder, did everything right. He tracked his sleep with a Whoop band, cross-referenced it with an Oura ring, and had nearly 100 biomarkers checked annually. At 35, he was as disciplined about longevity as anyone in his network.

Then, after a workout in 2025, his arm swelled—a discovery that led to surgery, pre-op imaging, and a diagnosis that changed everything: an aggressive, fast-growing non-Hodgkin's lymphoma affecting roughly one in 420,000 people.

Multiple Opinions, a Harder Path#

Christou's first oncologist recommended a lighter chemotherapy regimen with roughly a 60% success rate. But the night before his first infusion, he sought a second opinion—and got the opposite advice. A second doctor recommended a harder, six-month regimen with an 85% success rate. Over two days, Christou gathered 12 medical opinions. Eleven to one backed the aggressive treatment. He took it.

The decision reflected his approach: Christou treated chemotherapy like a company sprint cycle, logging data obsessively—blood results, wearable metrics from his Whoop, symptom journals via voice transcription, and scan images. He focused on three variables: sleep, nutrition, and psychology.

AI as a Conversation Partner#

Crucially, Christou fed all this data into Claude. He emphasizes that the AI "didn't replace the doctors," but "helped me ask the right questions." For a rare lymphoma an oncologist might see once yearly, he found that a model trained on the full body of medical literature offered something Google search could not.

The model proved critical at the end of treatment. His final PET scan came back ambiguous, and his oncologist began discussing radiotherapy near his heart and lungs—a serious escalation. Christou discovered that false-positive rates on end-of-treatment PET scans for his lymphoma subtype hover around 60%.

When he fed his scans into Claude, the model flagged a known but overlooked phenomenon: in patients under 40 recovering from this lymphoma type, the thymus gland can reactivate post-chemotherapy, mimicking active disease on imaging. Given his age and scan characteristics, Claude estimated a 90% probability of thymus rebound. A fourth medical opinion confirmed it—no active disease, no radiotherapy needed. He was clear.

A Broader Reckoning#

Christou is the founder of Keragon, an AI platform for medical practice administration. His experience as a patient has shifted his thinking. He watched clinicians buried in non-care tasks and saw chemotherapy protocols applied identically to patients of vastly different ages and circumstances.