ai
Man Uses AI Model to Help Interpret Ambiguous PET Scan After Lymphoma Chemotherapy
A patient diagnosed with rare non-Hodgkin’s lymphoma gathered multiple medical opinions and fed scan data into Claude to resolve an ambiguous final PET result. The model identified thymus rebound with high probability, later confirmed by a fourth doctor.
TechCrunchAudio version
Tap play to generate a narrated version.
Developing·Limited corroboration so far. This page will refresh as more sources emerge.
Connor Christou received confirmation that he had no active disease after an ambiguous final PET scan by feeding three prior scans and an MRI into the AI model Claude, which attributed the result to thymus rebound with roughly 90 percent probability.
TechCrunch reported that Christou, 35, discovered an 11-by-11-by-8 centimeter mass behind his sternum during pre-operative exams for arm swelling and blood clots. A biopsy confirmed aggressive non-Hodgkin’s lymphoma, a diagnosis affecting roughly one in 420,000 people, with the tumor estimated to have existed for about three months.
His first oncologist recommended a lighter chemotherapy regimen with an estimated 60 percent success rate for his presentation. Christou booked the first infusion three days later but sought a second opinion the night before, which favored a harder regimen of continuous in-hospital infusion cycling every three weeks for six months, citing an 85 percent success rate.
Christou collected 12 opinions from hematologists and oncologists in the United States and abroad.
Eleven recommended the more aggressive path, which he completed over six cycles while wearing a Whoop band and maintaining a voice-transcribed symptom journal. He input blood results, scan data, wearable output, and journal entries into Claude throughout treatment. After the final PET scan returned ambiguous, the model flagged thymus rebound rather than active disease.
A fourth doctor confirmed the finding, eliminating the need for radiotherapy near the heart and lungs. Christou lives part time in Athens and built Keragon, an AI-powered platform that automates administrative operations for medical practices. He received the same chemotherapy protocol as an 80-year-old woman and now takes Sundays off.
