Pratik Desai, an AI expert, faced a daunting challenge when his mother was diagnosed with stage IV duodenal adenocarcinoma. After her diagnosis, the doctors released her without connecting her to an oncologist, indicating that her prognosis was grim. Taking matters into his own hands, Desai employed advanced AI tools to meticulously examine his mother’s extensive medical documentation, which spanned over 1,600 pages. Utilizing NotebookLM and Claude, he sought out specialists, educated himself on her condition, and monitored her symptoms and concerns.
His AI-driven analysis uncovered significant inaccuracies within her CT scan report, including three instances of misidentified cancer types. Furthermore, the technology revealed that changing his mother’s diet from liquid to solid food just 48 hours post-blood transfusion led to severe bleeding, placing her in critical condition on two occasions.
Desai is convinced that AI has granted his mother additional time, as she lived for 76 days following her diagnosis, allowing her to say farewell to family and friends. He has since shared his AI-assisted system with others; a friend employed it to gain insights into his own mother’s treatment, impressing her medical team with his newfound understanding, despite having no medical background.
While some medical professionals have voiced concerns regarding the reliability of AI, noting it has an accuracy of only about 70%, Desai counters this by suggesting that the medical field also has its share of errors. "I replied: let’s assess the healthcare system with the same standards. We expect AI to be flawless, yet the medical community is not," Desai asserts.
Would you place your trust in AI for medical decisions? ❤️ — Yes, as doctors can also err 🤔 — No, only a qualified physician
Informational material. 18+.
