Cardiff Woman Uses ChatGPT to Identify Rare Genetic Condition Following Four Years of Medical Misdiagnosis

Phoebe Tesoriere, 23, used ChatGPT to correctly identify her rare neurological condition after doctors repeatedly misdiagnosed her physical symptoms as anxiety.

By: AXL Media

Published: Apr 10, 2026, 8:11 AM EDT

Source: Information for this report was sourced from BBC Wales

The Failure of Conventional Diagnostic Pathways

For nearly half a decade, Phoebe Tesoriere experienced a series of debilitating physical symptoms that were consistently attributed to mental health issues rather than physiological pathology. Despite suffering from a lifelong limp, balance struggles, and a traumatic 2022 collapse that resulted in a three-day coma, Tesoriere claims that clinicians repeatedly labeled her condition as "anxiety" and "depression." At one point, she was reportedly warned that further visits to the emergency department would result in her being treated exclusively as a mental health patient, leaving her isolated and without a definitive clinical explanation for her deteriorating mobility.

Artificial Intelligence as a Diagnostic Catalyst

Exasperated by the lack of progress and a recent misdiagnosis of Todd’s Paralysis, Tesoriere entered her specific symptom set, including rigid limbs, loss of sensation, and hair loss, into the ChatGPT interface. The chatbot generated a list of potential rare conditions, prominently featuring Hereditary Spastic Paraplegia (HSP). Armed with this specific lead, Tesoriere presented the possibility to her GP, who agreed to facilitate genetic testing. In August 2025, the results confirmed she suffers from the complex quadriplegic form of HSP, a progressive condition caused by a rare gene mutation triggered by a childhood MRSA infection.

Systemic Pressures and the Rare Disease Gap

The case of Phoebe Tesoriere underscores the extreme difficulty the NHS faces in identifying rare diseases, which are often overlooked in high-pressure primary care environments. Hereditary Spastic Paraplegia is frequently misdiagnosed as cerebral palsy or multiple sclerosis due to its overlapping symptoms of muscle stiffness and weakness. Dr. Rebeccah Tomlinson, a Cardiff-based GP, noted that while it is "difficult for GPs to know everything," patients arriving with AI-generated information can help guide clinical discussions more effectively. However, the Cardiff and Vale University Health Board has officially stated it cannot comment on individual cases, though it has invited Tesoriere to discuss her care through their internal concerns team.