An artificial intelligence tool has assisted in identifying a uncommon neurological disorder in a woman from Wales after she spent four years being misdiagnosed by healthcare practitioners. Phoebe Tesoriere, 23, from Cardiff, was first informed by doctors that she was experiencing anxiety, epilepsy and depression, despite presenting with increasingly severe symptoms such as seizures, movement difficulties and balance problems. Following a major seizure that resulted in a coma for three days in July 2025, Phoebe consulted ChatGPT to investigate her condition. The AI tool suggested several conditions, among them hereditary spastic paraplegia—a uncommon inherited condition affecting the nervous system. After presenting this finding to her GP, genetic analysis verified the finding, at last offering answers after prolonged periods of frustration and poor management in the NHS.
A Four-Year Journey Through Medical Uncertainty
Phoebe’s health issues began long before her identification of the condition. During her childhood, she went through a ongoing difficulty walking, which she traced back to being born without a hip socket and undergoing remedial surgery as an infant. She also struggled with coordination issues and was screened for dyspraxia, a neurological condition impacting physical coordination, though the findings proved negative. These initial signs would eventually turn out to be noteworthy in understanding her root cause, yet at the point in time they continued to be unaccounted for and largely dismissed by healthcare practitioners.
The situation declined considerably when Phoebe was 19 years old. She collapsed and suffered a seizure whilst at work, a frightening experience that should have prompted comprehensive enquiry. Instead, doctors linked the episode with anxiety—a diagnosis that was thereafter recorded in her medical records despite Phoebe having no previous experience with anxiety disorders. She described herself as “a really happy, bubbly person” before this incident, making the diagnosis seem particularly incongruous. This misdiagnosis would establish the pattern for extended periods of unsuitable care and mounting frustration.
- Childhood gait abnormality resulting from hip surgery, not underlying neurological condition
- Balance difficulties assessed for dyspraxia but results proved negative
- First seizure at 19 incorrectly identified as anxiety episode
- Anxiety diagnosis added to medical records without proper assessment
The Game-Changing Breakthrough: ChatGPT’s Unforeseen Discovery
After passing 72 hours in a coma following a severe seizure in July 2025, Phoebe found herself at a critical juncture. Upon regaining consciousness, a doctor provided a bewildering statement: she did not have epilepsy after all, but rather anxiety. This conflicted with extensive prior treatment and the previous epilepsy diagnosis she had received in 2022. Exasperated with the cyclical pattern in her medical journey and feeling unheard by healthcare professionals, Phoebe made the choice to turn to an unconventional source for answers. She inputted her full list of symptoms into ChatGPT, the AI chatbot that has become more common in healthcare discussions.
The AI tool’s output proved remarkably comprehensive. ChatGPT produced a list of possible conditions that might explain Phoebe’s constellation of symptoms—advancing loss of strength, coordination problems, seizures, and periods of paralysis. Among the suggestions was a rare genetic neurological disorder affecting the spinal cord, a uncommon hereditary neurological disorder that affects the spinal cord and causes progressive stiffness and weakness in the legs. What distinguished this suggestion from earlier clinical assessments was its specificity and the way it cohesively explained multiple symptoms that had previously been fragmented across various diagnoses. Phoebe quickly recognised that this condition might finally provide the single coherent understanding she had been seeking.
From Scepticism to Affirmation
Armed with the recommendation from ChatGPT, Phoebe approached her GP with details regarding hereditary spastic paraplegia. Rather than outright rejecting the AI-generated hypothesis, her doctor treated the recommendation with seriousness and arranged genetic testing. This practical strategy turned out to be transformative. The genetic tests came back positive, confirming that Phoebe genuinely had hereditary spastic paraplegia—validating both the AI chatbot’s analysis and, more importantly, finally delivering a definitive diagnosis after four years spent dealing with medical misdiagnosis and uncertainty.
The confirmation marked a significant moment for Phoebe, though it also underscored the limitations of her previous medical care. Her GP, Dr Rebeccah Tomlinson, has subsequently acknowledged the role artificial intelligence can have in healthcare research, whilst emphasising the importance of expert validation. She pointed out that when people use artificial intelligence tools to explore medical issues, these results should always be discussed with qualified medical professionals before drawing conclusions. This balanced perspective acknowledges both the possible advantages of artificial intelligence in medicine and the irreplaceable value of professional medical expertise.
Learning about Hereditary Spastic Paraplegia
Hereditary spastic paraplegia (HSP) is a uncommon hereditary neurological condition characterised by increasing weakness and rigidity in the legs. The condition affects the spinal cord, particularly the nerve fibres responsible for leg movement. HSP presents in multiple forms, with more than 80 genetic variations identified, presenting diagnostic challenges for medical professionals. Symptoms usually appear gradually and can involve problems with walking, coordination issues, reduced muscle strength, and in some cases, seizures. The disease’s scarcity means most clinicians have limited experience identifying it, which partly explains why Phoebe’s condition was undiagnosed for so long despite presenting characteristic features of the disorder.
| Aspect | Details |
|---|---|
| Primary Affected Area | Spinal cord and nerve fibres controlling leg movement |
| Genetic Variants | Over 80 known genetic forms of the condition |
| Common Symptoms | Progressive leg weakness, stiffness, balance difficulties, and occasionally seizures |
| Inheritance Pattern | Can be inherited in autosomal dominant, autosomal recessive, or X-linked patterns depending on genetic variant |
The intricacy of HSP’s hereditary variation presents substantial diagnostic difficulties. With various modes of inheritance and inconsistent symptom manifestations across different forms, even experienced neurologists can find it difficult to recognise the condition without DNA analysis. Phoebe’s case highlights how uncommon hereditary conditions can be overlooked when symptoms coincide with frequently occurring disorders like epilepsy or anxiety disorders, emphasising the essential significance of thorough genetic investigation when conventional diagnostic approaches do not adequately account for a patient’s clinical presentation.
The Expanding Conversation Concerning AI in Medical Care
Phoebe’s situation has revived discussions about the function of artificial intelligence in medical diagnosis and patient care. Whilst her case demonstrates AI’s capacity to detect unnoticed disorders, medical professionals and researchers advise against regarding chatbots as diagnostic tools. A recent Oxford University study showed that individuals pursuing healthcare advice through AI were given unreliable recommendations, spanning from useful guidance to risky suggestions. This inconsistency poses significant challenges for individuals trying to distinguish reliable guidance from incorrect suggestions, particularly when handling rare or complex ailments that demand expert expertise and thorough medical assessment.
The incident also highlights important questions about patients’ right to make decisions and the responsiveness of healthcare systems to individuals who feel unheard. Many patients turn to AI tools out of disappointment when conventional medical routes don’t work, highlighting potential gaps in diagnostic processes. Phoebe’s decision to turn to ChatGPT arose from experiencing profound loneliness during her medical journey and the exhaustion of fighting to be believed. This points to a wider issue that patients more often turn to alternative resources when traditional medical systems fail to deliver solutions, indicating that improvements in diagnostic protocols and communication with patients may be equally important as setting out clear rules for AI tool usage in medical contexts.
Expert Perspectives on Artificial Intelligence Healthcare Solutions
Dr Rebeccah Tomlinson, a general practitioner, acknowledges that patients may legitimately use AI chatbots to investigate medical issues but emphasises the critical importance of discussing findings with qualified medical professionals. This measured approach acknowledges individuals’ entitlement to obtain knowledge whilst preserving professional oversight. The BMA and other healthcare bodies have likewise recommended for AI integration within formal clinical systems rather than as a substitute for clinical assessment. Experts stress that artificial intelligence systems should complement rather than bypass clinical expertise, particularly given the complexity of uncommon hereditary disorders requiring specialist knowledge and genetic testing confirmation.
Cardiff and Vale University Health Board’s handling of Phoebe’s case acknowledged her difficult experience whilst implicitly defending the challenges doctors face when identifying uncommon disorders affecting a small number of patients. Medical professionals contend that hereditary spastic paraplegia’s scarcity and multiple genetic variants make it inherently difficult to identify without targeted genetic analysis. However, the case has prompted reflection within the medical sector about improving diagnostic pathways for patients with atypical presentations. Healthcare leaders increasingly recognise that developing improved communication frameworks and reduced barriers for referrals for genetic testing could prevent similar diagnostic delays whilst upholding strict clinical criteria.
- AI should support clinical decision-making, not replace medical expertise and diagnostic assessment
- Patients employing AI-based tools must discuss findings with certified medical practitioners prior to taking action
- Healthcare systems must improve diagnostic protocols for uncommon disorders and unusual symptom patterns
Moving Forward: Life Following Diagnosis
Since getting her confirmed diagnosis of hereditary spastic paraplegia in 2025, Phoebe Tesoriere has begun the process of adjusting to life with a clear understanding of her condition. The genetic confirmation has provided her with answers after years of uncertainty and misdiagnosis, allowing her medical team to develop a more targeted treatment approach. Phoebe has become an advocate for improved diagnostic pathways, publicly sharing her experience to increase understanding of hereditary spastic paraplegia amongst both patients and healthcare professionals. Her experience has underscored the importance of listening to patients who consistently describe symptoms that don’t match conventional diagnoses, and she continues to work with healthcare providers to manage her condition effectively.
Phoebe’s path has also sparked broader dialogue within the NHS about diagnostic protocols for rare neurological conditions. Whilst she accepts the real difficulties doctors deal with when identifying uncommon genetic disorders, she remains committed to helping others avoid the four-year diagnostic journey she experienced. Her case has prompted reflection amongst medical professionals about reducing barriers for referrals for genetic testing and improving communication with patients presenting with unusual symptoms. Going forward, Phoebe believes her story will encourage both clinicians and patients to keep searching for answers, illustrating that uncommon conditions, though difficult to diagnose, should never be regarded as merely psychological.