Facebook
Britain's News Portal
Around The Clock
BREAKING
Loading latest headlines…

AI chatbot suggested diagnosis for toddler's rare condition

A 19-month-old girl from Bath with a rare genetic condition reportedly received a diagnosis suggestion from ChatGPT, after her mother entered her symptoms.

  • Lily, 19 months old, was diagnosed with Multisystemic Smooth Muscle Dysfunction Syndrome (MSMDS) in January.
  • Her mother, Rosie, stated she used ChatGPT to input Lily's symptoms, which suggested MSMDS.
  • Genetic Alliance UK and University of Oxford research warn against using AI for medical diagnoses, especially for rare conditions.

The family of a 19-month-old girl from Bath, Lily, states that an AI chatbot helped them gain a diagnosis for her rare genetic condition. Lily, who had surgery for an unknown heart condition at four months old, was diagnosed with Multisystemic Smooth Muscle Dysfunction Syndrome (MSMDS) in January.

Her mother, Rosie, said she entered Lily's symptoms into ChatGPT, which suggested MSMDS. However, when she informed doctors, some had not heard of it, while others considered it too rare to be likely.

Nick Meade, chief executive at Genetic Alliance UK, urged caution regarding the use of ChatGPT, noting that rare diseases are particularly affected by AI modelling problems. Research from the University of Oxford earlier this year echoed this warning, with Dr Rebecca Payne stating that AI is not yet ready to take on the role of a physician and can give wrong diagnoses.

MSMDS is caused by a change in the ACTA2 gene and affects organs such as the heart and kidneys. Lily is one of only 70 known cases worldwide and one of six in the UK. Her parents have established a charity, ACTA2 Alliance UK, to fund research into the condition, with international research into potential treatments and a cure being led from Boston, America.

Why this matters: The case highlights the potential, but also the risks, of using AI tools for medical diagnoses, particularly for rare conditions where AI models may lack sufficient training data.

Related Articles

Get the news that matters.

Join thousands of readers getting the best of British news straight to their inbox.