In an extraordinary medical case, ChatGPT, an artificial intelligence application, successfully diagnosed a four-year-old child’s chronic illness, a feat that had eluded 17 different doctors over three years. This diagnosis was not only a relief for the desperate family but also a moment of validation for AI’s potential in healthcare, as it was later confirmed by medical professionals.
The child in question, four-year-old Alex (surname withheld for privacy reasons), resides in the state of Michigan, USA. He had been suffering from intense chronic body pains and growth retardation for nearly three years. Despite repeated visits to various healthcare professionals since he was one, no conclusive diagnosis had been reached. In an attempt to alleviate his pain, Alex’s family had turned to a pain reliever called Motrin. Unfortunately, instead of providing long-term relief, the medication only led to several side effects and, as its effectiveness waned, Alex’s pains intensified.
Adding to the complexity, Alex began to bite on unusual objects – a behavior that perplexed both dental and orthodontic specialists. In a decisive move borne out of mounting frustration and concern, Alex’s mother, Courtney, turned to the artificial intelligence application ChatGPT for answers.
Courtney inputted a brief description of Alex’s symptoms and MRI results into the AI interface and hit ‚enter‘. Within seconds, ChatGPT identified Alex’s condition as „tethered cord syndrome“, a diagnosis that was subsequently confirmed by several doctors, leading to the initiation of proper treatment for Alex.
Reflecting on the journey, Courtney shared, „We had consulted many doctors, but all in vain. Some dentists even suggested the possibility of cavities, but that wasn’t the correct diagnosis either. However, when ChatGPT identified ‚tethered cord syndrome‘, it made absolute sense. I researched this condition online and later shared the findings with doctors, who agreed with the diagnosis.“
Courtney’s experience underscores the diverse diagnoses and treatments Alex underwent. From dental professionals suggesting palate issues to pediatric experts observing disproportionate growth between his body’s right and left sides, the range of conclusions was vast and disjointed. One specialist even erroneously identified Alex’s condition as the very rare Chiari malformation (cerebellar ptosis).
„But ChatGPT got it right in seconds,“ Courtney emphasized. After the AI’s diagnosis, she joined a Facebook group dedicated to families affected by the same rare condition, gaining insights and support from a community familiar with the challenges she was facing.
The American Association of Neurological Surgeons (AANS) describes tethered cord syndrome as a condition arising when tissue attachments within the spinal cord limit its movement, abnormally stretching it, and is closely associated with ’spina bifida‘, a birth defect where a child’s spinal cord fails to develop properly.
Dr. Holly Gilmer, a pediatric neurosurgeon at the Michigan Head and Spine Institute, who is involved in Alex’s treatment, explained the syndrome further: „In cases of tethered cord, the spinal cord is stuck to something. There could be a tumor in the spinal canal, a thickening on the bone protrusion, or too much fat at the end of the cord. This abnormality can’t remain and will eventually cause discomfort in some part of the body.“
According to the US Centers for Disease Control and Prevention (CDCP), while many children with spina bifida have a noticeable opening on their back, Alex’s condition was of the closed type, known as spina bifida occulta, and remained ‚hidden‘, complicating the diagnosis.
So, how did ChatGPT manage to correctly diagnose Alex’s condition when numerous specialists could not? Harvard University Epidemiology Professor Andrew Beam explains that ChatGPT operates by predicting the „next word in a sentence or sequence of words“ based on existing online text data. He emphasized that ChatGPT, which scans inputted data, is essentially an AI program designed to produce text similar to what it has been fed, likening it to a „supercharged medical search engine.“
So, how did ChatGPT manage to correctly diagnose Alex’s condition when numerous specialists could not? Harvard University Epidemiology Professor Andrew Beam explains that ChatGPT operates by predicting the „next word in a sentence or sequence of words“ based on existing online text data. He emphasized that ChatGPT, which scans inputted data, is essentially an AI program designed to produce text similar to what it has been fed, likening it to a „supercharged medical search engine.“
„This application’s premium version delivers even better results. We could even say it performs better than Google in certain aspects,“ Professor Beam added, suggesting that ChatGPT could be incredibly beneficial for patients with complex, hard-to-diagnose conditions.
This story isn’t just about a family finding relief after years of searching; it’s a testament to the growing capabilities of AI in healthcare. It underscores the potential of technology like ChatGPT in supporting medical diagnostics, especially in cases where traditional methods falter. While AI doesn’t replace human doctors, it serves as a powerful tool in enhancing healthcare, providing hope for patients like Alex and countless others facing medical uncertainties.
https://www.today.com/health/mom-chatgpt-diagnosis-pain-rcna101843