Recent advancements in artificial intelligence (AI) are enabling some patients to receive long-awaited diagnoses after years of grappling with unexplained illnesses, as highlighted in a new study.
Researchers from OpenAI in collaboration with Boston Children's Hospital have harnessed existing genetic information from 18 pediatric patients—many of whom are now adults—utilizing a novel AI model to resolve cases that had perplexed healthcare professionals for years. The hope is that this innovative model will assist countless children across the United States, as statistics show that one in ten Americans—over 30 million individuals, including many children—suffer from rare diseases.
The goal of the researchers is not only to diagnose some of the most elusive conditions but also to expedite the diagnosis process overall. The findings were published in the latest issue of the New England Journal of Medicine.
Among the patients, 28-year-old Kyra obtained a diagnosis of Myofibrillar Myopathy (MFM) after nearly 20 years of uncertainty. MFM encompasses a group of genetic disorders that lead to progressive muscle weakness. Reflecting on her experience, Kyra shared her disbelief at finally receiving an answer: "It felt very surreal at the time because I just didn’t expect to get an answer in my lifetime." Although the disease remains incurable, she expressed gratitude for gaining clarity about her condition.
While the AI model proposed potential explanations, the final diagnoses were always confirmed by medical professionals. This distinction is crucial, as the researchers emphasized that AI is not intended to replace physicians or genetic specialists. Instead, it serves as a supplementary resource, enabling healthcare providers to analyze complex data more efficiently, often within six to ten minutes per case.
Lead researcher Catherine Brownstein, a research associate in genetics at Boston Children's Hospital, noted that leveraging AI technology allows medical professionals to focus their time on critical tasks, like in-depth data analysis, rather than exploring numerous potential diagnoses without direction.
Kyra is optimistic about AI's potential to aid patients like herself but stresses the importance of careful oversight and ethical considerations in its use. "I think for assisting researchers in such complicated situations, AI can be a very useful tool," she remarked, adding that caution is necessary.
Privacy and ethical safeguards are foundational to this technology, as reiterated by Brownstein, who stated, "We're not removing any human guardrails here. A human has to review everything that the AI does."
This study also sheds light on the value of revisiting previous genetic test results. Given the rapid advancements in genetics, outcomes that were once deemed inconclusive may now be clarified with new knowledge and methodologies. Brownstein explained, "A genetic test that's negative today might not be viewed the same way in the future."
She acknowledged the challenge of clinicians keeping up with these swift developments but underscored AI’s effectiveness in navigating through genetic complexities. "The genome is being decoded more every day," she said, illustrating AI's ability to process this expanding information.
Notably, only 60 cases of MFM have been documented in medical literature, according to the non-profit organization CureMFM13, which focuses on tackling the issues surrounding the condition.
However, the study has its limitations, primarily as it comprised a review of existing cases, which doesn’t conclusively demonstrate the tool’s efficacy in real-time applications. Additionally, the number of diagnoses made was limited, and the research did not evaluate whether the AI tool resulted in time savings, reduced costs, or improved patient care.
The authors emphasize the need for larger, forward-looking studies across multiple medical institutions as the next step in evaluating this approach.
For Kyra, while the diagnosis does not erase her years of uncertainty, it has provided her and her family with newfound clarity—a name for her condition and a sense of connection with others facing similar challenges. She highlighted the irreplaceable value of the human aspect of medicine, stating, "When it comes to health matters that really change your life, you kind of want that human touch present. You want to feel like people care, and that they listen to you, and you’re not just a condition."
