More than 7 thousand “unusual” conditions have been detected.
Highlights:
- Rare diseases are common, affecting up to 446 million people globally, with patients facing high costs, long diagnosis times, and difficulty finding specialists.
- A diagnosis offers emotional relief, personalized care, improved outcomes, access to support groups, participation in research, and family planning advice.
- AI has the potential to enhance the understanding, diagnosis, and management of rare diseases by utilizing comprehensive patient data, improving diagnostic accuracy, reducing diagnosis time, identifying candidate genes, and aiding in drug discovery and biomarker identification.
- Addressing data underrepresentation, regulatory, legal, public policy, methodological, and technological issues is essential for effective AI implementation in diagnosing and treating rare diseases.
Rare diseases are, in reality, not so “rare.” There are more than seven thousand of them, and they can affect up to 446 million people worldwide, according to data published in the European Journal of Human Genetics. Those living with one of these conditions face multiple challenges, from high treatment costs to extremely long processes to find a diagnosis, which often turns out to be incorrect. Those fortunate enough to receive a correct diagnosis frequently need help finding medical specialists knowledgeable about their condition. Many patients even have to immigrate to receive care. However, the reality is that the vast majority do not have access to such care and go through the process in isolation and without support.
Rare diseases are, in reality, not so “rare.”
For the past few weeks, I have been reflecting on why it is so important to have a diagnosis. Even though there is still no treatment for many of these diseases, knowing the diagnosis allows people to understand their condition and know what to expect. It provides them emotional relief. Additionally, it enables them to receive personalized medical care, improve their outcomes with prevention, access support groups, participate in research programs for future treatments, and obtain advice for family planning. “Well, now I have a label. I have a Facebook group I can belong to,” explains a patient living with one of these rare diseases.
Well, now I have a label. I have a Facebook group I can belong to.
While artificial intelligence does not currently solve the problem of access to healthcare, it has the potential to improve diagnosis and research for treatment development. This technology mimics human cognitive processes and is no longer just a simple promise or hype. Using each patient’s information, such as their demographic data, history, clinical test results, genetic data, clinical images, and data from health records and mobile devices, helps improve our understanding of managing these diseases. Several studies have shown advances in using artificial intelligence to improve diagnostic accuracy. Considering that genetic alterations cause between 70% and 80% of these disorders, systems can help identify the disease, prioritize candidate genes, and offer a differential diagnosis based on each patient’s symptoms. These innovations improve the accuracy of diagnosing rare diseases and reduce the time between the onset of the first symptoms and diagnosis. Additionally, artificial intelligence is being used to discover new drugs and identify biomarkers and genes, opening more opportunities in clinical research.
Despite these advances, several challenges must be addressed. It is necessary, for example, to address the characteristics of the data. Underrepresentation of the population in research studies generates biases. In this regard, most participants in genetic studies are of European descent, which limits diversity and, therefore, the understanding of diseases across all populations. Furthermore, human, regulatory, legal, public policy and methodological and technological aspects must be considered. Finally, more specific regulation and quality control must be implemented in the use of artificial intelligence to diagnose and treat these types of diseases.
