Arrow
use cases

Incease diagnosis and Treatment of Rare Diseases with AI

Diagnosing rare diseases is difficult and time-consuming, and there is often no cure. AI improves diagnosis and treatment for patients.

Problem

The realm of rare diseases encompasses a spectrum of around 7,000 distinct conditions, each affecting fewer than 200,000 individuals in the United States and presenting a diverse array of symptoms, from benign to severe life-threatening manifestations (1). These conditions, often rooted in genetic mutations or environmental factors, present significant diagnostic challenges due to their heterogeneity and rarity. Compounding the issue, the journey to a diagnosis is typically prolonged—averaging seven years—and less than 5% of these diseases have an FDA-approved treatment, underscoring the persistent unmet medical needs within this patient population (1)(4).

Globally, rare diseases impact 1 in 10 people, translating into an estimated 475 million affected individuals, with a notable impact on pediatric health, as 30% of children with a rare disease do not survive past their fifth birthday (2). In pediatric care facilities, a third of hospital beds are occupied by children fighting rare diseases (3). These statistics not only highlight the prevalence of rare diseases but also the critical need for continued research, treatment development, and advocacy to better the lives of those affected by these uncommon disorders (5)(6).

‍

Why it matters

  • Rare diseases encompass about 7,000 conditions, each affecting fewer than 200,000 individuals in the U.S., with diverse and often severe symptoms.
  • Diagnosis is typically prolonged, averaging seven years, and less than 5% of rare diseases have FDA-approved treatments.
  • Globally, rare diseases affect 1 in 10 people (475 million), with significant pediatric impact, including 30% of affected children not surviving past their fifth birthday and a third of pediatric hospital beds occupied by these patients.

Solution

A predictive analytics model has been designed to simulate clinical scenarios for rare diseases, enabling faster and more accurate diagnoses. This model is refined using a synthetic dataset that includes key variables such as age, gender, genetic markers, clinical symptoms, environmental exposures, family history, prior diseases, and lifestyle factors.

Discover more and interact with our AI!

Datasources

The model design is based on what was reported in the studies by Khoury et al. (5) and GĂłmez-Cabezas et al. (6), which detail the importance of the chosen variables in the diagnosis and treatment of rare diseases. These studies ensure that the synthetic data mimics the real-world scenarios that doctors face when treating rare diseases, providing a reliable foundation for the model's predictive capabilities.

Citations

  1. National Organization for Rare Disorders. (n.d.). National Organization for Rare Disorders. Retrieved May 16, 2023, from https://rarediseases.org/2.
  2. World Economic Forum. (2023). Global data access for solving rare disease: A health economics value framework. Retrieved from https://www.weforum.org/reports/global-access-for-solving-rare-disease-a-health-economics-value-framework/3.
  3. The New York Times. (2022, November 1). US Children's Hospitals Are Overwhelmed by RSV. Retrieved from https://www.nytimes.com/2022/11/01/science/rsv-children-hospitals.html4.
  4. U.S. Food and Drug Administration. (2023, May 19). Rare Diseases at FDA. Retrieved from https://www.fda.gov/patients/rare-diseases-fda5.
  5. Khoury, M. J., et al. (2020). Artificial intelligence and precision medicine for rare diseases. Nature Medicine, 26(11), 1679-1686. doi:10.1038/s41591-020-0994-06.
  6. GĂłmez-Cabezas, M. A., et al. (2021). Challenges and opportunities of artificial intelligence for rare diseases. Frontiers in Medicine, 8, 669561. doi:10.3389/fmed.2021.669561

Book a Free Consultation

Trusted by the world's top healthcare institutions