Speaker at Gartner IT Symposium/Xpo™

Wednesday, October 18th

Sofya Drives Clinical Terminology with AI in SNOMED-CT

Sofya highlights the vital role of clinical terminologies in AI, presenting her clinical reasoner that brings cognitive technologies developed by health experts. Hospital Sírio-Libanês researchers will also share Sofya use cases.

Mayhara Nogueira


Sofya Presents Clinical Terminologies Driven by AI at SNOMED-CT The company combines cognitive technologies led by healthcare experts to enhance clinical reasoning

Sofya is taking a central role in the transformation of global healthcare, as it prepares to participate in the innovative SNOMED-CT event on October 25 in Atlanta, USA, discussing the impact of clinical terminologies on AI technologies.

Clinical terminologies are fundamental in understanding and interpreting health data, promoting accurate communication among healthcare systems, medical professionals, and AI technologies. However, the cost associated with the lack of generation and utilization of valuable information has been incalculable.

The growth rate in health data volume will be 36% by 2025 (International Medicine Journal), and according to IBM, a surprising statistic reveals that approximately 70% of data generated in the healthcare field lacks proper structuring. Equally concerning is the fact that recommendations provided by AI systems, especially in cancer treatment contexts, have an error rate of 30% (JAMA Oncol).

Without clear terms and structured data, the consequences are imminent, such as misunderstandings and medical errors that affect diagnoses and treatments. Ensuring safe and reliable decisions involves following complex rules and standards, which can be challenging but crucial for making the right choices for patients. These numbers demonstrate the urgency of adopting approaches where AI technologies and data structuring based on clinical terminologies work together, which is what Sofya is doing with a team of researchers consisting of PhDs in medical ontologies, computer science, and data science.

Sofya has developed the Protocol Editor and Clinical Reasoner, superintelligences programmed with clinical language by medical professionals and healthcare experts. The goal is to create an environment that can capture medical knowledge in natural language and transform it safely and in compliance with clinical decision support protocols into powerful tools that assist with everything from automating routine tasks to aiding in complex clinical decisions.

* > "Placing cognitive technologies under the leadership of healthcare experts to enhance clinical reasoning is a vital achievement in an industry where the number of new drugs and treatments has increased tenfold in the past decade, adding to the already heavy workload of healthcare teams," says Igor Couto, CEO of Sofya.

Sofya uses bioinformatics terminologies and standards such as SNOMED, ICD, LOINC, as well as the FHIR interoperability standard, to ensure not only the reliability of information but also plug-and-play connectivity.


"Developing AI that generates reliable responses is a monumental task. We face this challenge by applying security and usability, integrating clinical terminologies and biomedical language into a natural language clinical programming environment, in adherence to strict digital health and AI ethics standards," explains Couto.

  • Sofya empowers healthcare professionals to create efficient, accurate, and secure clinical protocols with AI, effectively addressing systemic inefficiencies and the lack of personalized care.


"It's easier to make the reasoning-building environment friendly for doctors than to blindly trust in a magic box or believe that programmers deeply understand medicine," he says. "With Sofya, we are not just envisioning a better healthcare system; we are actively equipping those with the qualities to build it, uniting technology with the best human experiences to elevate personalization, accuracy, and clinical excellence," Couto concludes.

Igor Couto will present the panel discussion: "SNOMED-CT IA Symposium: The Role of Terminologies in AI Technologies" on October 25. In the same month, medical experts from Hospital Sírio-Libanês will also present Sofya's use cases and scientific works during the exhibition: Dr. Beatriz de Faria Leão will lead the lecture "Construction on the immunization component of the Brazilian IPS: mapping of local vaccines terminologies to SNOMED-IPS," to be held at Room Station on October 26 at 11 a.m. Dr. Fabio Lario will present the lecture: "Clinical Recommendation of allergies and adverse reactions from SNOMED-CT sub-ontologies and natural language processing," which will take place in Room Hub 1 on October 26 at 3 p.m. SNOMED International, a renowned non-profit organization based in the United Kingdom, is leading the way in defining global health standards in terms, with the aim of promoting improvements in human health.