Jan Sevcik - Author

Jan Sevcik

Contributor, The Yuan

Jan Sevcik is the co-founder and chief executive of Medical Search Technologies which develops AI software for healthcare industry. He has extensive knowledge of the content of electronic health record data, especially unstructured text, and over 20 years’ experience developing algorithms which he uses to develop both clinical and administrative applications. He is an avid reader of white papers and research on varied topics, including graph data theory, biology, and medicine.

Cognition

Classification Hones AI’s Effectiveness in Healthcare

When applying artificial intelligence and machine learning to healthcare, classification is an important way of identifying and labeling datasets to achieve more accurate diagnoses, and better treatments and outcomes for patients. This is especially true when data is missing or unknown.

Jan Sevcik | May 24, 2022

New era

NLP Falls Short in Healthcare Advancement

Debate swirls about how natural language processing can help advance artificial intelligence in the healthcare sector, where most data are unstructured. Jan Sevcik takes a closer look at NLP, the analysis of unstructured text, and how to understand its strategy.

Jan Sevcik | Apr 05, 2022

Emerging markets

AI in Healthcare, Tangible Benefits and Return on Investment

Where to begin for those delving into artificial intelligence in healthcare for the first time may be a bit of a conundrum, but Jan Sevcik shows organizations a path that provides tangible benefits and creates momentum.

Jan Sevcik | Jan 21, 2022

Optimization

Dataset’s Critical Role in Creating Correct Predictive Models

The digitization of electronic medical records, new machine learning techniques, and more robust hardware are creating opportunities to solve many unanswered questions in healthcare, but using the correct dataset is critical to creating predictive models which draw the correct conclusions. Even the most advanced AI techniques will yield poor results if the wrong dataset is utilized. Jan Sevcik discusses the variables’ conundrum and the problems scientists face.

Jan Sevcik | Oct 21, 2021

Optimization

Lessons Learned from Sarbanes Oxley IT Controls Can Improve AI Deployment Compliance

The process of developing and deploying artificial intelligence models is a complex one that needs closer inspection. Jan Sevcik considers the best practices, compliance, and risk mitigation, and suggests ways to implement and improve the AI model development and deployment processes.

Jan Sevcik | Oct 04, 2021

Domain knowledge

Reading RWD from EHR

Electronic health records (EHRs) contain a diverse array of structured and unstructured data which can be used in research requiring real world data, but most data are unstructured. Some is pre-extracted, meaning select EHR data are captured and entered into repositories like disease or product registers. Data of this kind can have limitations, being only a subset of all that are available, and may not contain all variables needed to yield correct conclusions.

Jan Sevcik | Sep 07, 2021