E-norms and AI in clinical neurophysiology

Abstract

Objective

To describe the use of Artificial Intelligence (AI) to automate the e-norms method, a technique used to derive normative data from patient studies, mixed datasets that contain both normal and abnormal data. Multiple studies have shown that normal values collected with the e-norms method compare favorably with those collected from healthy volunteers using traditional methods.

Methods

OpenAI’s ChatGPT was used by the author to build a Python script to automate the e-norms method’s plateau identification, the area of the e-norms curve where a variable’s normal values lie. To date, e-norms plateau identification has been done visually using an Excel Macro developed for that purpose.

Results

E-norms normal values derived from the visual e-norms plateau identification with the Excel Macro compared favorably with those derived from the OpenAI’s Python script developed by the author to automate the e-norms plateau identification.

Conclusions

OpenAI’s ChatGPT Python scripts can be developed by users with little to no experience in programming to automate the collection of e-norms normal values.

Significance

A Neurologist with no experience in programming, and a limited knowledge of statistics was able to develop this on his own without the help of any programmers or statisticians.