Related Experiment Video
Updated: Aug 13, 2025
12:18
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
7.6K
A Machine Learning-Based Severity Prediction Tool for the Michigan Neuropathy Screening Instrument
Fahmida Haque1,2, Mamun B I Reaz1,3, Muhammad E H Chowdhury4
1Centre of Advanced Electronic and Communication Engineering, Department of Electrical, Electronic and Systems Engineering, Universiti Kebangsaan Malaysia, Bangi 43600, Malaysia.
Diagnostics (Basel, Switzerland)
|January 21, 2023
Summary
A new grading system for diabetic sensorimotor polyneuropathy (DSPN) severity was developed using the Michigan neuropathy screening instrument (MNSI). This tool aids in predicting DSPN prognosis and classifying severity levels for better patient management.
Area of Science:
- Neurology
- Diabetology
- Medical Informatics
Background:
- Diabetic sensorimotor polyneuropathy (DSPN) is a major diabetes complication, potentially leading to severe outcomes like amputation.
- The Michigan neuropathy screening instrument (MNSI) is a common screening tool but lacks a standardized severity rating.
- Accurate DSPN severity assessment is crucial for effective patient management and prognosis.
Purpose of the Study:
- To develop and validate a DSPN severity grading system using the MNSI.
- To identify key MNSI features predictive of DSPN severity through machine learning.
- To create a nomogram for predicting DSPN probability and classifying severity levels.
Main Methods:
- Utilized 19-year longitudinal data from the Epidemiology of Diabetes Interventions and Complications (EDIC) trial.
- Employed machine learning (extra tree model) to identify significant MNSI features.
- Developed and validated a nomogram based on multivariable logistic regression for DSPN probability prediction.
Main Results:
- Identified seven key MNSI features: vibration perception (R/L), 10-gm filament, previous neuropathy, callus, deformities, and fissure.
- Achieved high diagnostic accuracy with AUCs of 0.9421 (internal) and 0.946 (external) for the nomogram.
- Established four DSPN severity levels (absent, mild, moderate, severe) based on predicted probability thresholds.
Conclusions:
- The developed MNSI-based grading system provides a straightforward, reproducible method for assessing DSPN severity.
- This tool can aid clinicians in determining patient prognosis and tailoring treatment strategies.
- The validated nomogram offers a reliable approach to predict DSPN risk and guide clinical decision-making.
