Time-frequency visualization of alcohol withdrawal tremors

Area of Science:

• Biomedical Engineering
• Neurology
• Signal Processing

Background:

• Alcohol withdrawal (AW) syndrome can cause tremors, impacting patient assessment.
• Current methods for assessing tremor severity in AW are subjective.
• Differentiating real tremors from factitious ones is clinically important, especially in emergency settings.

Purpose of the Study:

• To propose a signal processing method for assessing alcohol withdrawal (AW) tremor severity.
• To develop an iOS application for objective tremor assessment using accelerometers.
• To evaluate the accuracy of this method compared to expert clinicians and its potential in differentiating real from factitious tremors.

Main Methods:

• Developed an iOS application to capture iPod movements via accelerometer.
• Calculated the Clinical Institute Withdrawal Assessment (CIWA) score, incorporating accelerometer data for tremor severity.
• Analyzed 84 recordings from 61 subjects (49 patients, 12 nurses) to characterize AW tremor and assess accuracy.

Main Results:

• Found a linear relationship between accelerometer-measured energy (4.4-10 Hz) and expert tremor severity ratings.
• Demonstrated that 75% of actual AW tremors had a mean peak frequency >7 Hz, compared to only 17% of factitious tremors.
• Suggests a potential discriminator for real versus factitious tremors based on frequency.

Conclusions:

• An iOS application using accelerometers can reliably estimate AW tremor severity.
• Signal processing of accelerometer data shows promise in differentiating real AW tremors from factitious ones.
• This technology could enhance objective assessment of AW syndrome and improve clinical decision-making.