Single-trial classification of motor imagery differing in task complexity: a functional near-infrared spectroscopy study
2011

Classifying Motor Imagery Tasks Using fNIRS

Sample size: 12 publication 10 minutes Evidence: moderate

Author Information

Author(s): Lisa Holper, Martin Wolf

Primary Institution: University Hospital Zurich

Hypothesis

Can we classify brain signals from motor imagery tasks of varying complexity using functional near-infrared spectroscopy?

Conclusion

The study found that it is possible to classify motor imagery tasks of different complexities with an average accuracy of 81%, although there is significant variability among subjects.

Supporting Evidence

  • Classification accuracy averaged 81% across subjects.
  • Significant differences in brain activity were observed between simple and complex tasks.
  • Subject-to-subject variability was noted in the classification combinations.

Takeaway

The researchers used a special device to read brain signals while people imagined moving their fingers in simple and complex ways, and they found they could tell the difference most of the time.

Methodology

12 subjects imagined simple and complex finger-tapping tasks while their brain activity was recorded using fNIRS, and the data were analyzed using Fisher's linear discriminant analysis.

Potential Biases

Variability in individual brain signal responses may introduce bias in classification accuracy.

Limitations

The classification accuracy was subject to considerable variability between individuals, and the study had a limited number of trials per subject.

Participant Demographics

12 healthy subjects (6 males, mean age 29 years, range 26-33 years), all right-handed.

Statistical Information

P-Value

p ≤ 0.001

Confidence Interval

CI 95%

Statistical Significance

p ≤ 0.001

Digital Object Identifier (DOI)

10.1186/1743-0003-8-34

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication