Estimation of Noise Standard Deviation Using an Otsu-Based Approach in Extracellular Neural Recordings Hucre Disi Sinirsel Kayitlarda Otsu Temelli Bir Yaklasim Kullanilarak Gurultu Standart Sapmasinin Kestirimi


Kabakci K. A., Töreyin B. U., Okatan M.

28th Signal Processing and Communications Applications Conference, SIU 2020, Gaziantep, Turkey, 5 - 07 October 2020 identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/siu49456.2020.9302243
  • City: Gaziantep
  • Country: Turkey
  • Keywords: computational neuroscience, extracellular neural recording, noise estimation, Otsu thresholding, scale estimation, signal detection
  • Istanbul Technical University Affiliated: Yes

Abstract

In extracellular neural recordings, the actual signal is separated into noise and action potentials using some thresholding methods. Generally, the threshold is determined as 3 to 5 times the estimated standard deviation of the noise in the filtered recordings. However, the value of the standard deviation estimate in all of these methods depends on the spike density (firing rate) of the signal. In this study both the dependence on firing rate has been eliminated and the ratio of the estimated to true value of the standard deviation has been shown to converge to 1 using a method that is based on Otsu's thresholding method, which has been used as a thresholding method for decades in the field of image processing. This method gives better results in terms of frequency dependence and ratio to actual value of standard deviation than truncation thresholds method, which is the current best method. These results are important for the development of an appropriate method for accurate estimation of noise and amplitude thresholding in brain-machine-interfaces.