An Improved Detrended Moving-Average Method for Monitoring the Depth of Anesthesia

The detrended moving-average (DMA) method is a new approach to quantify correlation properties in nonstationary signals with underlying trends. This paper monitored the depth of anesthesia (DoA) using modified DMA (MDMA) method. MDMA provides a power-law relation between the fluctuation function F
_MDMA
(s) and the scale s: F
_MDMA
( s)αs
^α
, where α is the slope of F
_MDMA
( s) in the logarithm scale. We applied the MDMA to monitor the DoA by computing the scaling exponent Fα and F
_min
values. To validate the proposed method, we compared our results with the bispectral index (BIS) monitor. We found a close correlation between our results and BIS with r( F
_min
) = 0.9346, r
²
(F
_min
) = 0.9183, and r(F
_α
) = 0.9458, r
²
(F
_α
) = 0.8855. Our method reflects the state of consciousness of a patient undergoing general anesthesia faster than BIS as observed clinically. The minimum time delay between the BIS and F
_min
trends was 12 s and the maximum was 178 s. Furthermore, in the case of poor signal quality, our results agreed with clinical observation, which indicates that our method can accurately estimate a patient's hypnotic state in such circumstances. Fα and F
_min
trends are responsive and their movement seems similar to changes in the clinical state of the patients.