Automatic muscle impedance and nerve analyzer (AMINA) as a novel approach for classifying bioimpedance signals in intraoperative pelvic neuromonitoring

Abstract

Frequent complications arising from low anterior resections include urinary and fecal incontinence, as well as sexual disorders, which are commonly associated with damage to the pelvic autonomic nerves during surgery. To assist the surgeon in preserving pelvic autonomic nerves, a novel approach for intraoperative pelvic neuromonitoring was investigated that is based on impedance measurements of the innervated organs. The objective of this work was to develop an algorithm called AMINA to classify the bioimpedance signals, with the goal of facilitating signal interpretation for the surgeon. Thirty patients included in a clinical investigation underwent nerve-preserving robotic rectal surgery using intraoperative pelvic neuromonitoring. Contraction of the urinary bladder and/or rectum, triggered by direct stimulation of the innervating nerves, resulted in a change in tissue impedance signal, allowing the nerves to be identified and preserved. Impedance signal characteristics in the time domain and the time-frequency domain were calculated and classified to develop the AMINA. Stimulation-induced positive impedance changes were statistically significantly different from negative stimulation responses by the percent amplitude of impedance change A_max in the time domain. Positive impedance changes and artifacts were distinguished by classifying wavelet scales resulting from peak detection in the continuous wavelet transform scalogram, which allowed implementation of a decision tree underlying the AMINA. The sensitivity of the software-based signal evaluation by the AMINA was 96.3%, whereas its specificity was 91.2%. This approach streamlines and automates the interpretation of impedance signals during intraoperative pelvic neuromonitoring.

Figure 1

Pelvic neuromonitoring principle based on direct nerve stimulation with a hand probe and bioimpedance measurement of the bladder and rectum with a bipolar electrode montage. Contraction of the bladder and rectum smooth muscle in response to direct stimulation of the innervating autonomic nerves results in a change in the muscle tissue impedance compared to the impedance level before contraction. Assessment of stimulation-induced tissue impedance changes allows the nerves to be identified and preserved by the surgeon.

Figure 2

Signal examples of pre-classified signals of: (1) a positive stimulation-induced characteristic impedance change of the rectum (upper left, a), from case 7 of the clinical investigation/patient LB-09) and the urinary bladder (upper right, b), from case 1 of the clinical investigation/patient LB-01). The voltage drop (U(t)–U(0))/U_a across the smooth muscle changes during stimulation of autonomic nerves in the surgical field until the maximum impedance change is reached. This is followed by a relaxation phase of the muscle with the decrease of the voltage drop to the initial level. U(t) is the voltage drop as a function of time, U(0) is the voltage drop at time t = 0s, and U_a is the impedance level at baseline (mean value of the first 15 samples acquired within 1.5 s). Signal examples of pre-classified signals of: (2) a negative stimulation response (lower left, c.), from case 26 of the clinical investigation/patient LB-29; and (3) a non-specific artifact (lower right, d), from case 2 of the clinical investigation/patient LB-02).

Figure 3

Flowchart of the algorithm design process. Segmented, normalized, and digitally filtered signal sections were used as input time series for subsequent offline signal analysis in time and time–frequency domain. Classification of the extracted signal features was used to design a decision tree underlying the Automatic muscle impedance and nerve analyzer (AMINA).

Figure 4

Flowchart Part I of the Automatic muscle impedance and nerve analyzer (AMINA). Pre-processed samples of the impedance signal are collected in a signal buffer during nerve stimulation. Signal analysis in time domain is performed for signal sections with > 50 samples in length. Discrimination between (2) negative stimulation responses and (1) positive or (3) non-specific responses is made based on an amplitude threshold A_max of 0.9%.

Figure 5

Flowchart Part II of the Automatic muscle impedance and nerve analyzer (AMINA). Signal sections, which are assessed as (1) positive or (3) non-specific stimulation responses are forwarded to signal analysis in the time–frequency domain using discretized CWT. The distinction between the two signal classes is based on the wavelet scale associated with the magnitude maximum of the transformation coefficients of the CWT.

Figure 6

Boxplot of amplitude Amax (a), gradient m (b) and onset latency t₀ (c) for comparison of (1) positive impedance changes, (2) negative stimulation responses, and (3) artifacts (n = 131 per signal class from case 1 to 15 of the clinical investigation/patients LB-01–LB-16).

Figure 7

Signal sections with artifacts in the time domain (top left, a and c), and corresponding CWT scalograms (top right, b and d), signal examples from case 3 of the clinical investigation/patient LB-04), and signal sections with positive stimulation responses in the time domain (bottom left, e and g) and corresponding CWT scalograms (bottom right, f and h), signal examples from case 1 and case 15 of the clinical investigation/patient LB-01 and LB-16). The examples show clear differences in the wavelet scale at maximum magnitudes of the CWT transformation coefficients.

Figure 8

Boxplots of wavelet scales at maximum magnitudes of transformation coefficients of CWT of (1) positive characteristic impedance changes and (3) artifacts; n = 131 per signal class from cases 1 to 15 of clinical investigation/patient LB-01–LB-16).