Advanced deconvolution method for the estimation of motor nerve conduction block: test on patients and comparison with standard clinical methods

Conduction block (CB) is the failure of an action potential (AP) to propagate through an intact axon of a motoneuron. The routinely used methods in clinical practice for CB estimation are based on the comparison of the area and the amplitude of two compound muscle action potentials (CMAPs) elicited with transcutaneous electrical nerve stimulation at sites proximal and distal to the nerve segment in which CB is suspected. The CB estimation obtained with these standard clinical methods is altered by the phase cancellations produced by temporal dispersion of the MUAPs constituting the CMAP. The temporal dispersion is abnormal in subjects with neuropathies that cause important reduction of peripheral nerves conduction velocity by disrupting the myelin sheath (demyelinating diseases). The proposed method is based on deconvolution of CMAPs and provides the delay distributions which convolved with a representative waveform named kernel approximately reconstruct the CMAPs. The slow afterwave of muscle potentials was studied and included in the kernel model and the delay distribution integral was used to estimate the CB. The method was tested on experimental signals obtained from both healthy and pathological subjects. The pathological group was composed of subjects with Chronic Inflammatory Demyelinating Polyneuropathy (CIDP), with Multifocal Motor Neuropathy (MMN) and with uncertain diagnosis. The results show that the proposed method is stable to temporal dispersion, allowing to obtain CB estimation with a lower confidence threshold with respect to the area and amplitude methods.