Research Article

Machine Learning Application of Transcranial Motor-Evoked Potential to Predict Positive Functional Outcomes of Patients

Table 1

The summary of studies that indicated that TcMEP can be used as a prognostic tool [6].

NumberReferenceNumber of samplesIONM modalities usedStimulation parametersMuscles used to monitor MEPImprovement criteriaResults

1Barley et al. [7]One (15-month-old boy)TcMEP and SSEPC1-C2 scalp electrode positioning, current stimulation (145 mA to 187 mA for the left extremities and 175 mA to 200 mA for the right extremities)Bilateral quadriceps femoris, tibialis anterior, gastrocnemius, sphincter, abductor pollicis brevis, and abductor hallucisNot mentionedTcMEP response of the left APB had an increment in amplitude. The patient had observable left upper extremity improvement
2Voulgaris et al. [8]25 (2 had no IONM results)TcMEP and EMGC1-C2 with multipulse current stimulation, 0 mA to 200 mA, stimulus duration 0.2 ms to 0.5 msNot mentioned>50% MEP amplitude improvement17 patients with >50% improvement had better VAS score improvement
3Rodrigues et al. [9]One (case report)SSEP, MEP, and free running EMGC3-C4 stimulationNot mentioned muscles’ names specifically but monitoring covered L3-S2 myotomesNot mentionedMEP improved as much as 30%, and patient had returned to sports
4Raynor et al. [10]386 patients had IOM signals improvement out of 12375 patients who had spinal surgeries over 25 yearsDNEP, TcMEP, spontaneous EMG, triggered EMG, and dermatomal SSEPC3-C4 TcMEP scalp electrode stimulation montageUpper extremity TcMEP was recorded from deltoid, flexor/extensor carpi radialis, and/or abductor digiti minimi/abductor pollicis brevis. Lower extremity TcMEP was recorded from anterior tibialis, medial gastrocnemius, and/or extensor hallucis longusNot mentionedThe results did not mention specifically TcMEP improvement, but out of the modalities used, 88.7% of patients had IOM signals improvement, but one patient out of this percentage had permanent neurological deficit
5Visser et al. [11]74 patientsTcMEPCz-Fz with monophasic stimulation and C3-C4 with biphasic stimulationFor the lower limbs, the quadriceps muscle (L2-L4), the tibialis anterior muscle (L4-L5), the hamstrings (L5-S1), or the gastrocnemius muscle (S1–S2). For cervical, the bilateral trapezoid muscle (C2–C4), the biceps (C5–C6), and triceps muscle (C7–C8) of the arm; the extensor muscles of the forearm (C6–C7); or the abductor digitus V muscle (C6–C8)>200% of amplitude incrementThere is a correlation between the duration of symptoms onset and the MEP improvement. MEP improvement can be accurate if the symptoms’ onset duration is less than half a year
6Wang et al. [12]59 patients who had cervical myelopathy who underwent laminoplasty or laminectomyMEP and SSEPNot mentionedNot mentionedNot mentionedPatients who had MEP signals improvement had a significant mJOA improvement rate. MEP amplitude was found to be a more accurate parameter compared to MEP latency in predicting surgery outcome
7Dhall et al. [13]32EMG, MEP, and SSEP (not used for the study)100 V–1000 V constant voltage stimulation, C1-C2 anodal stimulation, double train with a total of 9 pulses, 50 ms pulse width, 1.7 ms interstimulus, and 13.1 ms ISINot mentionedComparison with AIS grade and BASIC score of MRI imagesMEP outcome (present) highly correlated with better AIS grade and BASIC grade
8Piasecki et al. [2]18MEP and SSEP (not used for the study)50 V–150 V C1-C2 biphasic stimulation, 5 to 7 train pulses, 500 Hz, and 1 ms interstimulus pulseOne upper limb muscle (control), bilateral tibialis anterior/bilateral abductor hallucis>20% of AUC MEP; > 50% of ZCQ scoreThe MEP improvement was related to the early follow-up functional outcome
9Wi et al. (2019) [14]29 patients who had improvement in IONM signals out of 317 casesMEP and SSEPNot mentionedUpper extremity TcMEP was recorded from deltoid, triceps, and thenar muscles. Lower extremity TcMEP was recorded from anterior tibialis and abductor hallucesComparison with MISS, SF-36, JOA, NDI, and Oswestry Disability IndexThe patients with MEP improvement had a better MISS improvement rate, while the patients with SSEP improvement only had a better SF-36 improvement rate
10He et al. [15]One (case report)MEP and free running EMGNot mentionedBilateral iliopsoas, rectus femoris, tibialis anterior, and medial gastrocnemiusNot mentionedMEP improvement aligned with the patient’s relieved symptoms