Visual Evoked Potential Monitoring in Long-term Anesthesia Maintenance and Challenging Environments for Cerebral Arteriovenous Malformation Treatment: A Technical Case Report

Abstract

The monitoring of intraoperative optic nerve protection using visual evoked potential has been increasingly used in neurosurgical procedures. Although visual evoked potential is a valuable tool, its application is often hindered by anesthetic limitations and challenges related to measurement and recording environments. This case study indicates the successful use of visual evoked potential monitoring during long-term anesthesia maintenance and in artifact-prone environments for the treatment of a ruptured cerebral arteriovenous malformation. We report the case of a woman in her 30s who underwent surgical treatment for a ruptured cerebral arteriovenous malformation in the right occipital lobe, adjacent to the optic radiation. Visual evoked potential monitoring was essential during both transarterial embolization and subsequent microsurgical resection. During transarterial embolization, the implementation of improved measurement environments, localized light stimulation, and optimized recording conditions facilitated stable visual evoked potential monitoring despite the high levels of environmental noise in the angiography suite. Post-embolization, deep sedation was required to mitigate the risk of postoperative bleeding, causing an 18-hr duration of anesthesia before microsurgical resection. Despite the prolonged anesthesia, visual evoked potential monitoring was successfully maintained by carefully managing anesthesia depth. This case shows that visual evoked potential monitoring can be reliable and reproducible during extended periods of anesthesia and in challenging, artifact-rich environments. These findings underscore the feasibility of using visual evoked potential in complex neurosurgical cases, even under less-than-ideal conditions.

Introduction

Intraoperative neurophysiology monitoring (IONM) is used to detect nerve injuries resulting from surgical invasion and to preserve neurological function in the postoperative period.¹⁾ In neurosurgery, IONM is indicated and performed on the basis of the surgical site and procedure.²⁾ Visual evoked potential (VEP) is a form of IONM that assesses visual function and is mainly used in surgical procedures near the occipital lobe and in transnasal endoscopic surgeries for brain tumors.^3,4) Intraoperative VEP monitoring is highly sensitive to anesthetics and is only possible with total intravenous anesthesia using propofol.^5,6) Neurosurgical procedures often require long-term anesthesia maintenance, exceeding 8 to 10 hr, making anesthetic tolerance of IONM an important issue. However, there are no reports of VEP monitoring during extended anesthesia maintenance. Furthermore, performing reliable VEP monitoring under adverse conditions, such as outside the operating room, remains an area with unresolved challenges. This report describes a case in which reproducible VEP monitoring was achieved during long-term anesthesia maintenance and successfully conducted in an angiography room without noise control.

Case Presentation

A woman in her 30s was incidentally diagnosed with an unruptured cerebral arteriovenous malformation (AVM) in the right occipital lobe, classified as Spetzler-Martin grade 3. During follow-up, the AVM ruptured, leading to intraventricular hemorrhage (Fig. 1A). Transarterial embolization (TAE) and microsurgical resection of the AVM were planned with computed tomography angiography⁷⁾ and magnetic resonance tractography. The feeder artery was the P2 segment of the right posterior cerebral artery, which had 3 branches (Fig. 1B). One branch was the main feeder of the AVM, whereas another contributed to normal cerebral perfusion. The AVM was located near the optic radiation (Fig. 1C). We determined that VEP monitoring would be necessary during both TAE and microsurgical resection.

Fig. 1

Computed tomography (CT) of the head revealed intraventricular hemorrhage and hydrocephalus (A). CT angiography of the head showed the right posterior cerebral artery with 3 branches and a cerebral arteriovenous malformation (AVM) (B). Magnetic resonance tractography showed the optic radiation near the cerebral AVM and hemorrhage (C). Cerebral angiography revealed the disappearance of the nidus and no occluded vessels aside from the accessed vessel (D, E). During surgery, the nidus was completely removed (F, G).

During TAE, general anesthesia was induced with propofol in the angiography room. Because the sedation level could not be monitored with a bispectral index sensor (BIS), the effect-site concentration (ESC) of propofol was set at 3.8 to 4.0 μg/mL using a target-controlled infusion system developed based on a compartment model from previous studies.^8,9) The infusion rate was adjusted to 7.8 to 9.5 mg/kg/hr (Fig. 2), which was higher than the usual sedation level. Feeder trunk embolization was performed with the liquid embolic agent. Postoperative cerebral angiography confirmed the disappearance of the nidus without occlusion of non-feeder blood vessels (Fig. 1D and E). To minimize the risk of postoperative bleeding, the patient was deeply sedated in the intensive care unit (ICU) until the microsurgical resection the next day. Propofol was initially administered at an infusion rate of 4.7 mg/kg/hr, and dexmedetomidine was also added. The propofol infusion rate was then adjusted between 1.2 and 3.5 mg/kg/hr (Fig. 2). The total duration of anesthesia before microsurgical resection was 18 hr.

Fig. 2

The amplitude and latency of visual evoked potentials (VEPs) during transarterial embolization and microsurgical resection are presented as changes over time. The effect-site concentrations (μg/mL) and infusion rates (mg/kg/hr) of propofol and dexmedetomidine for each period are shown correspondingly in the graph. Amplitude and latency are plotted as percentages of the baseline in each period, and the VEP test results for light stimulation were recorded with Oz as the recording electrode. An alert point, defined as a 50% amplitude decrease from the baseline, is indicated by a red dotted line. In all periods, the amplitude did not continuously decrease below this point.

During surgery, the ESC of propofol was set at 2.2 to 3.0 μg/mL, with the infusion rate adjusted to 3.6 to 6.8 mg/kg/hr to maintain a BIS value of 40 to 60 (Fig. 2). A right temporal craniotomy and mastoidectomy were performed. A subtemporal approach was used to visualize the trifurcation of the P2 segment (Fig. 1F), and the feeder was clipped. The right temporal lobe was retracted, and a corticotomy was performed from the medial side of the temporal lobe. Tiny feeding vessels were coagulated and cut. The nidus was completely removed (Fig. 1G). The total anesthesia time from TAE to microsurgical resection was 30 hr. Postoperatively, the patient exhibited no subjective visual field defects. However, a Goldmann perimetry measurement conducted 3 weeks after treatment revealed a very slight left homonymous hemianopsia, which had not been present before surgery (Fig. 3A and B). Six months later, the patient showed slight improvement in the homonymous hemianopsia on examination and reported no subjective visual field defects.

Fig. 3

A Goldmann perimetry measurement was conducted 2 weeks before (A) and 3 weeks after (B) surgery. Postoperative examination revealed a very slight left homonymous hemianopsia, which was not present preoperatively.

Assessment and measurement of VEP

A light-emitting diode (LED) flash stimulation device (LFS-101 III; Unique Medical Co., Ltd., Tokyo, Japan) and a recording device (Neuromaster MEE-2000; Nihon Kohden, Co., Ltd., Tokyo, Japan) were used. The power supply was independently secured using an isolation unit. LED pads were fixed to both eyelids and covered with black shield pads. The recording electrodes were implanted at Oz, O1, and O2, following the International 10-20 System. The reference was A1+A2, using bilateral earlobes as A1 and A2. VEP stimulation and recording conditions included a light stimulus illuminance of 500 to 20,000 Lx, duration of 20 ms, frequency of 1.0 Hz, addition average of 200, band-pass filter of 10 to 500 Hz, and analysis time of 200 ms. The VEP was N1 and N2 for the negative wave near 75 ms and 145 ms, and P1 for the positive wave near 100 ms; N1-P1 and P1-N2 were evaluated by their respective peak-to-peak amplitudes.

Electroretinogram (ERG) recordings were performed simultaneously to verify that the light stimulus reached the optic nerve. Active electrodes were implanted in the bilateral external canthus, with A1+A2 as the reference. Other conditions were the same as for VEP. Preoperative VEP before TAE showed clear recordings at a stimulus intensity of 4000 Lx without interference from surrounding noise. During TAE, VEP waveforms did not change (Fig. 2). After deep sedation management, a new VEP baseline was obtained before microsurgical resection. VEP showed a slight decrease in amplitude and prolongation of latency compared with the TAE phase. During microsurgical resection, VEP exhibited a mild decrease in amplitude and prolongation of latency, but visual function was monitored effectively throughout the procedure.

Discussion

The course of this case emphasized 2 clinical points. First, VEP monitoring may be tolerant of long-term anesthesia maintenance with propofol. Second, detailed VEP monitoring in the angiography room is feasible by preparing the recording environment and deriving clear waveforms.

A comprehensive search of the literature was conducted using PubMed to identify studies in patients who underwent anesthetic management with propofol and intraoperative VEP monitoring. The results of this search are listed in Table 1.^3,4,10-20) There have been no studies addressing how long VEP monitoring can be performed reproducibly without significant anesthetic effects. Previous reports of VEP monitoring under general anesthesia suggest that the appropriate ESC of propofol ranges from 2.0 to 4.0 μg/mL,^{4,14,17,18,20)} with concentrations higher than 4.0 μg/mL associated with a more significant decrease in VEP amplitude than those of 2.0 to 3.0 μg/mL.¹⁴⁾ Other reports have also noted successful VEP recordings at propofol infusion rates of 3.9 to 12.0 mg/kg/hr.^{3,10-13,15,16,19)} In this case, after admission to the ICU, the ESC of propofol was maintained at approximately 3.0 μg/mL or less, with an intraoperative infusion rate ranging from 3.6 to 6.8 mg/kg/hr, aligning with previously reported values. These findings suggest that VEP monitoring remains feasible during prolonged anesthesia, provided the ESC and infusion rate are maintained within the established ranges. In contrast, VEP monitoring has been reported to produce a dose-dependent decrease in reproducibility and amplitude with anesthetics.⁶⁾ This reduction is attributed to the suppression of the polysynaptic pathway involving the lateral geniculate body from the retina to the visual cortical area. Unlike motor evoked potentials, which are highly susceptible to anesthetic effects owing to accumulation in alpha-motoneurons in the spinal cord, producing a reduction in amplitude owing to fading phenomena and temporal dispersion,^21,22) VEP pathways bypass the spinal cord. The effect of anesthetics on VEP may be deemed similar to that on somatosensory evoked potentials (SEPs) because both involve sensory nerve monitoring. However, although temporal dispersion often affects SEPs,²³⁾ its impact on VEP amplitude is believed to be negligible, given the shorter distance from the stimulation device to the occipital lobe. We believe that these anatomical distinctions reduce the likelihood of anesthetic accumulation affecting VEP signals.

Table 1

Previous Studies in Patients Who Underwent Anesthetic Management with Propofol and Intraoperative VEP Monitoring

Groups	Propofol		Amplitude		Latency
	ESC [μg/mL]	CI [mg/kg/hr]	N1-P1/P1-N2	Alert point	N1/P1	Alert point
CI: continuous infusion; ESC: effect site concentration; VEP: visual evoked potential
Sasaki et al., 2010 (3)	-	6.0-9.9	-/-	50%/-	-/-	-
Kamio et al., 2014 (11)	-	6.0-9.9	4.6±1.8	50%/50%	76.8±6.4	-
			5.7±2.8		98.0±8.6
Luo et al., 2015 (10)	-	3.9-8.1	2.8 [0.7-19.4] /2.2 [0.2-14.1]	50%/-	87.0 [51.0-142.0] /106.0 [65.0-161.0]	-
Uribe et al., 2017 (12)	-	6/12	-/-	50%/-	-/-	10%
Houlden et al., 2019 (13)	-	6.0-9.0	-/-	50%/-	-/-	-
Tanaka et al., 2020 (14)	2.0-4.0	-	7.5±3.6/-	-	-/-	-
Qiao et al., 2021 (15)	-	3.9-12.0	-/-	50%/-	-/-	-
Tao et al., 2021 (16)	-	3.9-12.0	-/-	-	-/101.5 [94.0-109.5]	8.61%
Ma et al., 2022 (17)	3.0-4.0	-	-/3.8±1.3	-	-/108.0±7.0	-
Nakagawa et al., 2022 (18)	2.3-3.0	-	-/-	50%/-	-	-
Mattogno et al., 2023 (4)	2.5-3.0	-	2.6±1.4	90%/-	-/-	-
Yamada et al., 2023 (20)	2.7-3.5	-	3.1±2.1/-	50%/-	74.7±12.2 88.9±11.5	-
Tao et al., 2024 (19)	-	3.9-12.0	4.61±2.22 4.78±2.27	50%/60%	-/-	-

Prolonged and high-dose administration of propofol is a known risk factor for propofol infusion syndrome (PRIS).²⁴⁾ However, to maintain the quality of VEP monitoring, we needed to avoid inhalation anesthesia and benzodiazepines, which interfere more significantly with VEP signals. Propofol is generally reported to have low drug accumulation.²⁵⁾ The patient had mild liver dysfunction, necessitating careful consideration of the propofol infusion rate. However, liver dysfunction itself is not a known risk factor for PRIS developing.^24,26,27) We determined that the benefits of IONM under propofol anesthesia outweighed the risks of PRIS, provided that strict monitoring—including electrocardiogram changes, blood gas analysis, and urine color assessment—was implemented. Moreover, we minimized the propofol dose by incorporating dexmedetomidine in the ICU, causing an average infusion rate of 3.5 mg/kg/hr over the total anesthesia period, including sedation management after microsurgical resection. There were no preoperative or postoperative abnormalities in vital signs or blood tests, including serum creatine kinase, and PRIS did not occur. Nevertheless, PRIS has been reported even with propofol administration at rates lower than 4 mg/kg/hr or durations shorter than 48 hr.^24,28-30) Therefore, rigorous monitoring and preparation for PRIS remain essential.

Within VEP monitoring, it is crucial to control ambient noise and reconsider the stimulus environment and measurement methods to minimize intra-individual differences and ensure reproducibility.^3,6) In this case, the initial procedure was endovascular treatment conducted in the angiography room, where various limitations and artifacts necessitated meticulous attention to detail in VEP recording. First, we selected Flash-VEP with black shield pads. Although goggle-type light stimulators are inherently light-shielded and easily fixed, they are not suitable for endovascular treatment because they obstruct the surgical field. In contrast, Flash-VEP often has insufficient light shielding. To address this, we used black shield pads to shade the area around the eyelids. This method, previously reported for transsphenoidal surgery, has shown that VEP monitoring can be performed without interfering with the surgical field.³¹⁾ Light entered only 1 eye, allowing complete isolation and derivation of VEPs for the left and right eyes. The light stimulus intensity was kept low³⁾ and constant, producing negligible changes in the ERG and rendering Flash-VEP quantitative.

Next, we secured the independent power supply with isolation unit equipment for each stimulator and recorder as a measure against environmental noise. Flash-VEP is vulnerable to environmental artifacts owing to its low microvolt signal, and high-voltage devices such as angiography equipment and floor-mounted mobile angiographic tables are likely sources of interference. We determined that reducing artifacts from power lines in the facility or surrounding equipment was necessary. Implementing isolation for each individual device helped prevent noise returning through the power line. This approach was both safe and convenient for the patient and effectively reduced artifacts and variability in VEP waveforms.

Furthermore, we also addressed noise reduction caused by the reference electrode. Fz and A1+A2 are commonly used as references in pattern-reversal VEP,³²⁾ but the Fz electrode may pick up ERG noise. We selected A1+A2 as the reference electrode, a concatenation of A1 and A2, which allowed the measurement of VEP waveforms with less ERG noise.

Another interesting point in this case was the potential for detecting homonymous hemianopsia. Functional protection through VEP monitoring using red light has traditionally been limited, making it difficult to predict slight or quadrant visual field deficits in the postoperative period.¹⁰⁾ VEP monitoring using white light, which includes all colors and is not yet commonly used in Japan, has been reported to allow the detection of quadrant blindness when the warning criterion is set to a 10% to 30% amplitude decrease.^4,32) In this case, the warning criterion for VEP was set to a 50% decrease in amplitude as in previous reports, and the amplitude never decreased below this threshold. However, slight homonymous hemianopsia developed in the patient postoperatively. Retrospective analysis of the VEP waveforms revealed a 20% to 30% decrease in amplitude while operating near the optic radiation. These subtle changes in VEP waveforms were believed to reflect the effects of surgery. The results suggest that a stricter warning criterion, such as a 25% amplitude decrease, might improve detection of homonymous hemianopsia, even when using red-light VEP monitoring.

In summary, we reported a case in which neurological function was effectively assessed by VEP monitoring under long-term anesthesia and an angiography room during the treatment of cerebral AVM. The findings suggest that VEP monitoring can be reproducible and reliable, even in surgeries involving long-term anesthesia maintenance and environments with significant artifacts.

Acknowledgments

We thank the staff of the Division of Clinical Engineering for their invaluable support in the treatment of this patient.

Author Contributions

Shoto Yamada and Ayumu Yamaoka contributed equally to this work.

Informed Consent

This report is a technical note based on case experience and adjustments of IONM equipment, and informed consent was obtained from the patient. Therefore, approval from the institutional review board of Sapporo Medical University was not required.

Disclaimer

Nobuhiro Mikuni is a member of the Editorial Board of this journal. However, he was not involved in the peer-review or decision-making process for this article.

Conflicts of Interest Disclosure

There are no conflicts of interest.

References

• 1)   Howick  J,  Cohen  BA,  McCulloch  P, et al. Foundations for evidence-based intraoperative neurophysiological monitoring. Clin Neurophysiol. 2016;127 (1):81-90. doi: 10.1016/j.clinph.2015.05.033
• 2)   You  H,  Qiao  H. Intraoperative neuromonitoring during resection of gliomas involving eloquent areas. Front Neurol. 2021;12:658680. doi: 10.3389/fneur.2021.658680
• 3)   Sasaki  T,  Itakura  T,  Suzuki  K, et al. Intraoperative monitoring of visual evoked potential: introduction of a clinically useful method. J Neurosurg. 2010;112 (2):273-84. doi: 10.3171/2008.9.JNS08451
• 4)   Mattogno  PP,  D'Alessandris  QG,  Rigante  M, et al. Reliability of intraoperative visual evoked potentials (iVEPs) in monitoring visual function during endoscopic transsphenoidal surgery. Acta Neurochir (Wien). 2023;165 (11):3421-9. doi: 10.1007/s00701-023-05778-1
• 5)   Banoub  M,  Tetzlaff  JE,  Schubert  A. Pharmacologic and physiologic influences affecting sensory evoked potentials: implications for perioperative monitoring. Anesthesiology. 2003;99 (3):716-37. doi: 10.1097/00000542-200309000-00029
• 6)   Hayashi  H,  Kawaguchi  M. Intraoperative monitoring of flash visual evoked potential under general anesthesia. Korean J Anesthesiol. 2017;70 (2):127-35. doi: 10.4097/kjae.2017.70.2.127
• 7)   Mikami  T,  Hirano  T,  Sugino  T, et al. Presurgical planning for arteriovenous malformations using multidetector row CT. Neurosurg Rev. 2012;35 (3):393-9; discussion 399-400. doi: 10.1007/s10143-012-0383-4
• 8)   Marsh  B,  White  M,  Morton  N, et al. Pharmacokinetic model driven infusion of propofol in children. Br J Anaesth. 1991;67 (1):41-8. doi: 10.1093/bja/67.1.41
• 9)   Glen  JB. The development of 'Diprifusor': a TCI system for propofol. Anaesthesia. 1998;53 (suppl 1):13-21. doi: 10.1111/j.1365-2044.1998.53s115.x
• 10)   Luo  Y,  Regli  L,  Bozinov  O, et al. Clinical utility and limitations of intraoperative monitoring of visual evoked potentials. PLoS One. 2015;10 (3):e0120525. doi: 10.1371/journal.pone.0120525
• 11)   Kamio  Y,  Sakai  N,  Sameshima  T, et al. Usefulness of intraoperative monitoring of visual evoked potentials in transsphenoidal surgery. Neurol Med Chir (Tokyo). 2014;54 (8):606-11. doi: 10.2176/nmc.oa.2014-0023
• 12)   Uribe  AA,  Mendel  E,  Peters  ZA, et al. Comparison of visual evoked potential monitoring during spine surgeries under total intravenous anesthesia versus balanced general anesthesia. Clin Neurophysiol. 2017;128 (10):2006-13. doi: 10.1016/j.clinph.2017.07.420
• 13)   Houlden  DA,  Turgeon  CA,  Amyot  NS, et al. Intraoperative flash visual evoked potential recording and relationship to visual outcome. Can J Neurol Sci. 2019;46 (3):295-302. doi: 10.1017/cjn.2019.4
• 14)   Tanaka  R,  Tanaka  S,  Ichino  T, et al. Differential effects of sevoflurane and propofol on an electroretinogram and visual evoked potentials. J Anesth. 2020;34 (2):298-302. doi: 10.1007/s00540-020-02733-7
• 15)   Qiao  N,  Yang  X,  Li  C, et al. The predictive value of intraoperative visual evoked potential for visual outcome after extended endoscopic endonasal surgery for adult craniopharyngioma. J Neurosurg. 2021;135 (6):1714-24. doi: 10.3171/2020.10.JNS202779
• 16)   Tao  X,  Yang  X,  Fan  X, et al. Prediction of post-operative visual deterioration using visual-evoked potential latency in extended endoscopic endonasal resection of craniopharyngiomas. Front Neurol. 2021;12:753902. doi: 10.3389/fneur.2021.753902
• 17)   Ma  J,  Xiong  W,  Guo  D, et al. Effects of sevoflurane-propofol-balanced anesthesia on flash visual evoked potential monitoring in spine surgery: a randomized noninferiority trial. Anesth Analg. 2022;134 (5):1054-61. doi: 10.1213/ANE.0000000000005742
• 18)   Nakagawa  I,  Park  H,  Kotsugi  M, et al. Diagnostic impact of monitoring visual evoked potentials to prevent visual complications during endovascular treatment for intracranial aneurysm. Front Neurol. 2022;13:761263. doi: 10.3389/fneur.2022.761263
• 19)   Tao  X,  Fan  X,  Gui  S, et al. Efficacy of intraoperative visual evoked potential amplitude reduction in predicting visual outcome after extended endoscopic endonasal resection of craniopharyngiomas. J Neurosurg. 2024;140 (5):1270-5. doi: 10.3171/2023.9.JNS23457
• 20)   Yamada  S,  Hayamizu  K,  Akiyama  Y, et al. Effect of remimazolam on intraoperative neurophysiology monitoring of visual-evoked potential: a case series. J Anesth. 2023;37 (2):311-4. doi: 10.1007/s00540-022-03159-z
• 21)   Lyon  R,  Feiner  J,  Lieberman  JA. Progressive suppression of motor evoked potentials during general anesthesia: the phenomenon of "anesthetic fade". J Neurosurg Anesthesiol. 2005;17 (1):13-9.
• 22)   Macdonald  DB. Intraoperative motor evoked potential monitoring: overview and update. J Clin Monit Comput. 2006;20 (5):347-77. doi: 10.1007/s10877-006-9033-0
• 23)   Seyal  M,  Mull  B. Mechanisms of signal change during intraoperative somatosensory evoked potential monitoring of the spinal cord. J Clin Neurophysiol. 2002;19 (5):409-15. doi: 10.1097/00004691-200210000-00004
• 24)   Krajčová  A,  Waldauf  P,  Anděl  M, et al. Propofol infusion syndrome: a structured review of experimental studies and 153 published case reports. Crit Care. 2015;19:398. doi: 10.1186/s13054-015-1112-5
• 25)   Hughes  MA,  Glass  PS,  Jacobs  JR. Context-sensitive half-time in multicompartment pharmacokinetic models for intravenous anesthetic drugs. Anesthesiology. 1992;76 (3):334-41. doi: 10.1097/00000542-199203000-00003
• 26)   Fudickar  A,  Bein  B. Propofol infusion syndrome: update of clinical manifestation and pathophysiology. Minerva Anestesiol. 2009;75 (5):339-44.
• 27)   Hemphill  S,  McMenamin  L,  Bellamy  MC, et al. Propofol infusion syndrome: a structured literature review and analysis of published case reports. Br J Anaesth. 2019;122 (4):448-59. doi: 10.1016/j.bja.2018.12.025
• 28)   Liolios  A,  Guérit  JM,  Scholtes  JL, et al. Propofol infusion syndrome associated with short-term large-dose infusion during surgical anesthesia in an adult. Anesth Analg. 2005;100 (6):1804-6. doi: 10.1213/01.ANE.0000153017.93666.BF
• 29)   Merz  TM,  Regli  B,  Rothen  HU, et al. Propofol infusion syndrome--a fatal case at a low infusion rate. Anesth Analg. 2006;103 (4):1050. doi: 10.1213/01.ane.0000239080.82501.c7
• 30)   Eriksen  J,  Povey  HM. A case of suspected non-neurosurgical adult fatal propofol infusion syndrome. Acta Anaesthesiol Scand. 2006;50 (1):117-9. doi: 10.1111/j.1399-6576.2006.00904.x
• 31)   Toyama  K,  Wanibuchi  M,  Honma  T, et al. Effectiveness of intraoperative visual evoked potential in avoiding visual deterioration during endonasal transsphenoidal surgery for pituitary tumors. Neurosurg Rev. 2020;43 (1):177-83. doi: 10.1007/s10143-018-1024-3
• 32)   Chen  X,  Li  Q,  Liu  X, et al. Visual acuity evaluated by pattern-reversal visual-evoked potential is affected by check size/visual angle. Neurosci Bull. 2012;28 (6):737-45. doi: 10.1007/s12264-012-1292-9