Motor network neurophysiological (TMS-EEG) markers as predictors of outcomes of ventriculoperitoneal shunt sustained for over a year in normal pressure hydrocephalus

This article was published in Fluids and Barriers of the CNS (2026) under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) licence and is reproduced here in full. Tables and supplementary material are available in the original article.

Abstract

Background and objectives. Current practice for normal pressure hydrocephalus (NPH) relies on clinical presentation, imaging, and various clinical procedures to determine the need for neurosurgical intervention, specifically the installation of a ventriculoperitoneal shunt (VPS). This study aimed to evaluate the utility of a TMS-evoked potential (TEP)-based assessment (Delphi-MD) for predicting sustained response (over one year) to VPS, serving as an ancillary test to the cerebrospinal fluid tap test (CTT).

Methods. Forty-one patients with possible iNPH underwent Delphi-MD evaluation of wave form adherence (WFA), a composite measure assessing how closely their TEPs resembled an ideal benchmark TEP. Of these, 18 patients received VPS treatment following a multidisciplinary team decision based on standard practice, blinded to the TEP results. Response to the intervention was evaluated over a period of more than one year using the clinical global impression of change (CGIC) scale, in addition to assessments with a 10-meter timed up and go test and the modified Rankin scale.

Results. Responders had higher M1L WFA values at baseline (prior to VPS) than non-responders (median 0.51 [95% CI 0.36–0.76], median 0.14 [95% CI -0.02-0.38], respectively p = 0.012). M1L WFA was successful in distinguishing responders (11/18) from non-responders (7/18) with an ROC-AUC of 0.935 [bootstrap 95% CI 0.792-1.00], p < 0.0001. M1L WFA cutoff was established with LOOCV and produced 90.9% sensitivity, 85.7% specificity and 88.9% accuracy. In contrast the CTT result was unsuccessful in predicting shunt responsiveness.

Conclusion. This study indicates that higher M1L WFA obtained through Delphi-MD (TMS-EEG), signifies better preservation and functional integrity of the motor network in patients with probable iNPH that had a sustained response to VPS than those that did not. Given replication of results in future independent cohorts, Delphi-MD shows promise as an objective, accurate, and accessible ancillary screening tool for iNPH candidates, prior to referral for VPS treatment.

Trial registration. Registration number: MOH_2021-12-08_010449. Registration date: 11.09.2021. https://clinicaltrial.health.gov.il/clinicaltrials/?sort=experimentapproveddate_tdt+desc&page=1&searchQuery=MOH_2021-12-08_010449.

Supplementary Information. The online version contains supplementary material available at https://doi.org/10.1186/s12987-026-00844-z.

Keywords: Normal pressure hydrocephalus, TMS evoked potentials, Ventriculoperitoneal shunt

Introduction

Normal pressure hydrocephalus (NPH) is a potentially treatable neurological syndrome presenting in older adults, marked by gait disturbances, cognitive decline, and urinary incontinence alongside ventriculomegaly with normal cerebrospinal fluid (CSF) pressure [1]. Though the exact pathophysiology remains incompletely understood, proposed mechanisms include disrupted CSF absorption, altered brain compliance, and periventricular white matter changes [2]. The current standard of care for NPH involves surgically implanting a ventriculoperitoneal shunt (VPS), which drains excess CSF. This procedure has shown significant clinical benefits, particularly in improving gait disturbances. Properly selected patients can experience initial improvement in symptoms, with success rates ranging from 80% to 90% [3–5]. However, long-term effects are often not systematically documented or evaluated in routine clinical care. Shunt surgery carries risks such as infection, over-drainage, and subdural hematoma [6], therefore accurate screening is required, to make an informed and objective evaluation of the risk-benefit ratio, to increase the rates of clinically meaningful and sustained benefit, and minimize unnecessary interventions. Current diagnostic screening for VPS candidacy relies on imaging markers (e.g., enlarged ventricles, narrow callosal angle) and invasive procedures such as the CSF tap test (CTT), extended lumbar drainage, infusion tests, or intracranial pressure [1, 7]. While these methods have reasonable positive predictive value, the CTT in particular is uncomfortable for patients, and its negative predictive value is limited, as up to 50% improve after shunting despite a negative CTT [8, 9]. However, these rates are typically derived from short-term follow-up studies (≤ 3–6 months), when postoperative gains are most pronounced. Longitudinal data indicate that many patients experience either waning benefit or shunt-related complications over time [6], underscoring the need for follow-up periods longer than 6 months to identify patients that may obtain long term benefit. This highlights the need for additional or alternative tools for prediction of a successful intervention, with superior performance over current standard practice; these should preferably be objective and accessible and more patient-friendly screening tools [10].

Transcranial magnetic stimulation combined with electroencephalography (TMS–EEG), is emerging as a promising neurodiagnostic tool that noninvasively measures cortical reactivity via TMS-evoked potentials (TEPs) [11, 12]. It holds promise as a novel direct methodology to test a real time status of brain’s neurophysiological function [13]. In a previous prospective study, by Davidy et al. (2024) [14], we reported the performance of Delphi TMS-EEG motor and prefrontal index in predicting clinical response 3 months following VPS. The Delphi index outperformed traditional screening methods, including MRI markers and CTT, in predicting VPS response at three-month follow-up (ROC AUC = 0.91), and provided superior accuracy coupled with greater tolerability for patients [14]. Building on these encouraging results, the current study extends this investigation by assessing the Delphi-MD (TMS–EEG) in the same cohort over a 1 to 2-year follow-up period.

Extending beyond the early postoperative window to test whether Delphi retains its predictive accuracy over time. This longitudinal analysis is critical for determining whether Delphi can reliably identify patients who achieve long-term benefit from VPS, ultimately enhancing patient selection and long-term management of NPH.

Materials and methods

Study participants

This was an exploratory, non-interventional longitudinal study, in possible iNPH patients arriving at the Movement Disorders Institute of the Department of Neurology at Sheba Medical Center (SMDI, Israel). The study received approval from the Sheba Medical Center local Helsinki committee, and all subjects signed an informed consent in accordance with GCP guidelines.

Patients were included based on the following eligibility criteria: Inclusion criteria: (a) Male and female patients at age of 60–85. (b) presenting with possible iNPH⁴ that consisted of a gait disorder in addition to cognitive impairment and/or incontinence (NPH triad of symptoms). (c) Demonstrating ventriculomegaly in a recent CT or MRI scan. (d) there were no identifiable secondary reasons for the hydrocephalus, such as a previous brain injury, hemorrhage, infection, radiation, or neurosurgical procedures.

Exclusion criteria: (a) Additional cerebral structural anomalies per imaging. (b) Other central nervous system disease. (c) Other major medical problems affecting gait.

Study procedures

Baseline symptom evaluation

Screening of patients for referral to VPS included collection of demographic and clinical information from patients and their medical records, a full neurological examination and focused evaluation of gait and balance, cognitive state and continence. Gait impairment was assessed through “timed up and go” test (TUG) [15] where the total time taken to complete the task is recorded for a 10-meter TUG. Urinary incontinence assessment was rated using a five rank scale from the iNPH grading scale [16]. Cognitive assessment was performed with Montreal cognitive assessment (MoCA) and MMSE [17, 18], and general functionality was assessed by modified Rankin Scale (MRS) [19].

MRI/CT brain scans

Each patient had either a brain CT or MR imaging scan before referral to the SMDI. A neuroradiologist evaluated and ranked the scans for Evans index (the ratio of the maximum width of the frontal horns of the lateral ventricles and the maximal internal diameter of the skull), Sylvian fissure dilation, crowding sulci at the vertex, temporal horns width, narrow callosal angle, peri-ventricular hypodensities and Fazekas scale score (quantifies the amount of deep white matter lesions), and presence of disproportionately enlarged subarachnoid-spaces (DESH) [20], which was positive when all 3 components of enlarged ventricles tight high convexity, and dilated Sylvian fissure, were present [1]. The imaging findings were considered as part of the diagnostic workup and VPS referral process, with only ventriculomegaly required for inclusion, while the remaining imaging measurements were present variably across patients. Brain MRI scans of HC were previously evaluated and ruled out for significantly pathological or remarkable findings per age.

CSF tap test (CTT)

The CTT consists of a lumbar puncture (LP) with drainage of 30–50 ml of CSF, performed by a neurologist or neurosurgeon under local anesthesia. Initial opening pressure is determined, and CSF specimens are collected for analysis. A 10-meter TUG test is performed before, and 3–4 h following the CTT. The CTT result is considered as the difference between baseline TUG and post CTT TUG times.

TMS-EEG (Delphi) procedure

TMS-EEG acquisition was performed with the Delphi system version 1.0 including Delphi acquisition and analysis software (QuantalX Neuroscience), EEG compatible TMS stimulator and 65 mm figure 8 coil (MagPro R30 stimulator (MagVenture, Denmark) with MCF-B65-HO figure-8 Coil (MagVenture, Denmark)), TMS compatible DC coupled amplifier with sampling rate of 5 kHz (Delphi Amplifier, QuantalX Neuroscience Ltd.) and 34 electrode cap with Ag\AgCl sintered electrodes (MCS Delphi cap, QuantalX Neuroscience Ltd.). The reference and ground electrodes were affixed to the ear lobes.

All the subjects included in the study performed the same TMS-EEG (Delphi) evaluation prior to CTT. The resting motor threshold (RMT) was obtained at the beginning of each session, by stimulating the left and right M1 according to guidelines⁶ and was defined as the intensity that produced a visible twitch in abductor policis brevis in 50% out of total stimulations. A 0.5 mm foam pad was attached to the magnetic coil and ear plugs were used to minimize electrode movement and bone-conducted auditory artefact. The magnetic coil was positioned over the left and right M1 at 45◦ toward the contralateral forehead.

Single pulse (< 0.3 Hz), 85% of RMT intensity, with 18 repetitions, was applied to left primary motor cortex (M1). Data acquisition, pre-processing, and cleaning of the transcranial evoked response was performed with Delphi software 1.0 [14, 21–26].

Based on our previous findings, we sought to test the hypothesis that M1L stimulation may be sufficient to identify iNPH patients likely to demonstrate sustained gait improvement following VPS. Accordingly, the current study focused on this predefined motor-related outcome and did not include further exploration of additional TMS-EEG metrics or stimulation sites, although prefrontal stimulations were also performed. Future studies with larger cohorts and longitudinal follow-up are planned to investigate whether DLPFC-derived TEP features may additionally predict cognitive outcomes following VPS.

TEP preprocessing

All preprocessing and quality control procedures were performed automatically by Delphi software (Ver 1.0). Preprocessing included filtering to an effective 1–45 Hz passband [27], powerline noise removal and TMS decay artifact removal using ICA (Independent component analysis) decomposition [28–30]. Electrooculography (EOG) artifacts were corrected using Canonical correlation analysis (CCA) which uses maximization of linear independence [31–33]. Data segments with high noise levels were automatically detected through processing stages using thresholding in the temporal and spectral domains. Trials with excessive noise were discarded, and noisy channels were interpolated using data from neighboring channels. Datasets were excluded when trial rejection exceeded 30% or when insufficient spatial coverage remained. Real-time alerts were provided to the operator for datasets classified as insufficient or noisy.

TEP analysis

The Delphi algorithm automatically analyses the regional and network TEP recorded in response to each stimulation site and extracts numeric output of TEP features [21–24]. These features include specific peak latencies, amplitudes and slopes of typical negative and positive peaks [34–36] detected automatically by the Delphi algorithm (Fig. 1A, B).

WFA is a composite measure that captures the entire TEP waveform and quantifies its adherence to a benchmark waveform by calculating the mean relative distance across all time points of the TEP. -1 (low adherence) to 1 (high adherence). The benchmark waveform was generated from a separate Delphi TMS-EEG database (similar TMS stimulation, recording and processing attributes) collected from 108 neurologically and psychiatrically healthy subjects, (Age range 25–55, mean age 43 ± 9, 32% females and 14% left-handed). This waveform serves as a static reference benchmark reflecting the TEP signature of a healthy, intact brain.

Referral to VPS

The final recommendation for a VPS was made by a multidisciplinary team comprising movement disorder specialists and a neurosurgeon. Factors considered include clinical characteristics, brain imaging, gait improvement in CTT, comorbidities, frailty, age, cognitive function, and overall functionality. The clinical team was blinded to the Delphi results.

VPS implantation and peri and post-operative shunt setting was performed according to clinical practice [37], for details please refer to supplementary materials.

Determination of clinical benefit

Follow-up visits were performed at around 3 months, 6 months and 12–18 months from VPS, for the evaluation of symptoms and adverse events and were performed by a multidisciplinary team of neurologists and a neurosurgeon. The primary outcome of this study was defined as response to treatment at 12 months and defined using the clinical global impression of change scale (CGI-C) [38] regarding gait (1 = marked improvement; 2 = moderate improvement; 3 = minimal improvement; 4 = no change; 5 = minimal worsening; 6 = moderate worsening; 7 = marked worsening). Evaluations of symptoms and adverse events were based on the healthcare professional’s clinical impression following patient and caregiver interviews and observation of clinical scales. The clinical team was blinded to the TMS-EEG automatic analysis all through study conduct.

In addition to CGI-C, baseline and follow up scores of the Modified Rankin scale (MRS) [19], and TUG score compared to baseline were collected.

The Clinical Global Impression of Change (CGI-C) scale provides a clinician-rated, holistic evaluation of overall patient improvement that integrates multiple domains of functioning. While tests such as the Timed Up and Go (TUG) or other quantitative gait measures are sensitive to changes in mobility, they capture only a narrow dimension of the NPH symptom complex and are prone to variable patient effort and clinical setting. In contrast, CGI-C allows the clinician to incorporate gait quality, balance, initiation, step pattern, and functional mobility in context with cognition and continence, reflecting the broader impact of shunt treatment on patient well-being.

See supplementary materials for detailed peri and post-surgical protocol.

Statistical analysis

Two-tailed independent t-tests or Fisher’s exact tests were used to examine differences in characteristic measures between the groups. Spearman’s rank correlation and Pearson correlation were used to test associations of disease, brain imaging and TMS-EEG measures. The effect of time since VPS on treatment effect (CGI-C scores) was evaluated using linear mixed-effects models with random intercepts for subjects, estimated with restricted maximum likelihood (REML). Both linear and quadratic time terms were tested to capture potential non-linear change trajectories.

The primary outcome was defined as CGI-C score regarding gait, more than one year after VPS implantation. Spearman correlation was used to assess association between the level of improvement and the Delphi-MD (TMS-EEG). Groups’ means were compared with the Mann-Whitney U test. Receiver operating curve (ROC) analysis was performed to establish baseline WFA ability to predict responders from non-responders. To obtain confidence intervals for the AUC, we applied nonparametric bootstrapping with 5,000 iterations and optimal cutoff using the Youden’s index. To evaluate the generalizability of the ROC-derived cutoff, we applied leave-one-out cross-validation (LOOCV) using a fixed threshold (optimal Youden-determined from the full dataset). For each fold, the left-out subject was classified using this predefined cutoff, and overall accuracy, sensitivity, and specificity were calculated across all subjects.

Statistical analysis was performed in IBM SPSS Statistics, version 30 and plots created with GraphPad Prism version 10.1.1.

Results

A total of 41 potential iNPH patients were enrolled in the study and underwent Delphi-MD (TMS-EEG) assessment. Of these, 22 patients were referred to VPS, based on clinical presentation, cranial imaging, and CTT. Ultimately, 18 patients proceeded with VPS implantation and completed one year or more of follow-up. A STROBE diagram is depicted in Supplementary Fig. 1. The demographic and clinical characteristics at baseline of the probable NPH patients who underwent VPS and followed up for over a year, are summarized in Table 1. There was a predominance of male participants in our cohort (83.3%).

Table 1. Baseline characteristics of probable iNPH patients who underwent VPS and over 12 months follow-up. The table is available in the original article.

Response to treatment with VPS

CGI-C score was determined at each of the follow up visits that were completed within specific time frames from VPS, 3 months, 6 months and ≥ 12 months (mean ± SD: 3.4 ± 1.3, 6.4 ± 1.7 and 16.1 ± 3.6, respectively), with percentage of patients improved ranging from 77.23%, 89.47% and 61.11%, respectively (Figs. 2A, B,C). While most of the patients were classified as ‘responders’ in the first few months (CGI-C score < 4), the rates of improvement diminished with extended follow-up periods. The CGI-C score over time per each patient is described in Fig. 2D. A linear mixed-effects model with random intercepts showed no significant linear effect of time on CGI-C scores (p = 0.08). However, when a quadratic term was added, time exerted a significant curvilinear effect (linear: p < 0.001; quadratic: p < 0.001), indicating a curvilinear trajectory of treatment response where after some months the benefit plateaus and may diminish for some patients (Fig. 2D, E).

Prediction of treatment response

The primary endpoint analysis included the 18 patients that completed the ≥ 1 year (12–18 months) follow up and had the CGI-C rank determined. CGI-C score at 1–2 year follow up was used to determine the ‘ground truth’ and was used to classify each case: CGI-C score < 4 [responder], CGI-C score ≥ 4 [non-responder]. CTT results showed no significant correlation with the CGI-C score at more than one-year post-VPS (r = 0.024, p = 0.92) (Fig. 3A). Additionally, there were no significant differences in mean CTT result scores between the groups (p = 0.859) (Fig. 3B). The CTT results did not significantly differentiate responders from non-responders. (AUC = 0.56 [95% CI: 0.23–0.88] (Fig. 3C) maximal Youden index cutoff contributed to a 90.9% sensitivity and 57.1% specificity and 63% accuracy. Correlation between 1 year CGI-C score and WFA demonstrated significant negative correlation (r = 0.58, p = 0.014) (Fig. 3D). Responders had higher M1L WFA values than non-responders (median 0.51 [95% CI 0.36–0.76], median 0.14 [95% CI -0.02-0.38], respectively p = 0.012) (Fig. 3E). M1L WFA at baseline (prior to VPS) was successful in discriminating responders (11/18) from non-responders (7/18) with an AUC of 0.935 [95% CI 0.792-1.00], p < 0.0001, as assessed by the CGI-C at ≥ 12 months follow up (Fig. 3F). The Youden-optimal cutoff of 0.36 yielded 90.9% sensitivity and 85.7% specificity in the full dataset. To address potential optimism bias, nested leave-one-out cross-validation (LOOCV) was subsequently performed with fold-specific re-derivation of the Youden cutoff. Under this more conservative validation framework, specificity remained stable at 85.7%, while sensitivity and overall accuracy were 72.7% and 77.8%, respectively, supporting the robustness of the discriminatory signal despite the small sample size. The ‘responders’ versus ‘non-responders’ mean ± 95% CI TEP waveform is presented in Fig. 4.

Discussion

This study aimed to assess patients who are iNPH candidates for VPS surgery, using Delphi-MD (TMS-EEG). We hypothesized, based on preliminary findings, that the neurophysiological data obtained from TMS-EEG could help predict which patients would benefit from VPS in the long term. The results of the TMS-EEG evaluation suggest that patients with preserved motor network function may experience improvements in gait after VPS surgery. This finding indicates that neurophysiological assessment of motor networks, presenting preserved motor network function, could potentially serve as a predictor for postoperative gait improvement in NPH patients. This composite measure of Delphi-MD in response to M1L stimulation, depicts the resemblance to an ideal benchmark waveform, is an indication of the preservation and neurophysiological functionality of the motor network. The higher the WFA the larger the improvement in gait. Furthermore, using a binary cutoff at baseline, this neurophysiological measure was able to predict patients who would respond to VPS after more than one year of follow-up. The prediction demonstrated a sensitivity of 90.9%, a specificity of 85.7%, and an overall accuracy of 88.9%. This suggests that the measure is potentially a highly effective predictor of long-term treatment response.

The ability of TMS–EEG to capture effective connectivity in real time has been demonstrated both in task-specific investigations [39, 40] and through multimodal imaging studies [41]. While the structural connectome, as assessed by conventional imaging, remains relatively stable, the functional connectome is dynamic and varies according to the current cognitive state, thereby altering the strength of functional connections. When evaluating resting-state functional connectivity, as modeled in the present study, neural activity appears to be organized within restricted specialized networks that can be conceptualized as interconnected hubs and nodes [42]. Previous studies have demonstrated that higher M1 WFA values corresponded to higher white matter integrity, of transcallosal fibers, in patients with post stroke and traumatic brain injury [23]. The full TEP waveform recorded contralateral to a given stimulation site (e.g., left M1 stimulation, as presented here) reflects the propagation of the evoked response across contralateral regions within the corresponding network. WFA reflects effective connectivity by quantifying the deviation of an individual subject’s TEP waveform from a reference TEP signature derived from an intact adult brain. This approach enables standardized characterization of the individual TEP response using a unified and comparable metric.

In many cases, the determination of whether the syndrome presented by a specific patient is in fact idiopathic NPH or whether it is a result another primary neurodegenerative entity or comorbidity cannot be ascertained with confidence at the initial evaluation [10]. Two previous studies demonstrated that both M1 and DLPFC WFA values are reduced in patients with Parkinson’s disease [22, 25]. In contrast, the iNPH responders in the present study exhibited higher WFA values. These findings further support the distinction between definite iNPH [1] and Parkinson’s disease, one of the neurological conditions most frequently confounded with iNPH in clinical practice.

Having an objective, quantitative tool to evaluate the brain’s potential for improvement, based on the neurophysiological state of relevant networks, is truly inspiring. Such a tool can enhance personalized treatment approaches and improve prediction of patient outcomes in neurological rehabilitation and management. Current standard practices rely on invasive CSF drainage tests, which can cause significant discomfort and carry potential complications. These procedures also require hospitalization and the involvement of specialized teams and facilities, making them resource-intensive and less patient-friendly. In this study, the CTT results failed to reliably predict treatment response at over one year of follow-up, achieving only a modest accuracy of 63%. This finding is consistent with previous meta-analyses, which also reported limited predictive value of the CTT for long-term outcomes [9].

In our previous publication, where we presented the results of a 3 month- follow-up [14] of this iNPH cohort, a composite measure of DLPFC and M1 Delphi TEP measures comprised by a multivariate logistic regression model predicted responders successfully. As our results show, the long-term benefit of VPS may decline over time. We observed a curvilinear pattern of symptom improvement, with gains reaching a plateau after 6 months, followed by partial decline in improvement among some patients by the 1-year post-shunting follow-up. A recent retrospective longitudinal report of NPH patients following VPS reported a similar percentage of responders at 1 year (61.40%) as in our study (61.11%) with additional decline that reached 44.0% around three years of follow-up [6]. While about half of the NPH patients undergoing VPS display marked benefits in symptoms and function in everyday living, more accurate and objective screening methods will help increase VPS success rates, saving costs and improving patient outcomes.

Our study is subject to several limitations. First, the iNPH sample size is relatively small. This introduces the possibility of optimism bias and overfitting of the ROC-derived performance estimates. Although leave-one-out cross-validation and bootstrap confidence intervals were used to improve internal validity, external validation in larger independent cohorts will be necessary to establish the reproducibility and generalizability of these findings. Second, patients underwent a CTT rather than extended lumbar drainage or infusion tests to compare with the TEP. Nevertheless, we favor the CTT in our routine practice, and it remains the predominant screening modality in current U.S. clinical practice [42]. Furthermore, our outcome assessments focused exclusively on gait, leaving potential changes in urinary incontinence or cognitive impairment unevaluated. Finally, although we utilized three distinct clinical outcome measures—the CGI-C, the 10-meter TUG, and the mRS—none can fully capture the complex, multi-faceted impact of NPH-related gait disorders. To validate these findings, a prospective, multicenter study with a larger patient cohort is warranted. Such a future study should ideally incorporate broader secondary outcomes relevant to NPH, as well as sensor-based clinical and home monitoring of gait and physical activity.

Supplementary Information

Supplementary material is available in the original article.

Acknowledgements

The Delphi-MD device was loaned to the research facility by QuantalX Neuroscience, the company that manufactures Delphi-MD.This study received no funding.

Abbreviations

• TEPs: TMS Evoked Potentials
• EEG: Electroencephalogram
• TMS: Transcranial Magnetic Stimulation
• NPH: Normal Pressure Hydrocephalus
• VPS: Ventriculoperitoneal Shunt

Author contributions

TD contributed to conceptualization, methodology, writing of original draft, investigation, data curation, writing – review & editing. SA contributed to conceptualization, writing – review & editing. AS: contributed to the investigation. YZ contributed to the investigation, writing - original draft. TFK contributed to writing – review & editing, AS contributed to project administration. NZ contributed to writing-original draft, formal analysis. HF contributed to software, writing – review & editing, OLL contributed to software, writing – review & editing, OLS contributed to data curation, investigation, writing – review & editing, SHB contributed to conceptualization, methodology, project administration, data curation, supervision, writing – original draft, visualization, and writing – review & editing.

Funding

No funding was received towards this work. The Delphi-MD (TMS-EEG) system was loaned to the research facility by QuantalX Neuroscience, the company that manufactures Delphi-MD.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available due to privacy and regulatory restrictions but are available from the principal investigator upon reasonable request and subject to institutional approval.

Declarations

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Sheba Medical Center local Helsinki committee. All procedures were conducted in accordance with the ethical standards of the institutional research committee and the Declaration of Helsinki. Written informed consent was obtained from all participants prior to participation.

Consent for publication

Written informed consent was obtained from all participants for the use of their anonymized data for scientific publication.

Competing interests

QuantalX Neuroscience is the company developing Delphi software used for acquisition and analysis of the TMS-EEG data. The company loaned the medical institute the hardware and software and supported the performance of the Delphi evaluations. NZ, HF, OLL are employees of QuantalX Neuroscience.TD was reimbursed by QuantalX for her travel and registration expenses to the American Academy of Neurology meeting in 2023. Remaining authors have no disclosures or potential competing interests regarding this paper. No other funding was provided by the company for the conduct of the study. There are no competing interests to disclose.

Previous Presentations

International Hydrocephalus Society; European Association of Neurosurgical Societies.

References

1. Nakajima M, Yamada S, Miyajima M, Ishii K, Kuriyama N, Kazui H, et al. Guidelines for management of idiopathic normal pressure hydrocephalus (Third Edition): endorsed by the Japanese Society of Normal Pressure Hydrocephalus. Neurol Med Chir (Tokyo). 2021;61(2):63–97
2. Oliveira LM, Nitrini R, Román GC. Normal-pressure hydrocephalus: A critical review. Dement Neuropsychol. 2019;13(2):133–43.
3. Williams MA, Malm J. Diagnosis and Treatment of Idiopathic Normal Pressure Hydrocephalus. Continuum (Minneap Minn). 2016;22(2 Dementia):579–99.
4. Halperin JJ, Kurlan R, Schwalb JM, Cusimano MD, Gronseth G, Gloss D. Practice guideline: Idiopathic normal pressure hydrocephalus: Response to shunting and predictors of response: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Neurology. 2015;85(23):2063–71.
5. Luciano MG, Williams MA, Hamilton MG, Katzen HL, Dasher NA, Moghekar A, et al. A randomized trial of shunting for idiopathic normal-pressure hydrocephalus. N Engl J Med. 2025;393(22):2198–209.
6. Chen J, Xian J, Wang F, Zuo C, We L, Chen Z, et al. Long-term outcomes of ventriculoperitoneal shunt therapy in idiopathic normal pressure hydrocephalus. BMC Surg. 2025;25(1):157.
7. Skalický P, Mládek A, Vlasák A, De Lacy P, Beneš V, Bradáč O. Normal pressure hydrocephalus-an overview of pathophysiological mechanisms and diagnostic procedures. Neurosurg Rev. 2020;43(6):1451–64.
8. Thavarajasingam SG, El-Khatib M, Rea M, Russo S, Lemcke J, Al-Nusair L, et al. Clinical predictors of shunt response in the diagnosis and treatment of idiopathic normal pressure hydrocephalus: a systematic review and meta-analysis. Acta Neurochir (Wien). 2021;163(10):2641–72.
9. Mihalj M. CSF tap test - Obsolete or appropriate test for predicting shunt responsiveness? A systemic review. J Neurol Sci. 2016;370:157.
10. Fasano A, Espay AJ, Tang-Wai DF, Wikkelsö C, Krauss JK. Gaps, Controversies, and Proposed Roadmap for Research in Normal Pressure Hydrocephalus. Mov Disord. 2020;35(11):1945–54.
11. Fogel H, Zifman N, Hallett M. Utilization of Single-Pulse Transcranial-Evoked Potentials in Neurological and Psychiatric Clinical Practice: A Narrative Review. Neurol Int. 2024;16(6):1421–37.
12. Tremblay S, Rogasch NC, Premoli I, Blumberger DM, Casarotto S, Chen R, et al. Clinical utility and prospective of TMS-EEG. Clin Neurophysiol. 2019;130(5):802–44.
13. Ferreri F, Pasqualetti P, Määttä S, Ponzo D, Ferrarelli F, Tononi G, et al. Human brain connectivity during single and paired pulse transcranial magnetic stimulation. NeuroImage. 2011;54(1):90–102.
14. Davidy T, Anis S, Suminski A, Zauberman Y, Fay-Karmon T, Saar A, et al. Tms-evoked potentials: Neurophysiological biomarkers for diagnosis and prediction of response to ventriculoperitoneal shunt in normal pressure hydrocephalus. J Clin Neurosci. 2025;136:111234.
15. Schoene D, Wu SM, Mikolaizak AS, Menant JC, Smith ST, Delbaere K, et al. Discriminative ability and predictive validity of the timed up and go test in identifying older people who fall: systematic review and meta-analysis. J Am Geriatr Soc. 2013;61(2):202–8.
16. Kubo Y, Kazui H, Yoshida T, Kito Y, Kimura N, Tokunaga H, et al. Validation of grading scale for evaluating symptoms of idiopathic normal-pressure hydrocephalus. Dement Geriatr Cogn Disord. 2008;25(1):37–45.
17. Crum RM, Anthony JC, Bassett SS, Folstein MF. Population-based norms for the Mini-Mental State Examination by age and educational level. JAMA. 1993;269(18):2386–91.
18. Nasreddine ZS, Phillips NA, Bédirian V, Charbonneau S, Whitehead V, Collin I, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53(4):695–9.
19. Quinn TJ, Dawson J, Walters MR, Lees KR. Reliability of the modified Rankin Scale: a systematic review. Stroke. 2009;40(10):3393–5.
20. Kockum K, Lilja-Lund O, Larsson EM, Rosell M, Söderström L, Virhammar J, et al. The idiopathic normal-pressure hydrocephalus Radscale: a radiological scale for structured evaluation. Eur J Neurol. 2018;25(3):569–76.
21. Zifman N, Levy-Lamdan O, Suzin G, Efrati S, Tanne D, Fogel H, et al. Introducing a Novel Approach for Evaluation and Monitoring of Brain Health Across Life Span Using Direct Non-invasive Brain Network Electrophysiology. Front Aging Neurosci. 2019;11:248.
22. Maidan I, Zifman N, Hausdorff JM, Giladi N, Levy-Lamdan O, Mirelman A. A multimodal approach using TMS and EEG reveals neurophysiological changes in Parkinson’s disease. Parkinsonism Relat Disord. 2021;89:28–33.
23. Levy-Lamdan O, Zifman N, Sasson E, Efrati S, Hack DC, Tanne D, et al. Evaluation of White Matter Integrity Utilizing the DELPHI (TMS-EEG) System. Front Neurosci. 2020;14:589107.
24. Fogel H, Levy-Lamdan O, Zifman N, Hiller T, Efrati S, Suzin G, et al. Brain Network Integrity Changes in Subjective Cognitive Decline: A Possible Physiological Biomarker of Dementia. Front Neurol. 2021;12:699014.
25. Zifman N, Levy-Lamdan O, Hiller T, Thaler A, Dolev I, Mirelman A, et al. TMS-evoked potentials unveil occipital network involvement in patients diagnosed with Parkinson’s disease within 5 years of inclusion. NPJ Parkinsons Dis. 2024;10(1):182.
26. Lassman S, Zifman N, Fogel H, Hassin-Baer S, Anis S. TMS-evoked potentials provide novel neurophysiological features of Tourette syndrome. Parkinsonism Relat Disord. 2024;130:107217.
27. Cline CC, Lucas M, Sun Y, Menezes M, Etkin A. Advanced artifact removal for automated TMS-EEG data processing. In: 2021 10th International IEEE/EMBS Conference on Neural Engineering (NER); 2021 May 4–6.
28. Rogasch NC, Thomson RH, Farzan F, Fitzgibbon BM, Bailey NW, Hernandez-Pavon JC, et al. Removing artefacts from TMS-EEG recordings using independent component analysis: importance for assessing prefrontal and motor cortex network properties. NeuroImage. 2014;101:425–39.
29. Metsomaa J, Sarvas J, Ilmoniemi RJ. Multi-trial evoked EEG and independent component analysis. J Neurosci Methods. 2014;228:15–26.
30. Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods. 2004;134(1):9–21.
31. Jiang X, Bian GB, Tian Z. Removal of artifacts from EEG signals: a review. Sens (Basel). 2019;19(5).
32. Safieddine D, Kachenoura A, Albera L, Birot G, Karfoul A, Pasnicu A, et al. Removal of muscle artifact from EEG data: comparison between stochastic (ICA and CCA) and deterministic (EMD and wavelet-based) approaches. EURASIP J Adv Signal Process. 2012;2012(1):127.
33. Soomro MH, Badruddin N, Yusoff MZ, Jatoi MA. Automatic eye-blink artifact removal method based on EMD-CCA. In: Proceedings of the 2013 ICME International Conference on Complex Medical Engineering (CME); 2013. p. 186–190.
34. Lioumis P, Kicić D, Savolainen P, Mäkelä JP, Kähkönen S. Reproducibility of TMS-Evoked EEG responses. Hum Brain Mapp. 2009;30(4):1387–96.
35. Komssi S, Kähkönen S. The novelty value of the combined use of electroencephalography and transcranial magnetic stimulation for neuroscience research. Brain Res Rev. 2006;52(1):183–92.
36. Kähkönen S, Komssi S, Wilenius J, Ilmoniemi RJ. Prefrontal transcranial magnetic stimulation produces intensity-dependent EEG responses in humans. NeuroImage. 2005;24(4):955–60.
37. Yamada S, Ishikawa M, Nakajima M, Nozaki K. Reconsidering Ventriculoperitoneal Shunt Surgery and Postoperative Shunt Valve Pressure Adjustment: Our Approaches Learned From Past Challenges and Failures. Front Neurol. 2021;12:798488.
38. Studenski S, Hayes RP, Leibowitz RQ, Bode R, Lavery L, Walston J, et al. Clinical Global Impression of Change in Physical Frailty: development of a measure based on clinical judgment. J Am Geriatr Soc. 2004;52(9):1560–6.
39. Johnson JS, Kundu B, Casali AG, Postle BR. Task-dependent changes in cortical excitability and effective connectivity: a combined TMS-EEG study. J Neurophysiol. 2012;107(9):2383–92.
40. Daly I, Williams N, Nasuto SJ. TMS-evoked potential propagation reflects effective brain connectivity. J Neural Eng. 2024;21(6).
41. Atalay AS, Fecchio M, Edlow BL. Visualizing Effective Connectivity in the Human Brain. bioRxiv. 2025.
42. Bortoletto M, Veniero D, Thut G, Miniussi C. The contribution of TMS-EEG coregistration in the exploration of the human cortical connectome. Neurosci Biobehav Rev. 2015;49:114–24.

Citation and licence

Tal Davidy, Alexandra Suminski, Yakov Zauberman, Tsvia Fay-Karmon, Saar Anis, Adi Saar, Noa Zifman, Hilla Fogel, Ofri Levy-Lamdan, Orit Lesman-Segev, Sharon Hassin-Baer. Motor network neurophysiological (TMS-EEG) markers as predictors of outcomes of ventriculoperitoneal shunt sustained for over a year in normal pressure hydrocephalus. Fluids and Barriers of the CNS. 2026;23:119. doi:10.1186/s12987-026-00844-z

© The Author(s) 2026. Licence: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0). Changes: the text and figures are reproduced without changes; the layout has been adapted for this website and tables are linked to the original article.