Net optic nerve sheath diameter as a factor for predicting intracranial hypertension
Introduction
The evaluation of intracranial pressure (ICP) is crucial for the management of neurological emergencies. Excessively high ICP leads to focal or global hypoperfusion, cerebral ischemia, and brain herniation and is reported to correlate with increased mortality (1,2). In clinical settings, ICP elevation is mostly recognized by its consequences, such as deteriorated consciousness, pupillary light reflex impairment, and vital instability, which occur relatively late for further management.
Over the past few decades, invasive ICP monitoring has been applied for the early detection of intracranial hypertension. External ventricular drainage has been established as the gold standard for continuous ICP monitoring due to its positive outcomes and drainage usage. Antihypertensive treatments can be used for ICP approaching 20–25 mmHg, and early recognition and intervention due to ICP monitoring can improve clinical outcomes in certain—but not all—patients (2,3). According to the literature, up to 22% of patients who receive external ventricular drains (EVDs) develop procedure-associated infection, with (4-7) hematoma following the removal of an EVD tube being observed in 26% of patients (7). Device malfunction or transducer malposition during monitoring is also common (8). Moreover, this technique can be difficult to perform in patients with small ventricles or midline shifts.
Routine invasive ICP monitoring remains controversial for several reasons; therefore, efforts have been made to estimate ICP using noninvasive approaches. Radiological studies have suggested that a set of optic parameters are correlated with ICP, including optic nerve sheath diameter (ONSD), optic nerve diameter (OND), and ONSD/eyeball transverse diameter (EBTD) ratio (ONSD/EBTD) (9-12). To further refine these measurements, we introduced the net optic nerve sheath diameter (nONSD), which is calculated as ONSD minus OND. This parameter reflects the net width of the perioptic subarachnoid space, thereby reducing the confounding influence of optic nerve size. Compared with ONSD alone, nONSD may better capture sheath distension due to raised ICP and provide improved diagnostic accuracy. Given that no previous studies have systematically evaluated this parameter, we designed the present study to specifically examine whether nONSD, along with other optic parameters, offers additional value in predicting elevated ICP. The rationale for both ONSD and nONSD as predictors of ICP elevation lies in the anatomical continuity between the cranial and perioptic subarachnoid space. As ICP increases, cerebrospinal fluid (CSF) is driven by a pressure gradient and moves unidirectionally from the cranial to the peri-ON subarachnoid space. A retrospective study estimated the correlation between ONSD measured using cranial magnetic resonance imaging (MRI) and ICP elevation in patients with traumatic brain injury (TBI). At a cutoff of value of 5.82 mm, the ONSD had a sensitivity of 90% and a negative predictive value (NPV) of 92% for predicting ICP ≥20 mmHg (9). In an ultrasound study, at an optimal cutoff value of 6.3 mm, ONSD yielded a sensitivity and specificity of 94.7% and 90.9%, respectively, for detecting ICP ≥250 mmH2O (13). In the majority of studies, ONSD is typically evaluated by ultrasound, whereas much fewer studies have employed MRI.
In this study, we aimed to investigate in detail the correlations between optic parameters measured using cranial magnetic resonance (MR) and quantitative ICP obtained from lumbar puncture (LP). We present this article in accordance with the TRIPOD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1671/rc).
Methods
Setting and population
This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Beijing Tiantan Hospital, Capital Medical University (No. KY2021-163-02). The requirement for individual consent for this retrospective analysis was waived. We reviewed and screened patients treated in the emergency department between October 2020 and September 2022. Adult patients admitted for clinically suspected intracranial infections who underwent LP and cranial MR (the LP was carried out after the cranial MR) were eligible for the study. Patients in whom LP or cranial MR examinations failed or who were uncooperative in this regard, who had preexisting optic pathology or psychological disorders, who underwent LP before MR, or who had a time interval between LP and cranial MR longer than 8 hours were excluded from the study cohort.
Parameter measurements and data collection
All clinical data, LP records, and imaging information were obtained from the hospital’s electronic medical information system and retrospectively reviewed for analysis. Radiographic parameters were measured using cranial MR images (3.0 Tesla, 5-mm slice thickness, including at least axial T1- and T2-weighted sequences) that were extracted from the institutional radiographic grid. Optic parameters measured in the MR images (T2-weighted sequence) included the OND and the ONSD (Figure 1). The OND and ONSD were bilaterally measured 3 mm behind the eyeball, perpendicular to the linear axis of the optic nerve (10). The side with a larger ONSD (including the OND) was recorded. The maximum transverse diameter of the EBTD was also measured and recorded (14). The nONSD was developed as a novel variable to reflect the net diameter of the optic nerve sheath and was calculated by subtracting the OND from the ONSD. The quotients of the optic nerve sheath and eyeball transverse diameters were calculated and recorded (ONSD/EBTD). All optic parameters were measured and recorded by an investigator (Y.W.).
The ICP was obtained from the LPs through chart review. LPs were performed by neurology specialists, and the details of each procedure were documented in the electronic medical record. Patients were placed in a left lateral position, flexing their hips and knees and leaning their heads as close as comfortable to their knees. The patients’ lower back area was prepared with the application of aseptic technology. After the puncture needle had entered the subarachnoid space, the CSF opening pressure was recorded with legs fully extended. ICP elevation (intracranial hypertension) was defined as a CSF pressure >250 mmH2O during the LP procedure (13). According to the routine workflow in our emergency department, MR images required additional time for reconstruction and uploading to the hospital system. LPs were typically performed shortly after imaging in urgent cases, and the MRI results were not yet available to the operator at the time of the procedure. The interval should not have exceeded 8 hours.
Data analysis
All statistical analyses were performed using R version 4.3.1 (R Project for Statistical Computing, Vienna, Austria) and associated packages. Data with a normal distribution were presented as the mean ± standard deviation or as the median and interquartile range (IQR) for continuous variables and as frequencies (percentages) for categorical variables. The Student t-test and Mann-Whitney test were used to compare continuous variables. The chi-squared test and Fisher’s exact test were used for categorical variables. Linear correlations between optic parameter and ICP were examined, and the Pearson correlation coefficient was calculated. The receiver operating characteristic (ROC) curve was estimated to detect the diagnostic accuracy of the radiographic diameter. Patients were grouped to determine the detailed linear correlations between optic parameters and ICP and the accuracy of optic parameters in predicting intracranial hypertension. For all comparisons, a two-tailed P value <0.05 was deemed statistically significant.
Results
Study population
During the study period between October 2020 and September 2022, a total of 10,426 patients were reviewed, of whom 126 fulfilled the eligibility criteria. Among them, 14 patients were excluded due to failed LP (n=2), failed MR (n=5), optic pathology (n=1), psychological disorders (n=2), or a time interval between LP and cranial MR longer than 8 hours (n=4) (Figure 2). The final cohort included 112 patients, 41 of whom were found to have elevated ICP [median 290 (IQR 267–310) mmH2O]. Others were confirmed to have a normal ICP [median 150 (IQR 107.5–177.5) mmH2O] during LP (Table 1). In addition, the number of people in the normal ICP group who received anti-infection treatment was greater than that in the group with intracranial hypertension (P=0.02).
Table 1
| Characteristics | Normal ICP (n=71) | Elevated ICP (n=41) | P value |
|---|---|---|---|
| Demographics | |||
| Age (years) | 55.00 (37.00, 67.50) | 51.00 (34.00, 59.00) | 0.13 |
| Sex (male) | 48 (67.61) | 26 (63.40) | 0.81 |
| Current status | |||
| ICP | 150.00 (107.50, 177.50) | 290.00 (267.00, 310.00) | <0.001 |
| Epilepsy | 3 (4.23) | 2 (4.88) | >0.99 |
| Infection within 1 month | 29 (40.85) | 15 (36.59) | 0.84 |
| Comorbidities | |||
| Hypertension | 26 (36.62) | 12 (29.27) | 0.56 |
| Diabetes mellitus | 17 (23.94) | 9 (21.95) | 0.99 |
| Malignancy | 3 (4.23) | 4 (9.76) | 0.44 |
| Coronary artery disease | 29 (40.85) | 13 (31.71) | 0.47 |
| Stroke | 33 (46.48) | 15 (36.59) | 0.44 |
| Autoimmune encephalitis | 28 (39.44) | 12 (29.27) | 0.40 |
| Confirmed diagnosis | |||
| Autoimmune encephalitis | 19 (26.76) | 8 (19.51) | 0.49 |
| Intracranial infection | 29 (40.85) | 24 (58.54) | 0.08 |
| Cerebral infarction | 5 (7.04) | 3 (7.32) | >0.99 |
| Epilepsy | 9 (12.68) | 1 (2.44) | 0.09 |
| Metabolic encephalopathy | 2 (2.82) | 1 (2.44) | >0.99 |
| Cerebral hemorrhage | 4 (5.63) | 0 (0) | 0.30 |
| Trauma | 1 (1.41) | 0 (0) | >0.99 |
| Poisoning | 0 (0) | 1 (2.44) | 0.37 |
| Tumor | 1 (1.41) | 1 (2.44) | >0.99 |
| Peripheral neuropathy | 1 (1.41) | 2 (4.88) | 0.55 |
| Treatment | |||
| Hormone therapy | 21 (29.58) | 14 (34.15) | 0.67 |
| Anti-infective therapy | 33 (46.48) | 29 (70.73) | 0.02 |
| Hyperosmotic therapy | 11 (15.49) | 2 (4.88) | 0.13 |
| Sedatives, and/or analgesics | 23 (32.39) | 11 (26.83) | 0.07 |
Continuous data are presented as median (25th percentile, 75th percentile), and categorical data are presented as number (n) and percentage (%). Measured with cerebrospinal fluid pressure during lumbar puncture, with an ICP higher than the maximum limit of measurement of 330 cmH2O being recorded as 330 cmH2O. ICP, intracranial pressure.
In comparison with patients with normal ICP, those with intracranial hypertension were found to have a higher ONSD [median 5.23 (IQR 4.85–5.83) vs. 4.83 (IQR 3.54–5.17) mm; P<0.001], nONSD [median 2.62 (IQR 2.60–3.07) vs. 2.22 (IQR 1.85–2.40) mm; P<0.001], and ONSD/EBTD [median 0.23 (IQR 0.21–0.24) vs. 0.20 (IQR 0.15–0.23); P<0.001]; however, no statistical significance was detected for EBTD or OND in the between-group comparisons (Table 2).
Table 2
| Measurement result | Normal ICP (n=71) | Elevated ICP (n=41) | P value |
|---|---|---|---|
| ONSD (mm) | 4.83 (3.54, 5.17) | 5.23 (4.85, 5.83) | <0.001 |
| nONSD (mm) | 2.22 (1.85, 2.40) | 2.62 (2.60, 3.07) | <0.001 |
| OND (mm) | 2.62 (1.50, 2.86) | 2.62 (2.01, 3.01) | 0.50 |
| ONSD/EBTD | 0.20 (0.15, 0.23) | 0.23 (0.21, 0.24) | <0.001 |
| EBTD (mm) | 23.31 (22.56, 23.77) | 22.89 (22.31, 24.12) | 0.65 |
Data are presented as median (25th percentile, 75th percentile). EBTD, eyeball transverse diameter; ICP, intracranial pressure; nONSD, net optic nerve sheath diameter; OND, optic nerve diameter; ONSD, optic nerve sheath diameter.
Moderate significant linear correlations were found between ONSD and ICP (r=0.36; P<0.001; Figure 3A). However, no relationship was found between OND and ICP (r=0.04; P=0.65; Figure 3B). Moderate positive linear associations were also observed between nONSD and ICP (r=0.61, P<0.001; Figure 3C), and ONSD/EBTD and ICP (r=0.36; P<0.001; Figure 3D).
In predicting an ICP greater than 250 mmH2O, nONSD yielded a relatively high predictive accuracy for the general patient group; at an optimal cutoff value, the nONSD had a specificity of 88.7% and a positive predictive value (PPV) of 80%. nONSD also yielded the highest area under the curve (AUC) of 0.863 [95% confidence interval (CI): 0.785–0.921] for the whole patient group (Table 3 and Figure 4).
Table 3
| Parameters | Cutoff | AUC (95% CI) | Sensitivity | Specificity | PPV | NPV | P value |
|---|---|---|---|---|---|---|---|
| ONSD (mm) | 4.935 | 0.681 (0.586, 0.767) | 0.744 | 0.648 | 0.537 | 0.821 | 0.24 |
| OND (mm) | 1.580 | 0.500 (0.403, 0.596) | 0.846 | 0.310 | 0.402 | 0.786 | >0.99 |
| nONSD (mm) | 2.520 | 0.863 (0.785, 0.921) | 0.821 | 0.887 | 0.800 | 0.900 | <0.001 |
| ONSD/EBTD | 0.218 | 0.681 (0.586, 0.767) | 0.462 | 0.803 | 0.563 | 0.731 | 0.24 |
AUC, area under the curve; CI, confidence interval; EBTD, eyeball transverse diameter; ICP, intracranial pressure; nONSD, net optic nerve sheath diameter; NPV, negative predictive value; OND, optic nerve diameter; ONSD, optic nerve sheath diameter; PPV, positive predictive value.
Discussion
In this retrospective study of 112 patients with suspected intracranial infection, we investigated the diagnostic value of optic nerve-related MRI parameters for predicting elevated ICP. We found that patients with intracranial hypertension had significantly larger ONSD, nONSD, and ONSD/EBTD compared with patients with normal ICP. In contrast, OND and EBTD showed no significant differences. Among these parameters, nONSD demonstrated the strongest discriminative ability.
In addition to optic parameters, our study also observed a statistically significant difference in the use of anti-infective therapy between the two groups. Patients with normal ICP were more likely to receive anti-infective treatment, which may be explained by the higher proportion of central nervous system or other systemic infections in this group. By contrast, patients with elevated ICP were more often affected by non-infectious causes.
In patients with elevated ICP, the enlargement of the ONSD (representing the intraorbital subarachnoid space) is consistent with its underlying anatomical and physiological mechanisms. The intraorbital subarachnoid space is directly connected to the intracranial cavity, and elevated ICP leads to sheath expansion, as previously reported in both cadaveric and clinical studies (15-17). A study using ultrasound shear-wave elastography demonstrated that optic nerve stiffness, reflecting the mechanical response of the perioptic subarachnoid space, may provide complementary information to diameter-based indices for evaluating intracranial or orbital pathology (18). In our study cohort, the median ONSD was 4.83 mm in patients with normal ICP and 5.23 mm in those with ICP greater than 250 mmH2O. This result is consistent with previous research (19). However, individual anatomical variations, such as eyeball size, may influence ONSD measurements. To address this, we calculated the ONSD/EBTD ratio, which showed significant differences between groups and may help to adjust for individual variability, thus serving as a supplementary parameter for detecting raised ICP.
Correlation analysis further revealed a significant positive relationship between ICP and ONSD, nONSD, and the ONSD/EBTD ratio, whereas OND showed no correlation with ICP. Notably, nONSD demonstrated the strongest correlation (r=0.61, P<0.001), suggesting that subtracting OND from ONSD can reduce measurement bias and better capture sheath expansion. These findings are consistent with the pathophysiological expectation that elevated ICP primarily affects the sheath rather than the optic nerve itself.
In terms of diagnostic performance, nONSD demonstrated the highest predictive accuracy for ICP exceeding 250 mmH2O, with an AUC of 0.863, a sensitivity of 82.1%, and a NPV of 90.0%. These findings indicate that nONSD could serve as a sensitive and reliable screening tool for excluding intracranial hypertension. Compared with invasive monitoring or LP, MRI-based assessment of nONSD is safer and may be particularly valuable in patients for whom invasive procedures are unsuitable. Another practical advantage is that, unlike external ventricular drainage, this method does not require neurosurgical expertise and may therefore be especially useful in emergency settings or hospitals without neurosurgical services.
Several limitations must be acknowledged. First, this was a retrospective, single-center study. Although patients were grouped according to their ICP status, the absence of prospective randomization may have introduced bias. Second, patients with elevated ICP were generally older, and age-related bias could not be eliminated. Third, the MRI slice thickness was 5 mm, which may have reduced measurement accuracy compared with thinner slices (2–3 mm). Fourth, although other MRI features of elevated ICP (e.g., empty sella, posterior scleral flattening, and transverse sinus stenosis) were occasionally observed, these findings were inconsistent and were not systematically analyzed in this study. MRI features of intracranial infection were not analyzed for the same reason.
Conclusions
ONSD, nONSD, and ONSD/EBTD measured on cranial MRI were significantly associated with ICP elevation, with nONSD showing the strongest correlation and best predictive accuracy. These findings suggest that nONSD may serve as a valuable non-invasive biomarker for intracranial hypertension. Future prospective, multicenter studies incorporating a broader range of MRI features are warranted to confirm these results and extend their applicability across diverse clinical settings.
Acknowledgments
The authors thank all participants who kindly took part in this study.
Footnote
Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1671/rc
Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1671/dss
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1671/coif). The authors have no conflicts of interest to declare.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Beijing Tiantan Hospital, Capital Medical University (No. KY2021-163-02). Individual consent for this retrospective analysis was waived.
Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.
References
- Zoerle T, Beqiri E, Åkerlund CAI, Gao G, Heldt T, Hawryluk GWJ, Stocchetti N. Intracranial pressure monitoring in adult patients with traumatic brain injury: challenges and innovations. Lancet Neurol 2024;23:938-50. [Crossref] [PubMed]
- Duan S, Yuan Q, Wang M, Li R, Yuan H, Yao H, Hu J. Intracranial Pressure Monitoring in Patients with Spontaneous Intracerebral Hemorrhage: A Systematic Review with Meta-Analysis. World Neurosurg 2024;189:447-455.e4. [Crossref] [PubMed]
- Yuan Q, Wu X, Sun Y, Yu J, Li Z, Du Z, Mao Y, Zhou L, Hu J. Impact of intracranial pressure monitoring on mortality in patients with traumatic brain injury: a systematic review and meta-analysis. J Neurosurg 2015;122:574-87. [Crossref] [PubMed]
- Ramanan M, Lipman J, Shorr A, Shankar A. A meta-analysis of ventriculostomy-associated cerebrospinal fluid infections. BMC Infect Dis 2015;15:3. [Crossref] [PubMed]
- Bota DP, Lefranc F, Vilallobos HR, Brimioulle S, Vincent JL. Ventriculostomy-related infections in critically ill patients: a 6-year experience. J Neurosurg 2005;103:468-72. [Crossref] [PubMed]
- Lozier AP, Sciacca RR, Romagnoli MF, Connolly ES Jr. Ventriculostomy-related infections: a critical review of the literature. Neurosurgery 2002;51:170-81; discussion 181-2. [Crossref] [PubMed]
- Volovici V, Pisică D, Gravesteijn BY, Dirven CMF, Steyerberg EW, Ercole A, Stocchetti N, Nelson D, Menon DK, Citerio G, van der Jagt M, Maas AIR, Haitsma IK, Lingsma HFCENTER-TBI investigators. participants for the ICU stratum. Comparative effectiveness of intracranial hypertension management guided by ventricular versus intraparenchymal pressure monitoring: a CENTER-TBI study. Acta Neurochir (Wien) 2022;164:1693-705. [Crossref] [PubMed]
- Bratton SL, Chestnut RM, et al. Guidelines for the management of severe traumatic brain injury. VII. Intracranial pressure monitoring technology. J Neurotrauma 2007;24:S45-54. [Crossref] [PubMed]
- Geeraerts T, Newcombe VF, Coles JP, Abate MG, Perkes IE, Hutchinson PJ, Outtrim JG, Chatfield DA, Menon DK. Use of T2-weighted magnetic resonance imaging of the optic nerve sheath to detect raised intracranial pressure. Crit Care 2008;12:R114. [Crossref] [PubMed]
- Bäuerle J, Schuchardt F, Schroeder L, Egger K, Weigel M, Harloff A. Reproducibility and accuracy of optic nerve sheath diameter assessment using ultrasound compared to magnetic resonance imaging. BMC Neurol 2013;13:187. [Crossref] [PubMed]
- Kalantari H, Jaiswal R, Bruck I, Matari H, Ghobadi F, Weedon J, Hassen GW. Correlation of optic nerve sheath diameter measurements by computed tomography and magnetic resonance imaging. Am J Emerg Med 2013;31:1595-7. [Crossref] [PubMed]
- Kim DH, Jun JS, Kim R. Measurement of the Optic Nerve Sheath Diameter with Magnetic Resonance Imaging and Its Association with Eyeball Diameter in Healthy Adults. J Clin Neurol 2018;14:345-50. [Crossref] [PubMed]
- del Saz-Saucedo P, Redondo-González O, Mateu-Mateu Á, Huertas-Arroyo R, García-Ruiz R, Botia-Paniagua E. Sonographic assessment of the optic nerve sheath diameter in the diagnosis of idiopathic intracranial hypertension. J Neurol Sci 2016;361:122-7. [Crossref] [PubMed]
- Onder H, Goksungur G, Eliacik S, Ulusoy EK, Arslan G. The significance of ONSD, ONSD/ETD ratio, and other neuroimaging parameters in idiopathic intracranial hypertension. Neurol Res 2021;43:1098-106. [Crossref] [PubMed]
- Liu D, Kahn M. Measurement and relationship of subarachnoid pressure of the optic nerve to intracranial pressures in fresh cadavers. Am J Ophthalmol 1993;116:548-56. [Crossref] [PubMed]
- Hansen HC, Helmke K. The subarachnoid space surrounding the optic nerves. An ultrasound study of the optic nerve sheath. Surg Radiol Anat 1996;18:323-8. [Crossref] [PubMed]
- Hansen HC, Helmke K. Validation of the optic nerve sheath response to changing cerebrospinal fluid pressure: ultrasound findings during intrathecal infusion tests. J Neurosurg 1997;87:34-40. [Crossref] [PubMed]
- Chen SP, Dong Y, Ye L, Jiang YH, Jiang ZP, Xia YS. Increased optic nerve stiffness can aid in localizing ipsilateral brain or eye lesions in patients with secondary headaches. Quant Imaging Med Surg 2025;15:4608-15. [Crossref] [PubMed]
- Xu N, Zhu Q. Optic nerve sheath diameter measured by ultrasonography versus Magnetic Resonance Imaging for diagnosing increased intracranial pressure: a systematic review and meta-analysis. Med Ultrason 2023;25:270-8. [Crossref] [PubMed]
