BACKGROUND: Multimodal artificial intelligence (AI) models have recently expanded into video analysis. In ophthalmology, one exploratory application is the automated detection of extraocular movement (EOM) disorders. Thi...BACKGROUND: Multimodal artificial intelligence (AI) models have recently expanded into video analysis. In ophthalmology, one exploratory application is the automated detection of extraocular movement (EOM) disorders. This proof-of-concept study evaluated the feasibility of using Gemini 2.0 to recognize EOM abnormalities, identify the affected eye, and recognize specific movement limitations from publicly available, real-world clinical videos. METHODS: We retrospectively collected 114 YouTube videos of EOM disorders, including cranial nerve (CN) palsies, internuclear ophthalmoplegia (INO), supranuclear disorders, nystagmus, and ocular myasthenia gravis (MG), alongside 15 control videos demonstrating normal EOMs. Videos were trimmed to include only the pertinent clinical examinations, and audio was removed to avoid diagnostic cues. Using a standardized zero-shot prompt, Gemini 2.0 analyzed each video via the Google AI Studio platform. Gemini 2.0 was evaluated based on its ability to provide the correct diagnosis, identify the affected eye, and recognize the specific movement limitation (if any). Descriptive statistics, Spearman correlations, and comparative analyses were used to assess performance. RESULTS: Gemini 2.0 correctly identified the primary diagnosis in 43 of 114 videos, yielding an overall diagnostic accuracy of 37.7%. Diagnostic performance varied by condition, with the highest accuracies observed in third nerve palsy (81.1%), INO (80.0%), sixth nerve palsy (66.7%), and ocular MG (20.0%), whereas normal EOMs were correctly classified in 93.3% of cases. In misclassified cases, the correct diagnosis appeared in the differential diagnosis in 15.5% of instances. Laterality was correctly identified in 26.5% of eligible cases overall, 73.1% among correctly diagnosed cases vs. 9.6% in misclassified ones. Similarly, movement limitations were accurately identified in 30.3% of eligible cases overall, with a marked increase to 88.5% accuracy in correctly diagnosed cases compared to 9.6% in misclassified cases. Longer videos moderately correlated with longer processing time (ρ = 0.55, P < 0.001). Significant correlations were observed between correct diagnosis and correct laterality identification (ρ = 0.45, P < 0.001), correct diagnosis and correct movement limitation identification (ρ = 0.61, P < 0.001), and laterality and movement limitation (ρ = 0.51, P < 0.001). Processing time averaged 11.0 seconds and correlated with video length (ρ = 0.55, P < 0.001). CONCLUSIONS: This proof-of-concept study demonstrates the feasibility of using Gemini 2.0 for automated recognition of EOM abnormalities in clinical videos. Although performance was stronger in overt cases, overall diagnostic accuracy remains limited. Substantial validation on standardized, clinician-annotated datasets is needed before clinical application.
BACKGROUND: Papilledema and other optic neuropathies are critical findings in neuro-ophthalmology that require timely and accurate diagnosis. This study evaluates the performance of a deep learning system (DLS) to identi...BACKGROUND: Papilledema and other optic neuropathies are critical findings in neuro-ophthalmology that require timely and accurate diagnosis. This study evaluates the performance of a deep learning system (DLS) to identify papilledema and other optic neuropathies, when applied to a large dataset of retinal photographs acquired prospectively with a handheld nonmydriatic camera in a neuro-ophthalmology department. METHODS: International multiethnic, multicenter study including 20,533 retinal photographs (10,647 patients) from 31 centers worldwide. The training and internal validation datasets consisted of 18,981 mydriatic retinal photographs (9,830 patients) from 22 countries. The external testing dataset included 1,552 prospectively collected retinal photographs (817 patients) acquired with a handheld camera (Aurora, Optomed, Finland). The DLS segmented the optic disc and peripapillary region, then classified each photograph as 1/"papilledema," 2/"other" optic disc abnormalities (i.e., swelling because of other causes, atrophy, etc.), or 3/"normal." The performance of the DLS in classifying optic disc appearance was evaluated using area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and accuracy. Diagnostic outcomes were evaluated at the eye level and at the patient level. RESULTS: The DLS achieved an accuracy of 99.5% (95% confidence interval [CI], 99.1-99.8), a sensitivity of 81.0% (95% CI, 64.1-97.7), a specificity of 99.7% (95% CI, 99.5-99.9), and an AUC of 98.3% (95% CI, 97.6-99.9) for differentiating papilledema from "others" and healthy controls. CONCLUSIONS: A DLS trained on a large dataset of mydriatic photographs achieved excellent diagnostic performance for detection of papilledema and other optic disc abnormalities when applied to nonmydriatic retinal photographs acquired with a handheld camera in real life conditions.
BACKGROUND: There is a global shortage of neuro-ophthalmologists, driven by an aging workforce and declining interest among residents. Emerging data indicate that this supply-demand mismatch results in prolonged patient...BACKGROUND: There is a global shortage of neuro-ophthalmologists, driven by an aging workforce and declining interest among residents. Emerging data indicate that this supply-demand mismatch results in prolonged patient wait times, leading to evaluation by additional physicians, delayed/incorrect diagnosis, and postponed therapy. However, data on referral patterns are primarily from resource-rich countries. This study examines referral patterns for a neuro-ophthalmology clinic in Bogota, Colombia. METHODS: A retrospective chart review analyzed patients referred to the clinic between July 2020 and July 2021. Data included demographics, time from symptom onset to neuro-ophthalmology appointment (NOA), referral-to-NOA time, initial specialty consulted, number of providers seen before NOA, diagnostic tests performed, and diagnoses at referral vs after NOA. RESULTS: Among 535 patients (62.1% female), the median wait from symptom onset to NOA was 365 days (interquartile range [IQR]: 60-500) and 30 days ([IQR: 10-45]) from referral to NOA. Most patients (80.2%) first saw an ophthalmologist, with a mean of 1.1 providers seen before NOA. Common diagnostic tests before NOA included brain MRIs (28.6%), posterior segment optical coherence tomography (24.3%), and visual fields (22.8%). Three hundred twenty-six (63.9%) of patients were either partially or completely misdiagnosed before NOA. Time from symptom onset to NOA was a predictor for fully correct diagnosis by referring provider ( P < 0.01), with a shorter time corresponding to a greater likelihood of partially correct diagnosis. CONCLUSIONS: Referral delays are frequent, and misdiagnoses before neuro-ophthalmology consultations are common. Enhancing timely access to NOA could significantly improve patient outcomes and reduce unnecessary utilization of health care resources.
Photon-counting detector CT is an emerging technology that has the promise to dramatically transform clinical practice. It largely overcomes the limitations of traditional CT, offering ultra-high spatial resolution, supp...Photon-counting detector CT is an emerging technology that has the promise to dramatically transform clinical practice. It largely overcomes the limitations of traditional CT, offering ultra-high spatial resolution, suppression of metallic artifact, and reduced radiation dose. We report the case of a 44-year-old patient presenting with gaze-evoked vision loss and history of a retained orbital foreign body. Localization of the foreign body on conventional CT was limited by metallic artifacts, and magnetic resonance imaging was contraindicated. Photon-counting detector CT effectively localized the foreign body and guided management. The potential applications of this newer technology in neuro-ophthalmology are still being explored.
Leinonen J, Mikkola R, Peltonen K
… +2 more, Hokkanen L, Laitala T
J Neuroophthalmol
· 2024 Dec · PMID 40833794
·
Full text
BACKGROUND: Concussions are mild traumatic brain injuries that often cause vision problems. They have significant impacts on everyday life, cognitive capacity, and sports performance, and may affect injury prevalence in...BACKGROUND: Concussions are mild traumatic brain injuries that often cause vision problems. They have significant impacts on everyday life, cognitive capacity, and sports performance, and may affect injury prevalence in fast contact sports such as ice hockey. A functional vision questionnaire specifically designed for sports was used here to study the correlation between vision problems and concussion history. METHODS: In this national cross-sectional concussion study, 860 Finnish elite-level male adolescent ice hockey players (aged 13-21 years) answered a functional vision questionnaire and performed a computerized neurocognitive test, ImPACT. Totally 265 athletes reported a history of at least 1 concussion. All data were statistically compared with age-matched athletes with no concussion history (n = 595). For further analysis, athletes were divided into subgroups by age and number of previous concussions. RESULTS: Previously concussed athletes reported more general and eye-specific symptoms than their healthy controls. Increases in eye fatigue, frontal headaches, and blinking were statistically significant. Also statistically more problems with depth perception and evaluating distances, concentration problems, blurred vision, and losing the object in sight were observed among athletes with concussion history. CONCLUSIONS: Concussion history reflects an increase in the prevalence of vision deficits, as determined by multiple disturbances in the near triad. The significant number of vision problems in the concussion history groups strongly suggests that functional vision should routinely be evaluated in athletes. The vision problems observed in the athletes with concussion history may indicate an increased injury risk that should be addressed.
Disse LR, Bockisch CJ, Weber KP
… +1 more, Fierz FC
J Neuroophthalmol
· 2024 Nov · PMID 40833793
·
Full text
BACKGROUND: The differentiation of Horner syndrome from physiological anisocoria is important yet clinically challenging. We investigated the diagnostic accuracy of pupillometry to discriminate Horner syndrome from physi...BACKGROUND: The differentiation of Horner syndrome from physiological anisocoria is important yet clinically challenging. We investigated the diagnostic accuracy of pupillometry to discriminate Horner syndrome from physiological anisocoria compared to pharmacological testing with the alpha-2-agonist apraclonidine, which is considered the current gold standard. METHODS: Forty-four adult patients, mostly referred to our neuro-ophthalmology service for evaluation of anisocoria, were included. Automated binocular pupillometry was performed under standardized light conditions before and >30 minutes after instillation of 1% apraclonidine eye drops. A positive apraclonidine test indicating unilateral Horner syndrome was defined as an increase of pupil size in the smaller pupil and decrease of size in the larger pupil. Receiver operator characteristic curves were calculated to find the best pupillometric parameter discriminating Horner syndrome from physiological anisocoria. RESULTS: We found that the parameters measuring the pupillary dilation lag using pupillometry could reliably discriminate Horner syndrome from physiological anisocoria compared to pharmacological testing. Calculating the change of anisocoria at 3-4 seconds after light-off relative to the anisocoria at the end of the light-on period (Δ3-4) may be most suitable to rule out Horner syndrome reaching a sensitivity of 95% and specificity of 68% using a cutoff of 0.35 mm. CONCLUSIONS: Our results indicate that pupillometry is a robust tool to measure the dilation lag in Horner syndrome and, therefore, to distinguish pathological from physiological anisocoria obviating pharmacological testing. The high sensitivity of the test will allow to identify the patients with Horner syndrome requiring further investigation.
Madadi Y, Delsoz M, Lao PA
… +4 more, Fong JW, Hollingsworth TJ, Kahook MY, Yousefi S
J Neuroophthalmol
· 2024 Oct · PMID 40830998
·
Full text
BACKGROUND: To evaluate the accuracy of Chat Generative Pre-Trained Transformer (ChatGPT), a large language model (LLM), to assist in diagnosing neuro-ophthalmic diseases based on case reports. METHODS: We selected 22 di...BACKGROUND: To evaluate the accuracy of Chat Generative Pre-Trained Transformer (ChatGPT), a large language model (LLM), to assist in diagnosing neuro-ophthalmic diseases based on case reports. METHODS: We selected 22 different case reports of neuro-ophthalmic disorders from a publicly available online database. These cases included a wide range of chronic and acute diseases commonly seen by neuro-ophthalmologists. We inserted each case as a new prompt into ChatGPTs (GPT-3.5 and GPT-4) and asked for the most likely diagnosis. We then presented the exact information to 2 neuro-ophthalmologists and recorded their diagnoses, followed by comparing responses from both versions of ChatGPT. RESULTS: GPT-3.5 and GPT-4 and the 2 neuro-ophthalmologists were correct in 13 (59%), 18 (82%), 19 (86%), and 19 (86%) out of 22 cases, respectively. The agreements between the various diagnostic sources were as follows: GPT-3.5 and GPT-4, 13 (59%); GPT-3.5 and the first neuro-ophthalmologist, 12 (55%); GPT-3.5 and the second neuro-ophthalmologist, 12 (55%); GPT-4 and the first neuro-ophthalmologist, 17 (77%); GPT-4 and the second neuro-ophthalmologist, 16 (73%); and first and second neuro-ophthalmologists 17 (77%). CONCLUSIONS: The accuracy of GPT-3.5 and GPT-4 in diagnosing patients with neuro-ophthalmic disorders was 59% and 82%, respectively. With further development, GPT-4 may have the potential to be used in clinical care settings to assist clinicians in providing accurate diagnoses. The applicability of using LLMs like ChatGPT in clinical settings that lack access to subspeciality trained neuro-ophthalmologists deserves further research.
BACKGROUND: We define uveitic optic disc edema as disc edema that is partly or solely associated with uveitis. Our study describes the clinical and imaging characteristics of patients with UDE evaluated at the University...BACKGROUND: We define uveitic optic disc edema as disc edema that is partly or solely associated with uveitis. Our study describes the clinical and imaging characteristics of patients with UDE evaluated at the University of Minnesota. METHODS: We retrospectively reviewed medical records of patients with UDE seen by a single uveitis provider for a 3-year period. Inclusion criteria were (1) the presence of uveitis and optic disc edema in one or both eyes and (2) optical coherence tomography (OCT) optic disc raster and retinal nerve fiber layer (RNFL) thickness measurements obtained within 2 weeks of one other. Disc raster OCT scans were analyzed to determine retinal height at the disc, focus of thickening, and retinal reflectance. Automated visual field (VF) testing and fluorescein angiography (FA) images were reviewed when available. FA pixel intensity was used to quantify disc fluorescence. Brain MRI scans were reviewed when available. RESULTS: Fifty-five eyes from 31 patients were analyzed. Patients' ages ranged from 11 to 78 years. Uveitis was present in all anatomic compartments, including retinal vasculitis and choroiditis. A total of 24 patients (77.4%) presented with unilateral disc edema and 7 patients (22.6%) had bilateral disc edema. VF testing was organized into 7 descriptive categories based on severity: normal, scattered/nonspecific defects, blind spot enlargement, central/paracentral defects, nasal/arcuate defects, mixed defects, and generalized depression. Each eye was assigned a primary VF defect type with an associated severity score. Overall, 12.7% of eyes had no/minor VF defects, 40.0% had focal VF defects, and 47.3% had severe VF defects. The average RNFL thickness for all eyes was 149 μm. A statistically significant positive correlation was found between the severity of VF defects and RNFL thickness when the entire group was analyzed (P = 0.042). Structural optic disc raster OCT scans showed no focal thickening (7.3%), isolated nerve fiber layer thickening (5.5%), focal inner-middle thickening (32.7%), and diffuse retinal thickening (54.5%). Disc fluorescence on FA showed a statistically significant positive correlation with maximum disc height (P = 0.0177), but did not correlate with mean reflectance on OCT. We did not detect a relationship between OCT reflectance and maximum disc height. Twenty-nine of 31 patients underwent brain MRI and 5 of these patients with bilateral disc edema showed radiographic features, suggestive of elevated intracranial pressure (ICP). Only 4 of 31 patients had elevated opening pressure of greater than 25 cm H2O by lumbar puncture. CONCLUSIONS: UDE as a distinct clinical entity has not been well categorized in the literature. A multimodal imaging approach including OCT RNFL, OCT disc raster scan, VF testing, and FA can aid in diagnosis of UDE. OCT disc raster height may be used as a surrogate for FA intensity and may be a useful adjunctive modality for monitoring UDE severity along with serial OCT scans. Increased intracranial pressure was rare in our patient cohort so neuroimaging should not be obtained solely based on optic disc appearance and imaging abnormalities.
BACKGROUND: We investigated the initial macular ganglion cell complex (GCC) changes in a group of patients during the conversion from normal vision to visual loss from Leber hereditary optic neuropathy (LHON). METHODS: O...BACKGROUND: We investigated the initial macular ganglion cell complex (GCC) changes in a group of patients during the conversion from normal vision to visual loss from Leber hereditary optic neuropathy (LHON). METHODS: Optical coherence tomography scans from 17 eyes of 10 patients with genetically confirmed LHON obtained within 12 weeks of visual symptom onset were analyzed. The thickness of the GCC was measured in 6 macular sectors: superior nasal (SN), inferior nasal (IN), superior temporal, inferior temporal (IT), superior (SUP), and inferior (INF). Receiver operating characteristic (ROC) curves and area under the ROC curve (AUROC) analyses with DeLong comparison were used to evaluate the predictive reliability of GCC thinning in each sector. RESULTS: Selective, nasal (perifoveal) GCC thinning was observed in most LHON eyes during the early stages, with the SN and IN sectors showing the highest AUROC values. Thinning in these sectors provided significantly greater predictive value for the presence of LHON than other sectors. CONCLUSIONS: Nasal GCC thinning is a reliable early indicator of the conversion from normal vision to visual loss in LHON as demonstrated in 10 of 17 eyes. This pattern of GCC loss provides insights into the disease mechanism and highlights the utility of OCT analysis for early diagnosis. The nasal GCC loss suggests selective vulnerability of ganglion cells serving the foveal region in LHON. Early identification of nasal GCC changes may facilitate timely intervention as gene therapy and other treatments become increasingly accessible.
BACKGROUND: Neuromyelitis optica spectrum disorders (NMOSD) and myelin oligodendrocyte glycoprotein antibody disease (MOGAD) are autoimmune inflammatory disorders of the central nervous system. Central serous chorioretin...BACKGROUND: Neuromyelitis optica spectrum disorders (NMOSD) and myelin oligodendrocyte glycoprotein antibody disease (MOGAD) are autoimmune inflammatory disorders of the central nervous system. Central serous chorioretinopathy (CSCR) is characterized by a serous retinal detachment with leakage of fluid through the retinal pigment epithelium accumulating under the retina. This study investigated a potential association between CSCR and these neuroinflammatory disorders. METHODS: We included people with aquaporin-4 immunoglobulin G (AQP4-IgG) seropositive NMOSD (N = 39), multiple sclerosis (MS, N = 39), myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD, N = 13), seronegative NMOSD (SN-NMOSD, N = 9), and healthy controls (HC, N = 30). Using optical coherence tomography (OCT), we assessed CSCR frequency and the thickness of the peripapillary retinal nerve fiber layer (pRNFL). RESULTS: There was a higher CSCR frequency (21.3%) throughout all investigated subgroups (AQP4-IgG seropositive NMOSD, MOGAD, and SN-NMOSD) than in the HC group (p = 0.048), with a significant association between CSCR and arterial hypertension frequency but not with these diagnoses, retinal neuroaxonal loss, or history of optic neuritis. CONCLUSION: The high frequency of CSCR suggests a potential benefit of routine monitoring of CSCR in patients with NMOSD and MOGAD using the OCT technology, a reliable method to detect and monitor CSCR in patients with neuroinflammatory disorders. Further research is necessary to establish the underlying pathophysiology and potential effects on vision.
In this issue of JNO , Drs. Deborah I. Friedman and Mark L. Moster discuss the following 4 articles: Shir D, Lee N, McCarter SJ, Ramanan VK, Botha H, Knopman DS, Petersen RC, Boeve BF, Day GS, Graff-Radford NR, Jones DT,...In this issue of JNO , Drs. Deborah I. Friedman and Mark L. Moster discuss the following 4 articles: Shir D, Lee N, McCarter SJ, Ramanan VK, Botha H, Knopman DS, Petersen RC, Boeve BF, Day GS, Graff-Radford NR, Jones DT, Murray ME, Nguyen AT, Reichard RR, Dickson DW, Tajfirouz D, Machulda MM, Whitwell JL, Josephs KA, Graff-Radford J. Longitudinal evolution of posterior cortical atrophy: diagnostic delays, overlapping phenotypes, and clinical outcomes. Neurology . 2025;104:e213559. Fariselli L, Marzoli BS, Morlino S, Pinzi V, Pedone C, De Martin E, Romeo A, Tramacere I, Marchetti M. Hypofractionated radiosurgery (25 Gy/5 fractions) for optic nerve sheath meningiomas: results from an exploratory phase 2 prospective trial. Int J Radiat Oncol Biol Phys . 2025:S0360-3016(25)00432-8 (epub ahead of print). Dinoto A, Cacciaguerra L, Vorasoot N, Redenbaugh V, Lopez-Chiriboga SA, Valencia-Sanchez C, Guo K, Thakolwiboon S, Horsman SE, Syc-Mazurek SB, Tisavipat N, Mandrekar J, Tajfirouz D, Eggenberger ER, Pless ML, Chodnicki K, Tillema JM, Pittock SJ, Chen JJ, Flanagan EP. Clinical features and factors associated with outcome in late adult-onset myelin oligodendrocyte glycoprotein antibody-associated disease. Neurology . 27;104:e213557. Park HE, Shin HJ, Lee AG. Neuro-ophthalmic findings of visual snow syndrome in Korea. Jpn J Ophthalmol . 2025. doi: 10.1007/s10384-025-01196-1 (epub ahead of print).
Huang JJ, Lovasz D, Shi M
… +3 more, Guillot FH, Davis JB, Henderson AD
J Neuroophthalmol
· 2025 Dec · PMID 40767867
·
Full text
BACKGROUND: Idiopathic intracranial hypertension (IIH) disproportionately affects young, Black, female patients. However, there are limited data characterizing baseline and outcome features of IIH across racial and socio...BACKGROUND: Idiopathic intracranial hypertension (IIH) disproportionately affects young, Black, female patients. However, there are limited data characterizing baseline and outcome features of IIH across racial and socioeconomic groups. This study aimed to compare presenting features, initial MRI findings, and visual outcomes of IIH from different racial, ethnic, and insured groups. METHODS: We conducted a retrospective chart review of patients with IIH seen at an academic quaternary care center between July 1, 2013, and September 30, 2023. Patient characteristics collected were self-reported race and ethnicity, insurance status, age, and sex. Clinical features included lumbar puncture (LP) opening pressure; body mass index (BMI); visual acuity (VA), visual field mean deviation, and average retinal nerve fiber layer (RNFL) thickness at presentation and follow up; MRI findings; presence of symptoms (headaches, diplopia, transient visual obscurations, pulsatile tinnitus), treatment(s); and adherence to follow-up recommendations. Patients were excluded if they were already treated for IIH before initial presentation or if diagnostic data were incomplete. Linear, mixed effect, and logistic regression models were used to evaluate differences between racial, ethnic, and insurance status groups. Statistical significance was set at P < 0.05. RESULTS: Our study included 241 patients. There were no statistically significant differences in BMI, presence of symptoms of high intracranial pressure, visual function, or RNFL thickness at initial evaluation between groups stratified by race, ethnicity, and insurance status. Black patients were more likely to have Medicaid coverage, whereas White patients were more likely to have private insurance (Medicaid: 38% Black vs 14% White; private: 47% Black vs 67% White, P < 0.001). Black patients were more likely to have posterior globe flattening on initial MRI, and patients with Medicare coverage had significantly worse VA outcomes than those with Medicaid coverage. There were no other statistically significant differences in presenting features, clinical outcomes, or adherence to follow-up recommendations across patient subgroups. CONCLUSIONS: There were no statistically significant differences in presenting symptoms of IIH, BMI, LP opening pressure, or adherence to follow-up based on race, ethnicity, or insurance status. Patients with Medicare coverage had worse VA outcomes than those with Medicaid coverage. Otherwise, visual outcomes were similar across groups.
BACKGROUND: Ocular myasthenia gravis (OMG) is an autoimmune disease characterized by autoantibodies targeting postsynaptic proteins at the neuromuscular junction, leading to weakness and fatigability of the levator palpe...BACKGROUND: Ocular myasthenia gravis (OMG) is an autoimmune disease characterized by autoantibodies targeting postsynaptic proteins at the neuromuscular junction, leading to weakness and fatigability of the levator palpebrae superioris, orbicularis oculi and extraocular muscles. Although OMG is primarily a clinical diagnosis, serological antibody testing, predominantly acetylcholine receptor (AChR) antibodies, is usually performed. The clinical utility of muscle-specific kinase (MuSK) antibodies is less well established in OMG. This meta-analysis evaluates the use of anti-AChR and anti-MuSK in patients with OMG and the relative costs of simultaneous vs sequential testing. METHODS: Studies were extracted from Cumulative Index to Nursing and Allied Health Literature (CINAHL), Embase (Ovid), Medline (Ovid), and additional gray literature. A systematic review was conducted using Covidence with 2 independent reviewers for study selection and data extraction. The meta-analysis was conducted with R version 4.4.1 on RStudio, and the meta package. Depending on the level of heterogeneity, either a fixed-effects or random-effects model was used to pool the data. Funnel plots were used to assess publication bias. RESULTS: The pooled analysis of 44 studies (n = 4,937 patients with OMG) revealed 59% (95% confidence interval [CI]: 52%-66%) positivity for anti-AChR, whereas the pooled analysis of 34 studies with (n = 3,380) showed 5% (95% CI: 2%-9%) positivity for anti-MuSK. From 62 studies (n = 5,180), 4 patients (0.1%) were doubly seropositive for anti-AChR and anti-MuSK. In patients with OMG positive for AChR antibodies, 5 studies (n = 527) reported a thymoma prevalence of 35% (95% CI: 3%-90%), underscoring the clinical value of anti-AChR testing. Four studies (n= 259) showed that anti-AChR positive patients had a 1.82 (95% CI: 1.15-2.88) times greater risk of progressing from OMG to generalized myasthenia gravis. CONCLUSIONS: Almost two-thirds (59%) of the patients with OMG tested positive for AChR antibodies, but MuSK antibodies were only detected in 5% of patients. Positivity for anti-AChR in OMG was associated with a worse prognosis, including a higher prevalence of thymomas and an increased risk of disease generalization. Given the relatively low prevalence of anti-MuSK and the higher cost of anti-MuSK testing, clinicians could consider a stepwise approach to the serological diagnosis of OMG, where anti-MuSK is ordered only if the initial anti-AChR returns negative.