Automated translation accurately translates recorded pediatric neurosurgery clinic conversations between Spanish and English.

OBJECTIVE: The purpose of this study is to analyze an automated voice to text translation device by reporting the translation accuracy for recorded pediatric neurosurgery clinic conversations, classifying errors in translation according to their impact on overall understanding, and comparing the incidence of these errors in English to Spanish vs. Spanish to English conversations. METHODS: English and Spanish speaking patients at a single academic health system's outpatient pediatric neurosurgery clinic had their conversations recorded. These recordings were played back to a Google Pixel handheld smartphone with Live Translate voice to text translation software. A certified medical interpreter evaluated recordings for incidence of minor errors, errors impacting understanding, and catastrophic errors affecting patient-provider relationship or care. Two proportion t-testing was used to compare these outcomes. RESULTS: 50 patient visits were recorded: 40 English recordings translated to Spanish and 10 Spanish recordings translated to English. The mean transcript length was 4244 ± 992 words. The overall accuracy was 98.2% ± 0.5%. On average, 46 words were missed in translation (1.09% error rate), 31 understanding-altering translation errors (0.73% error rate), and 0 catastrophic errors were made. There was no significant difference in English to Spanish or vice versa. CONCLUSION: Voice to text translation devices using automatic speech recognition accurately translate recorded clinic conversations between Spanish and English with high accuracy and low incidence of errors impacting medical care or understanding. Further study should investigate additional languages, assess patient preferences and potential concerns with respect to device use, and compare these devices directly to medical interpreters in live clinic settings.

Low surgical weight associated with ETV failure in pediatric hydrocephalus patients.

The use of endoscopic third ventriculostomy (ETV) for treatment of pediatric hydrocephalus has higher failure rates in younger patients. Here we investigate the impact of select perioperative variables, specifically gestational age, chronological age, birth weight, and surgical weight, on ETV failure rates. A retrospective review was performed on patients treated with ETV - with or without choroid plexus cauterization (CPC) - from 2010 to 2021 at a large academic center. Analyses included Cox regression for independent predictors and Kaplan-Meier survival curves for time to-event outcomes. In total, 47 patients were treated with ETV; of these, 31 received adjunctive CPC. Overall, 66% of the cohort experienced ETV failure with a median failure of 36 days postoperatively. Patients aged < 6 months at time of surgery experienced 80% failure rate, and those > 6 months at time of surgery experienced a 41% failure rate. Univariate Cox regression analysis showed weight at the time of ETV surgery was significantly inversely associated with ETV failure with a hazard ratio of 0.92 (95% CI 0.82, 0.99). Multivariate analysis redemonstrated the inverse association of weight at time of surgery with ETV failure with hazard ratio of 0.76 (95% CI 0.60, 0.92), and sensitivity analysis showed < 4.9 kg as the optimal cutoff predicting ETV/CPC failure. Neither chronologic age nor gestational age were found to be significantly associated with ETV failure.In this study, younger patients experienced higher ETV failure rates, but multivariate regression found that weight was a more robust predictor of ETV failure than chronologic age or gestational age, with an optimal cutoff of 4.9 kg in our small cohort. Given the limited sample size, further study is needed to elucidate the independent role of weight as a peri-operative variable in determining ETV candidacy in young infants. Previous presentations: Poster Presentation, Congress of Neurological Surgeons.

Noninvasive Treatment of Venous Pulsatile Tinnitus with an Internal Jugular Vein Compression Collar.

OBJECTIVE: The study was conducted to evaluate the safety and efficacy of mild internal jugular (IJV) compression via an FDA approved compression collar for symptomatic treatment of venous pulsatile tinnitus. METHODS: This is a prospective study that recruited 20 adult patients with venous pulsatile tinnitus. Participants completed the Tinnitus Handicap Inventory (THI), were fitted with the collar, and rated symptom intensity on a 10-point tinnitus intensity scale before and during collar use. Once weekly for 4 weeks, they answered a survey quantifying days used, average tinnitus intensity before and after wearing the collar each day of use, and any safety concerns. Lastly, they completed an exit interview. The primary outcome was symptomatic relief, with secondary outcomes of safety, effect of treatment setting, effect of time, and quality of life assessed via nonparametric testing. RESULTS: 18 participants completed the study, and 276 paired daily before use/during use intensity scores were submitted. The median symptom intensity without the collar was 6 (IQR 4, 7), whereas with the collar it was 3 (IQR 2, 5), for a median symptomatic relief of 50%. The collar had a significant effect in reducing symptom intensity (p < 0.0001) and burden of illness via the THI (p < 0.0001). There was no effect of setting, frequency, or time on symptomatic relief with the collar. There were no adverse safety events reported aside from minor discomfort upon initial application. CONCLUSIONS: Venous compression collars offer acute symptom relief for patients with venous pulsatile tinnitus. Further study is needed to assess safety and efficacy of longitudinal use. LEVEL OF EVIDENCE: 4 Laryngoscope, 2024.

Machine learning for automated and real-time two-dimensional to three-dimensional registration of the spine using a single radiograph.

OBJECTIVE: The goal of this work was to methodically evaluate, optimize, and validate a self-supervised machine learning algorithm capable of real-time automatic registration and fluoroscopic localization of the spine using a single radiograph or fluoroscopic frame. METHODS: The authors propose a two-dimensional to three-dimensional (2D-3D) registration algorithm that maximizes an image similarity metric between radiographic images to identify the position of a C-arm relative to a 3D volume. This work utilizes digitally reconstructed radiographs (DRRs), which are synthetic radiographic images generated by simulating the x-ray projections as they would pass through a CT volume. To evaluate the algorithm, the authors used cone-beam CT data for 127 patients obtained from an open-source de-identified registry of cervical, thoracic, and lumbar scans. They systematically evaluated and tuned the algorithm, then quantified the convergence rate of the model by simulating C-arm registrations with 80 randomly simulated DRRs for each CT volume. The endpoints of this study were time to convergence, accuracy of convergence for each of the C-arm's degrees of freedom, and overall registration accuracy based on a voxel-by-voxel measurement. RESULTS: A total of 10,160 unique radiographic images were simulated from 127 CT scans. The algorithm successfully converged to the correct solution 82% of the time with an average of 1.96 seconds of computation. The radiographic images for which the algorithm converged to the solution demonstrated 99.9% registration accuracy despite utilizing only single-precision computation for speed. The algorithm was found to be optimized for convergence when the search space was limited to a ± 45° offset in the right anterior oblique/left anterior oblique, cranial/caudal, and receiver rotation angles with the radiographic isocenter contained within 8000 cm3 of the volumetric center of the CT volume. CONCLUSIONS: The investigated machine learning algorithm has the potential to aid surgeons in level localization, surgical planning, and intraoperative navigation through a completely automated 2D-3D registration process. Future work will focus on algorithmic optimizations to improve the convergence rate and speed profile.

Day one postoperative MRI findings following electrode placement for deep brain stimulation: analysis of a large case series.

Objective: This study sought to characterize postoperative day one MRI findings in deep brain stimulation (DBS) patients. Methods: DBS patients were identified by CPT and had their reviewed by a trained neuroradiologist and neurosurgeon blinded to MR sequence and patient information. The radiographic abnormalities of interest were track microhemorrhage, pneumocephalus, hematomas, and edema, and the occurrence of these findings in compare the detection of these complications between T1/T2 gradient-echo (GRE) and T1/T2 fluid-attenuated inversion recovery (FLAIR) magnetic resonance (MR) sequences was compared. The presence, size, and association of susceptibility artifact with other radiographic abnormalities was also described. Lastly, the association of multiple microelectrode cannula passes with each radiographic finding was evaluated. Ad-hoc investigation evaluated hemisphere-specific associations. Multiple logistic regression with Bonferroni correction (corrected p = 0.006) was used for all analysis. Results: Out of 198 DBS patients reviewed, 115 (58%) patients showed entry microhemorrhage; 77 (39%) track microhemorrhage; 44 (22%) edema; 69 (35%) pneumocephalus; and 12 (6%) intracranial hematoma. T2 GRE was better for detecting microhemorrhage (OR = 14.82, p < 0.0001 for entry site and OR = 4.03, p < 0.0001 for track) and pneumocephalus (OR = 11.86, p < 0.0001), while T2 FLAIR was better at detecting edema (OR = 123.6, p < 0.0001). The relatively common findings of microhemorrhage and edema were best visualized by T2 GRE and T2 FLAIR sequences, respectively. More passes intraoperatively was associated with detection of ipsilateral track microhemorrhage (OR = 7.151, p < 0.0001 left; OR = 8.953, p < 0.0001 right). Susceptibility artifact surrounding electrodes possibly interfered with further detection of ipsilateral edema (OR = 4.323, p = 0.0025 left hemisphere only). Discussion: Day one postoperative magnetic resonance imaging (MRI) for DBS patients can be used to detect numerous radiographic abnormalities not identifiable on a computed tomographic (CT) scan. For this cohort, multiple stimulating cannula passes intraoperatively was associated with increased microhemorrhage along the electrode track. Further studies should be performed to evaluate the clinical relevance of these observations.