Surgical treatment for ASD is complex and associated with significant perioperative risks. Despite advancements in surgical techniques and perioperative care, complications remain common. This study aims to develop a predictive model of 30-day mortality in patients undergoing surgery for ASD using data from the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) from 2014 to 2022. Patients included were aged ≥ 18 years undergoing ASD surgeries were included. Major complications were defined as deep surgical site infection (SSI), organ space SSI, wound dehiscence, prolonged ventilation ≥ 48 h, pulmonary embolism, cerebrovascular accident, renal failure, myocardial infarction, cardiac arrest, bleeding requiring transfusion, sepsis, septic shock, pneumonia, and unplanned reintubation. The primary outcome of interest was 30-day mortality. Categorical variables were analyzed using the Pearson χ2 test, and continuous variables were analyzed using the Mann–Whitney U test. The study cohort included 2471 adults. The overall mortality rate within 30 days of 0.5

OBJECTIVE:Cervical disc replacement (CDR) is an emerging alternative to arthrodesis for the treatment of degenerative cervical disc disease. The current surgical standard of care, anterior cervical discectomy and fusion (ACDF), allows for adequate decompression with good outcomes; however, this alters biomechanical stability and can eventually lead to adjacent segment disease and reoperation. By preserving motion at the pathological level, CDR might reduce the development of adjacent segment disease and the associated morbidity of reoperation. The aim of this study was to evaluate long-term functional outcomes with CDR compared with ACDF using a large federated electronic health record network. METHODS:The TriNetX database was queried for all patients ≥ 18 years of age who underwent CDR or ACDF for degenerative indications. Propensity score matching based on demographic and clinical characteristics was performed using a 1:1 ratio. Cohorts were evaluated by calculating relative risk of functional and clinical outcomes at multiple time points from 6 months to 5 years. Kaplan-Meier survival curves were used to examine reoperation and adjacent segment disease. RESULTS:After matching, 9458 patients (mean age 48 years) were included in both the ACDF and CDR groups. At all follow-up time points, there was a significantly lower risk of anterior reoperation (RR 0.62, 95% CI 0.52-0.74; p < 0.001) and new-onset cervical pain (RR 0.64, 95% CI 0.55-0.73; p < 0.001) among patients who underwent CDR compared with those who underwent ACDF. At the 1-, 2-, and 5-year follow-ups, there was significantly lower risk of surgical site infection (RR 0.62, 95% CI 0.48-0.78; p < 0.001) in the CDR group. After 2 years, there was no difference in codes for acquired kyphosis or lordosis of the cervical spine between the two groups (p > 0.05), and after 5 years, there was significantly lower risk of opioid use (RR 0.66, 95% CI 0.54-0.82; p < 0.001) among patients who underwent CDR. CONCLUSIONS:These propensity-matched results suggest that CDR offers superior long-term outcomes compared with ACDF. While these findings suggest potential advantages associated with motion-preserving surgery, the results should be interpreted cautiously given the inherent limitations of retrospective electronic health record-based analyses. As the US population continues to age and technologies continue to improve, CDR should be considered as a viable alternative to motion-sacrificing procedures for the treatment of degenerative cervical spine pathologies. Future studies should investigate long-term clinical and radiographic outcomes to further elucidate the findings of this study.

OBJECTIVE CSF leaks are a common complication of spinal surgery, occurring in 3%-16% of elective cases and up to 12% of operatively managed traumatic spinal injuries. They can also occur following lumbar puncture. However, there are limited data on utilization and reimbursement trends within Medicare for CSF leak repairs. Characterizing the economic burden and market of spinal CSF leak repair has implications for optimizing effective management strategies. This study evaluates reimbursement and utilization trends for both surgical repair and epidural blood patch (EBP) management of spinal CSF leaks between the years 2000 and 2021. METHODS The 2000-2021 provider utilization and payment files from the Centers for Medicare & Medicaid Services was queried to identify all spinal CSF leak repair procedures billed to Medicare part B. For each management strategy, the number of procedures, total charges submitted by providers, and total Medicare reimbursements were extracted. Monetary data were adjusted to 2021 US dollars using changes to the Consumer Price Index. Spearman rank correlation coefficients were used to analyze trends in reimbursement and utilization. RESULTS From 2000 to 2021, 62,826 open surgical repairs were reported with $37,082,407 paid in reimbursement (45.9% repaired without additional laminectomy, 28.8% with laminectomy, and 25.3% requiring a spinal dural graft). An additional 63,227 EBPs were reported for CSF leak repair with $7,567,776 in reimbursement. Reimbursement for CSF leak repair without laminectomy totaled $15,707,094; repairs with laminectomy, $12,429,690; and use of spinal dural grafts, $8,945,444. Both a significant 162% increase in the utilization of spinal dural grafts (333 to 872, p < 0.001) and a decrease in reimbursement ($671 to $496, p = 0.02) was observed from 2000 to 2021. There was a significant 31% decrease in reimbursement for EBP ($148 to $102, p < 0.001). CONCLUSIONS Significant variability exists in the management strategy and rates of utilization and reimbursement for treating spinal durotomy. Despite increasing utilization of spinal dural grafts, reimbursement has declined between the years 2000 and 2021. Further studies are required to optimize both the cost effectiveness and efficacy of spinal durotomy repair techniques.

Predictive modeling has the potential to improve preoperative planning and resource allocation in lumbar fusion surgery. This study aimed to identify the 20 most important variables for predicting prolonged postoperative length of stay (pLOS) using machine learning (ML). The ACS-NSQIP database was queried for lumbar fusion procedures performed between 2012 and 2022, including ALIF, PlatIF, PLIF, and combined PLIF+PlatIF. Variable selection was performed using MUVR and Boruta, followed by hierarchical clustering and 5-fold cross-validation to ensure feature robustness. The 20 selected features were used to train multiple ML models, including tree-based classifiers (Random Forest, XGBoost, CatBoost, LightGBM), support vector classifiers, neural networks, ensemble methods, and logistic regression. A total of 114,892 patients were included. Eleven patient-specific and nine procedural variables were identified as most predictive of prolonged pLOS. Among patient factors, dialysis, congestive heart failure, and bleeding disorders were strongest predictors. Among procedural factors, osteotomy, billing of additional fusion codes, and longer operation time had the greatest impact. The neural network achieved the highest accuracy (71.2%), recall (79.4%), and F1-score (73.8%), though all models performed similarly, with minimal variation in classification metrics. These findings underscore that model choice plays a limited role once optimal features are selected - feature selection was the most critical determinant of predictive performance.

Existing neurosurgical patient education materials (PEMs) can be complex for the average American that reads at an eighth-grade level and may contribute to poor health literacy. Large language model chat bots may help to re-write existing PEMs to improve readability in a cost-effective manner. Neurosurgical PEMs pertaining to stroke, laminectomy, pituitary tumors, epilepsy, and hydrocephalus published by the top 100 US hospitals as ranked by the U.S. News Health Report were collected. ChatGPT-4 was used to re-write 25 randomly selected PEMs at or near the reading level of the average American (eighth-grade reading level). Re-written PEMs were assessed using the following measures of reading level and difficulty: Flesch Kincaid Grade Level, Flesch Reading Ease (FRE), Gunning Fog Index (GFI), Automated Readability Index (ARI), Coleman-Liau Index, and the SMOG index readability score. The accuracy of all re-written PEMs was assessed by a senior neurosurgical resident. The mean FRE score for rewritten PEMs on each topic were significantly lower than non-rewritten materials (p<0.01) except spinal stenosis (p=0.104) and were validated for accuracy. For rewritten materials the mean Kincaid score was 7.58, the mean ARI was 9.53, the mean Coleman-Liau was 11.51, the mean GFI was 9.62, and the mean SMOG Index was 9.28. The ARI for rewritten hydrocephalus and Coleman-Liau score for pituitary tumors did not differ significantly compared to the original PEMs. All other comparisons indicated significantly improved readability of rewritten PEMs compared to original PEMs (p<0.05). Large language model chatbots, such as ChatGPT-4, can be used to efficiently re-write these PEMs at a lower reading level while maintaining the accuracy of the material.

Adequate understanding of health information is a stronger determinant of health than several demographic factors including age, income, or employment status. However, existing neurosurgical patient education materials (PEMs) may be too complex for the average American that reads at an eighth-grade level and may contribute to poor health literacy. Neurosurgical PEMs pertaining to stroke, laminectomy, pituitary tumors, epilepsy, and hydrocephalus published by the top 100 US hospitals as ranked by the U.S. News Health Report were collected. All PEMs were assessed, and common measures of reading level and difficulty were generated including: Flesch Kincaid Grade Level, Flesch Reading Ease (FRE), Gunning Fog Index (GFI), Automated Readability Index (ARI), Coleman-Liau Index, and the SMOG index readability score. The mean FRE score for PEMs on all 5 common neurosurgical conditions (stroke, laminectomy, pituitary tumors, epilepsy and hydrocephalus) were significantly greater than corresponding scores for an 8th-grade reading level (p<0.001). The mean Kincaid value ranged from 8.38 - 12.49. The mean ARI ranged from 10.25 - 14.30 across conditions. Mean Coleman-Liau scores range from 11.26 - 14.08. The mean GFI and SMOG Index ranged from 10.68 - 15.43 and 10.19 - 13.21, respectively. All mean scores were significantly greater than an 8th-grade reading level (p<0.01). Existing PEMs published by the top U.S. hospitals for common neurosurgical conditions may be too complicated for the average American that reads at an eighth-grade level. PEMs may not achieve the intended goal of health literacy and could be rewritten to better meet patient needs.

BACKGROUND AND OBJECTIVES:Adequate understanding of health information has been shown to be a stronger determinant of health than several demographic factors, including age, income, or employment status. However, existing neurosurgical patient education materials (PEMs) may be too complex for the average American and may contribute to poor health literacy. Large language model chatbots may provide a rapid and low-cost means of rewriting existing PEMs at a lower reading level to improve patient understanding and overall health literacy. METHODS:Neurosurgical PEMs pertaining to stroke, laminectomy, pituitary tumors, epilepsy, and hydrocephalus published by the top 100 US hospitals were collected. For all PEMs, common measures of reading level and difficulty were generated, including Flesch Kincaid Grade Level, Flesch Reading Ease (FRE), Gunning Fog Index, Automated Readability Index, Coleman-Liau Index, and the Simple Measure of Gobbledygook Index readability score. ChatGPT-4 was then used to rewrite 25 randomly selected PEMs at or near the reading level of the average American (eighth-grade reading level). The rewritten PEMs were assessed for readability using the same measures of reading level and difficulty. RESULTS:The mean FRE for PEMs on all 5 common neurosurgical conditions were significantly greater than corresponding scores for an eighth-grade reading level ( P < .001). The mean Kincaid value, Automated Readability Index, Coleman-Liau score, Gunning Fog Index, and Simple Measure of Gobbledygook Index for PEMs on each condition were all significantly greater than an eighth-grade reading level ( P < .01). The mean FRE score for rewritten PEMs on each topic were significantly lower than nonrewritten materials ( P < .01) except spinal stenosis ( P = .104) and were validated for accuracy. CONCLUSION:Existing PEMs published by the top US hospitals for common neurosurgical conditions may be too complicated for the average American that reads at an eighth-grade level. Large language model chatbots can be used to efficiently rewrite these PEMs at a lower reading level while maintaining the accuracy of the material.

BACKGROUND AND OBJECTIVES:Brachial plexus injuries are a devastating and often life-altering event. For patients with flail limb and no intraplexal donors, intercostal (IC) to musculocutaneous nerve transfer allows for restoration of elbow flexion although the surgery can be long with a painful recovery. In this article, we demonstrate a minimal incision video-assisted thoracoscopic surgery (VATS) method and a robotic approach to IC harvesting for nerve to biceps brachii reanimation. METHODS:This study was performed on cadaveric donors. We initially used the VATS technique and placed 3 thoracoscopic ports at IC levels 9 and 10. IC nerve 3 was harvested thoracoscopically under direct video visualization. The nerve was carefully neurolysed at the anterior-most edge and exited at the third midaxillary IC space. The nerve to biceps brachii was isolated and neurolysed proximally. The third IC nerve was coapted to the nerve to biceps brachii. On a separate cadaver, we then performed the robotic IC nerve harvest, dissecting out IC nerves 3, 4, and 5. RESULTS:The VATS technique was used to successfully harvest the third IC nerve. The total harvest time of the single nerve was 55 minutes. The robotic intrathoracic approach was used to successfully harvest IC nerves 3, 4, and 5. The average harvest time for each of the 3 nerves was 27 minutes. CONCLUSION:This demonstrates a cadaveric proof of concept for the minimally invasive intrathoracic approach to harvest IC nerves for brachial plexus injuries through both the VATS and robotic techniques. Future studies are warranted for intrathoracic approaches to reanimate other peripheral nerves.

Background:Arachnoid cysts are commonly congenital but they can also be acquired in the setting of surgery, trauma, hemorrhage, or infection. The pathophysiology of arachnoid cysts may involve adhesive arachnoiditis secondary to arachnoid tissue injury, leading to cerebrospinal fluid accumulation. While often asymptomatic, arachnoid cysts can produce symptoms due to mass effect or hydrocephalus. Intraventricular or periventricular cysts can lead to ventricular entrapment. Case Description:In this video, we present a case of a 28-year-old female patient who developed an intraventricular arachnoid cyst following resection of a Grade I meningioma. This led to left lateral ventricular entrapment and associated hemianopia, headache, nausea, and fatigue. The arachnoid cyst was managed with endoscopic fenestration with the aid of stereotactic neuronavigation. The operative video reveals multiple veils of adhered arachnoid firmly associated with choroid plexus as well as with the lateral walls of the ventricle. Using blunt instruments, bipolar cautery, endoscopic microscissors, and a Fogarty balloon, a wide communication between the body of the lateral ventricle and the trapped temporal horn and atrium was achieved. The patient experienced complete resolution of preoperative symptoms. Follow-up imaging revealed interval reduction in the size of the left lateral ventricle without evidence of cyst recurrence. The patient consented to the procedure, and the Institutional Review Board approval was not required as all identifiable patient information is removed. Conclusion:While there are no significant differences in clinical or radiologic outcomes between different surgical strategies in managing arachnoid cysts, endoscopic cyst fenestration offers a balance between definitive treatment and avoidance of shunt dependence.

Objective Durotomy, or dural tear, is common element of spine surgery. All surgeons should be comfortable with techniques for primary repair to avoid follow-up complications. As such, there is need to develop a readily available model of a durotomy that mimics human physiology and the intraoperative experience to enhance trainee skill in durotomy repair prior to encountering this scenario in clinical practice. Design A collagen sausage casing serving as dura was embedded in a 3D-printed lumbar spine with a multi-level laminectomy defect, mobile cauda equina and paraspinal tissues mimicking a standard open posterior approach to the spine. This model was connected to a closed loop of physiologically pressurized and pulsatile flow of artificial spinal fluid. Conclusions We present a lumbar spine durotomy model with physiologically pressurized and pulsatile fluid flow, paraspinal tissues and mobile nerve roots. This model provides a reproducible and reusable means for the trainee spine surgeon to gain experience in primary dural repair without the need for expensive cadaveric specimens or large dedicated lab space. Additionally, this model can serve as a conduit for novel device development to advance the field of spine surgery.

Peripheral nerve injuries (PNI) secondary to gunshot wounds (GSW) are a unique category of neurologic injury that can cause morbidity in young, otherwise healthy individuals, occurring in 25–36

Artificial intelligence (AI) is reshaping neurosurgery, offering unprecedented opportunities to enhance diagnostics, personalize treatment, and predict outcomes. At the heart of this transformation is the ability to effectively harness big data (BD) within the electronic medical record. Understanding these data structures is essential for making sense of the vast volumes of information generated in modern neurosurgical practice. Equally important are the machine learning (ML) models driving these advancements. From supervised learning and convolutional neural networks to generative AI, these tools are already making a mark in areas such as brain tumor segmentation and spine surgery outcome predictions. Their versatility highlights the potential of ML to complement clinical expertise and streamline decision-making in neurosurgery. However, adopting BD and ML also brings ethical challenges that cannot be ignored. Bias in algorithms threatens to reinforce health disparities, whereas concerns about data privacy demand vigilance in handling sensitive patient information. In addition, the question of liability looms large as ML increasingly influences clinical decisions. The aim of the study was to provide a roadmap for neurosurgeons navigating the evolving intersection of BD, ML, and ethical responsibility in the AI era.

BACKGROUND:Double crush syndrome (DCS) is defined as multiple sites of compression along a single nerve. The combination of a compressive proximal lesion in the lumbar spine and a distal common peroneal nerve entrapment may result in compound nerve dysfunction. METHODS:A retrospective analysis of 100 patients who underwent common peroneal nerve decompression with a diagnosis of L5 radiculopathy between January 2000 and April 2023 at two quaternary academic institutions was performed. Patients were included if they had both active L5 radiculopathy and active peroneal mononeuropathy on electromyography (EMG) or imaging findings. Ten patients had a "pure" DCS of the lower extremity, meaning the co-occurrence of both active diseases based on EMG or imaging. Descriptive statistics of patient demographics, clinical presentation, surgical details, and outcomes were performed. RESULTS:All 10 patients underwent common peroneal nerve decompression, 3 of the patients underwent a prior lumbar spine surgery to address their L5 radiculopathy. Preoperatively, 6 patients (60%) had a positive Tinel's sign, which reduced to 3 patients (30%) postoperatively. Relative to preoperative strength, dorsiflexion, extensor hallucis longus, and ankle eversion strength all improved on average following common peroneal nerve decompression. The presence of numbness in the lateral leg or dorsal foot decreased from 9 (90%) preoperatively to 4 (40%) postoperatively. CONCLUSIONS:This is the first series to report DCS with two active points of compression in the lumbar spine and lower extremity based on EMG and imaging findings. Common peroneal nerve decompression was found to improve average dorsiflexion strength.

A syrinx involves cystic dilation of the central canal of the spinal cord due to the accumulation of cerebrospinal fluid and often results in a neurological deficit. While treatment options vary, surgical management is often utilized and requires an open durotomy. A 70-year-old female presented with one year of progressive low back pain with associated leg numbness, urinary incontinence, bilateral foot drop, and imbalance resulting in multiple falls. MRI revealed a terminal syrinx at the level of the conus medullaris. She underwent an uncomplicated T12-L1 laminectomy for syrinx fenestration and syringo-subarachnoid shunt placement, resulting in improved bladder/bowel function, strength, sensation, and ambulation. The necessary midline durotomy was closed with a running 6-0 prolene suture fastened at the start and end of the suture line with the novel use of the COR-KNOT MICROTM device (LSI Solutions, Victor, New York, US). We present the successful surgical management of a progressively symptomatic terminal syrinx using a syringo-subarachnoid shunt and the novel use of the COR-KNOT MICROTM titanium fastener device for dural closure. Earlier shunting should be considered in the disease course of terminal syrinx, and the successful novel use of a titanium fastener on the dura warrants further investigation.

Study design: This study was a multicenter retrospective analysis of cervical spine gunshot wound (GSW) patients. Objective: The present study was conducted to evaluate the management and outcomes of vascular injuries following GSW involving the cervical spine. Summary of background data: Gunshot wounds (GSW) injuring the cervical spine are associated with high rates of vascular injury. Methods: Charts of patients with GSW involving the cervical spine at two Level 1 trauma centers were reviewed from 2010 to 2021 for demographics, injury characteristics, management and follow-up. Statistical analysis included T tests and ANOVA for comparisons of continuous variables and chi-square testing for categorical variables, non-parametric tests were used when indicated. Beta-binomial models were used to estimate the probabilities outcomes. Bayesian regression models were utilized to compute risk ratios (RR) and their 95 % confidence intervals (CI) to enhance the inferential robustness. Results: 40 patients with cervical spine GSW and associated cerebrovascular injury were included in our analysis. 15 % of patients had Biffl grade (BG) V injuries, 50 % grade IV, and 35 % grade III-I. Angiography was performed in 35 % of patients. 5 of these patients (BG V-III) required endovascular treatment for pseudoaneurysm obliteration or parent vessel sacrifice. 7 patients (22 %) showed evidence of progression. 70 % of patients were placed on antiplatelet therapy for stroke prevention. Bayesian regression models with a skeptical prior for cerebral ischemia revealed a mean RR of 4.82 (95 % CI 1.02-14.48) in the BG V group, 0.75 (95 % CI 0.13-2.26) in the BG IV group, and 0.61 (95 % CI 0.06-2.01) in the combined BG III-I group. For demise the mean RR was 3.41 (95 % CI 0.58-10.65) in the BG V group and 1.69 (95 % CI 0.29-5.97) in the BG IV group. In the high BG (V, IV) group, 54.55 % of patients treated with antiplatelet therapy had complications. None of the patients that were treated with antiplatelet therapy in the low BG (III-I) group had complications. Conclusions: Cervical spine GSWs are associated with high-grade vascular injuries and may require early endovascular intervention. Additionally, a high rate of injury progression was seen on follow up imaging, requiring subsequent intervention. Reintervention and demise were common and observed in high BG (V, IV) groups. The incidence of stroke was low, especially in low BG (I-III) groups, suggesting that daily aspirin prophylaxis is adequate for long-term stroke prevention.

BACKGROUND CONTEXT:Gunshot wounds (GSWs) to the vertebral column represent an important cause of morbidity and mortality in the United States, constituting approximately 20% of all spinal injuries. The management of these injuries is an understudied and controversial topic, given its heterogeneity and lack of follow-up data. PURPOSE:To characterize the management and follow-up of GSWs to the spine. STUDY DESIGN/SETTING:A multi-institutional retrospective review of the experience of two urban Level 1 trauma centers. PATIENT SAMPLE:Patients with GSWs to the spine between 2010-2021. OUTCOME MEASURES:Measures included work status, follow-up healthcare utilization, and pain management were collected. METHODS:Charts were reviewed for demographics, injury characteristics, surgery and medical management, and follow-up. Statistical analysis included T-tests and ANOVA for comparisons of continuous variables and chi-square testing for categorical variables. All statistics were performed on SPSS v24 (IBM, Armonk, NY). RESULTS:A total of 271 patients were included for analysis. The average age was 28 years old, 82.7% of patients were black, 90% were male, and 76.4% had Medicare/Medicaid. The thoracic spine (35%) was most commonly injured followed by lumbar (33.9%) and cervical (25.6%). Cervical GSW was associated with higher mortality (p<.001); 8.7% of patients developed subsequent osteomyelitis/discitis, 71.3% received prophylactic antibiotics, and 56.1% of cervical GSW had a confirmed vertebral or carotid artery injury. ASIA scores at presentation were most commonly A (26.9%), D (20.7%), or E (19.6%), followed by C (7.4%) and B (6.6%). 18.8% of patients were unable to be assessed at presentation. ASIA score declined in only 2 patients, while 15.5% improved over their hospital stay. Those who improved were more likely to have ASIA B injury (p<.001). Overall, 9.2% of patients underwent spinal surgery. Of these, 33% presented as ASIA A, 21% as ASIA B, 29% as ASIA C, and 13% as ASIA D. Surgery was not associated with an improvement in ASIA score. CONCLUSIONS:Given the ubiquitous and heterogeneous experience with GSWs to the spine, rigorous attempts should be made to define this population and its clinical and surgical outcomes. Here, we present an analysis of 11 years of patients presenting to two large trauma centers to elucidate patterns in presentation, management, and follow-up. We highlight that GSWs to the cervical spine are most often seen in young black male patients. They were associated with high mortality and high rates of injury to vertebral arteries and that surgical intervention did not alter rates of discitis/osteomyelitis or propensity for neurologic recovery; moreover, there was no incidence of delayed spinal instability in the study population.

BACKGROUND AND OBJECTIVE: Iatrogenic spinal durotomies occur at a rate of 1% to 17%. Surgical simulation for durotomy repair is needed to provide affordable, accessible, and validated practice. This study sought to design and validate a simple 3-dimensional printed model for spinal cerebrospinal fluid (CSF) leak repair and to introduce the Rochester original objective structured assessment of technical skills (OSATS) CSF leak (ROCL) repair criteria for assessment. METHODS: A spinal model was designed to mimic a lumbar laminectomy with the L3-5 lamina removed and 3-dimensional printed using Vero polymers. The model was paired with a porcine collagen “dura” that was pressurized using IV saline and overlayed with gel-molded fascial, muscle, and skin layers with an opening. Participants were provided a training model with a 1.5-cm midline durotomy, surgical microinstrument set, microscope, and 6-0 prolene suture. The 25-point ROCL repair criteria were adapted from the original OSATS principles to assess proficiency in surgical repair by 2 blinded neurosurgeons not participating in the trials. Postsimulation survey data regarding model realism were collected. RESULTS: Six residents and 4 attendings participated. Median operative time in minutes was 13 minutes among residents and 7 minutes among attendings. Moreover, the ROCL score was a median of 19/25 for attendings and 15/25 for residents. The suture angle was statistically more consistent among senior residents and attendings compared with junior residents. Participants agreed that the model was realistic (median 4/5), useful for improving the operative technique (median 5/5), and would increase comfort in spinal CSF leak repair procedures (median 5/5). Each reusable model had a cost of $19.99 if printed with polylactic acid and each replacement dura cost <3¢. CONCLUSION: This study presents an affordable, realistic, and educational spinal CSF leak repair model and introduces ROCL for assessment.

You have accessJournal of UrologyProstate Oncology/Penile & Testis Oncology/Misc. Oncology II (V12)1 May 2024V12-01 USE OF ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING TO DEVELOP AN EDUCATION SYSTEM FOR INSTRUCTORLESS SURGICAL TRAINING Nelson N. Stone, Steven Griffith, Eric Delmonico, Michael Wilson, Laura Kim, and Jonathan J. Stone Nelson N. StoneNelson N. Stone , Steven GriffithSteven Griffith , Eric DelmonicoEric Delmonico , Michael WilsonMichael Wilson , Laura KimLaura Kim , and Jonathan J. StoneJonathan J. Stone View All Author Informationhttps://doi.org/10.1097/01.JU.0001009480.90141.21.01AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: Hands-on training for new procedures is limited by the availability of proctors proficient in new medical devices. Training is further constrained by the costs and time associated with the need to travel to be present for the training. A new system utilizing artificial intelligence (AI) to create instructorless training programs could help ameliorate these limitations. METHODS: A prototype education system for an instructorless surgical training (ESIST) program was created for training in the insertion of a needle into the perirectal space prior to spacer device placement. ESIST was developed by utilizing machine learning to convert 4 machine learning models, with a total of 12 unique ultrasound states via AI to instructions and observations. The ESIST program provided instructions to an extended reality headset worn by the student and incorporated corrective measures when needle position errors were encountered (Figure 1). The extended reality (XD) headset was developed to project the visual and audio instructions from the ESIST program and combined them with the live ultrasound images during the procedure. The headset contained see-through reflective lenses allowing the student to observe both the instructions and ultrasound images as an overlay on top of the prostate phantom. When the procedure was completed, the program provided a score. RESULTS: The student placed a needle into the prostate phantom which was recognized by ESIST. The program instructed the student to angle the needle down 7-8 degrees once the perineum was pierced and then insert it above the rectal hump and into the perirectal space. Once the needle reached the base of the prostate with its bevel down the student was asked to switch imaging to axial to ensure the needle was in the midline. Imaging was then switched back to axial and the procedure was then scored by ESIST. CONCLUSIONS: We demonstrated the feasibility of using AI to create a program to train physicians in a surgical procedure without the need for a proctor to be present. AI can be used to develop more complex surgical training programs. Download PPT Source of Funding: The creation of ESIST was supported by NSF 1913911 and Viomerse, Inc © 2024 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 211Issue 5SMay 2024Page: e994 Advertisement Copyright & Permissions© 2024 by American Urological Association Education and Research, Inc.Metrics Author Information Nelson N. Stone More articles by this author Steven Griffith More articles by this author Eric Delmonico More articles by this author Michael Wilson More articles by this author Laura Kim More articles by this author Jonathan J. Stone More articles by this author Expand All Advertisement PDF downloadLoading ...