Impact of environmental pollutants on pediatric brain tumor incidence in New Jersey.

OBJECTIVE: The relationship between environmental contaminants and brain tumor incidence in adults has been thoroughly explored but research into how these contaminants affect pediatric brain tumor (PBT) incidence has not been explored. Children, typically having more limited geographical movement and thus more consistent environmental contaminant exposure, might offer more reliable insights into which environmental contaminants affect the incidence of brain tumors. The present study is the first to focus on exploring whether a possible association exists between the incidence of PBTs and exposure to environmental pollutants in New Jersey (NJ). METHODS: Linear regressions were run between PBT incidence and the concentration of air quality pollutants such as Ozone (O3), Particulate Matter 2.5 (PM2.5), Particulate Matter 10 (PM10), and Carbon Monoxide (CO). Similarly, linear regressions were run between PBT incidence and Elevated Blood Lead Levels (BLL). RESULTS: The study observed a significant positive relationship between O3 and PBT incidence (ß = 0.34, p = 0.028). However, the relationship between PBT incidence, and environmental pollutants such as CO (ß = 0.0047, p = 0.098), PM2.5 (ß = -0.2624, p = 0.74), and PM10 (ß = -0.7353, p = 0.073) were found to be nonsignificant. For elevated BLL, nonsignificant relationships with PBT incidence were observed at 10-14 µg/dL (ß = -39.38, p = 0.30), 15-19 µg/dL (ß = -67.00, p = 0.21), and 20-44 µg/dL (ß = -201.98, p = 0.12). CONCLUSIONS: The results indicate a possible impact of O3 on the incidence of PBTs in NJ. In contrast to the significant links found in prior studies of adult brain tumors, the associations between PBT occurrence and particulate matter were not significant. These findings highlight the importance of further investigating how environmental factors, especially O3, relate to PBTs.

Can Artificial Intelligence Mitigate Missed Diagnoses by Generating Differential Diagnoses for Neurosurgeons?

INTRODUCTION: Neurosurgery emphasizes the criticality of accurate differential diagnoses, with diagnostic delays posing significant health and economic challenges. As large language models (LLMs) emerge as transformative tools in healthcare, this study seeks to elucidate their role in assisting neurosurgeons with the differential diagnosis process, especially during preliminary consultations. METHODS: This study employed three chat-based LLMs, ChatGPT (versions 3.5 and 4.0), Perplexity AI, and Bard AI, to evaluate their diagnostic accuracy. Each LLM was prompted using clinical vignettes, and their responses were recorded to generate differential diagnoses for 20 common and uncommon neurosurgical disorders. Disease-specific prompts were crafted using Dynamed, a clinical reference tool. The accuracy of the LLMs was determined based on their ability to identify the target disease within their top differential diagnoses correctly. RESULTS: For the initial differential, ChatGPT 3.5 achieved an accuracy of 52.63%, while ChatGPT 4.0 performed slightly better at 53.68%. Perplexity AI and Bard AI demonstrated 40.00% and 29.47% accuracy, respectively. As the number of considered differentials increased from two to five, ChatGPT 3.5 reached its peak accuracy of 77.89% for the top five differentials. Bard AI and Perplexity AI had varied performances, with Bard AI improving in the top five differentials at 62.11%. On a disease-specific note, the LLMs excelled in diagnosing conditions like epilepsy and cervical spine stenosis but faced challenges with more complex diseases such as Moyamoya disease and ALS. CONCLUSION: LLMs showcase the potential to enhance diagnostic accuracy and decrease the incidence of missed diagnoses in neurosurgery.

Analysis of Concussions Reported in American Professional Football Injury Reports in the 2019 through 2023 Regular Seasons.

OBJECTIVE: The National Football League (NFL) has seen increasing scrutiny regarding its handling of concussions, especially following an on-field incident involving Miami Dolphins' (™) Quarterback Tua Tagovailoa in the 2022 Season. The authors hoped to elucidate recent trends in the diagnosis and management of concussions over the course of five NFL seasons across 2019-2023. METHODS: The authors queried NFL injury reports from the 2019 through 2023 database recording players listed with Concussions. Weeks missed were calculated using NFL game logs. Player's concussions that did not occur in games, complicated by other injuries, or roster status were excluded. RESULTS: Searches of NFL injury reports resulted in the identification of 664 of 692 (96%) concussions that occurred in regular season games across the 2019-2023 seasons. Over the course of these five seasons 31% of players returned without missing a game, 39% of players missed one game, and 30% of players missed two or games. No significant difference in the number of concussions per game or weeks missed was observed across the seasons observed. Players with concussions on teams that made the playoffs saw less weeks missed than those on non-playoff teams (0.86 v. 1.37, p=0.002). CONCLUSIONS: Since the start of the 2021 NFL season, an increasing incidence of concussions has been noted, yet there was no change observed in the number of weeks missed following concussions. Trends in the rates of concussions across seasons remain largely stable, despite increased scrutiny over concussions in the sport.

Assessing Diversity, Equity, and Inclusion in Patient-Facing Websites in Neurosurgical Departments in the United States.

BACKGROUND: Patient-facing websites serve as essential platforms for disseminating information, engaging with patients, and increasing access to neurosurgical resources and services. Diversity, Equity, and Inclusion are at the forefront of issues facing the field of neurosurgery, especially concerning race and gender disparities in regards to providers in the field. METHODS: Data were collected in regards to the race and gender of patients and providers displayed on the neurosurgery department's patient-facing website in addition to accommodations for disabilities, decreased ability to pay, and language. RESULTS: Patients who were White were depicted more commonly than those of color (69% vs. 31%, P < 0.00001). White patients also were over-represented when compared with the average demographics of the communities in which the hospitals served (P = 0.03846). Neurosurgical providers who were White outnumbered those of color (70% vs. 30%, P < 0.00001). The racial depiction of providers was comparable with racial disparities currently observed in neurosurgery (P = 0.59612). Female neurosurgery providers were seen less than male providers on patient-facing websites (P < 0.00001) but were seen more commonly on patient-facing websites than the percentage of practicing neurosurgeons they currently comprise (28% vs. 8%, P < 0.00001). CONCLUSIONS: The results of this study suggest that patient-facing websites of neurosurgical departments are an area of improvement in regards to Diversity, Equity, and Inclusion in the field of neurosurgery. Disparities are noted in regards to the racial depiction of patients and further call to attention racial and gender disparities in the field of neurosurgery.

Endoscopic Endonasal Transsphenoidal Surgery for the Resection of Pituitary Adenomas: A Prime Candidate for a Shortened Length of Stay Enhanced Recovery after Surgery Protocol? A Systematic Review.

BACKGROUND: Enhanced Recovery After Surgery (ERAS) is a perioperative model of care aimed at optimizing postoperative rehabilitation and reducing hospital length of stay (LOS). Decreasing LOS avoids hospital-acquired complications, reduces cost of care, and improves patient satisfaction. Given the lack of ERAS protocols for endoscopic endonasal transsphenoidal surgery (EETS) resection of pituitary adenomas, a systematic review of EETS was performed to compile patient outcomes and analyze factors that may lead to increased LOS, reoperation, and readmission rates with the intention to contribute to the development of a successful ERAS protocol for EETS. METHODS: The authors performed a Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines-based systematic review of the literature. Information was extracted regarding patient LOS, surgery complications, and readmission/reoperation rates. Pearson's correlations to LOS and reoperation/readmission rates were performed with variables normalized to the number of participants. Statistical significance was set at P value <0.05. RESULTS: Fourteen studies were included, consisting of 2083 patients. The most common complications were cerebrospinal fluid leaks (37%) and postoperative diabetes insipidus (DI) (9%). Transient DI was significantly correlated with shorter LOS. Functional pituitary adenomas were significantly correlated with lower readmission rates while nonfunctional pituitary adenomas were correlated with higher readmission rates. No other factor was found to be significantly correlated with a change in LOS or reoperation rate. CONCLUSIONS: EETS may be an ideal candidate for the development of ERAS cranial protocols. While our data largely supports the safe implementation of shortened LOS protocols in EETS, our findings highlight the importance of transient DI and nonfunctional pituitary adenomas management when formulating ERAS protocols.

Can Publicly Available Artificial Intelligence Successfully Identify Current Procedural Terminology Codes for Common Procedures in Neurosurgery?

OBJECTIVE: Coding for neurosurgical procedures is a complex process that is dynamically changing year to year, through the annual introduction and removal of codes and modifiers. The authors hoped to elucidate if publicly available artificial intelligence (AI) could offer solutions for neurosurgeons with regard to coding. METHODS: Multiple publicly available AI platforms were asked to provide Current Procedural Terminology (CPT) codes and Revenue Value Units (RVU) values for common neurosurgical procedures of the brain and spine with a given indication for the procedure. The responses of platforms were recorded and compared to the currently valid CPT codes used for the procedure and the amount of RVUs that would be gained. RESULTS: Six platforms and Google were asked for the appropriate CPT codes for 10 endovascular, spinal, and cranial procedures each. The highest performing platforms were as follows: Perplexity.AI identified 70% of endovascular, BingAI identified 55% of spinal, and ChatGPT 4.0 with Bing identified 75% of cranial CPT codes. With regard to RVUs, the top performer gained 78% of endovascular, 42% of spinal, and 70% of cranial possible RVUs. With regard to accuracy, AI platforms on average outperformed Google (45% vs. 25%, P = 0.04236). CONCLUSIONS: The ability of publicly available AIs to successfully code for neurosurgical procedures holds great promise in the future. Future development of AI should focus on improving accuracy with regard to CPT codes and providing supporting documentation for its decisions. Improvement on the existing capabilities of AI platforms can allow for increased operational efficiency and cost savings for practices.

Enabling Spike-Based Backpropagation for Training Deep Neural Network Architectures.

Spiking Neural Networks (SNNs) have recently emerged as a prominent neural computing paradigm. However, the typical shallow SNN architectures have limited capacity for expressing complex representations while training deep SNNs using input spikes has not been successful so far. Diverse methods have been proposed to get around this issue such as converting off-the-shelf trained deep Artificial Neural Networks (ANNs) to SNNs. However, the ANN-SNN conversion scheme fails to capture the temporal dynamics of a spiking system. On the other hand, it is still a difficult problem to directly train deep SNNs using input spike events due to the discontinuous, non-differentiable nature of the spike generation function. To overcome this problem, we propose an approximate derivative method that accounts for the leaky behavior of LIF neurons. This method enables training deep convolutional SNNs directly (with input spike events) using spike-based backpropagation. Our experiments show the effectiveness of the proposed spike-based learning on deep networks (VGG and Residual architectures) by achieving the best classification accuracies in MNIST, SVHN, and CIFAR-10 datasets compared to other SNNs trained with a spike-based learning. Moreover, we analyze sparse event-based computations to demonstrate the efficacy of the proposed SNN training method for inference operation in the spiking domain.

Rubus: A compiler for seamless and extensible parallelism.

Nowadays, a typical processor may have multiple processing cores on a single chip. Furthermore, a special purpose processing unit called Graphic Processing Unit (GPU), originally designed for 2D/3D games, is now available for general purpose use in computers and mobile devices. However, the traditional programming languages which were designed to work with machines having single core CPUs, cannot utilize the parallelism available on multi-core processors efficiently. Therefore, to exploit the extraordinary processing power of multi-core processors, researchers are working on new tools and techniques to facilitate parallel programming. To this end, languages like CUDA and OpenCL have been introduced, which can be used to write code with parallelism. The main shortcoming of these languages is that programmer needs to specify all the complex details manually in order to parallelize the code across multiple cores. Therefore, the code written in these languages is difficult to understand, debug and maintain. Furthermore, to parallelize legacy code can require rewriting a significant portion of code in CUDA or OpenCL, which can consume significant time and resources. Thus, the amount of parallelism achieved is proportional to the skills of the programmer and the time spent in code optimizations. This paper proposes a new open source compiler, Rubus, to achieve seamless parallelism. The Rubus compiler relieves the programmer from manually specifying the low-level details. It analyses and transforms a sequential program into a parallel program automatically, without any user intervention. This achieves massive speedup and better utilization of the underlying hardware without a programmer's expertise in parallel programming. For five different benchmarks, on average a speedup of 34.54 times has been achieved by Rubus as compared to Java on a basic GPU having only 96 cores. Whereas, for a matrix multiplication benchmark the average execution speedup of 84 times has been achieved by Rubus on the same GPU. Moreover, Rubus achieves this performance without drastically increasing the memory footprint of a program.

Weighted-incidence syndromic combination antibiograms to guide empiric treatment of critical care infections: a retrospective cohort study.

INTRODUCTION: Empiric antimicrobial selection for critical care infections must balance the need for timely adequate coverage with the resistance pressure exerted by broadspectrum agents. We estimated the potential of weighted incidence syndromic combination antibiograms (WISCAs) to improve time to adequate coverage for critical care infections. In contrast to traditional antibiograms, WISCAs display the likelihood of coverage for a specific infectious syndrome (rather than individual pathogens), and also take into account the potential for poly-microbial infections and the use of multi-drug regimens. METHODS: Cases of ventilator-associated pneumonia (VAP) and catheter-related bloodstream infection (CRBSI) were identified over three years using stringent surveillance criteria. Based on the susceptibility profile of the culprit pathogens, we calculated the WISCA percentages of infections that would have been adequately covered by common antimicrobial(s). We then computed the excess percentage coverage offered by WISCA regimens compared to the actual antimicrobials administered to patients by 12 h, 24 h, and 48 h from culture collection. RESULTS: Among 163 patients with critical care infection, standard practice only resulted in adequate coverage of 35% of patients by 12 h, 52% by 24 h, and 75% by 48 h. No WISCA mono-therapy regimen offered greater than 85% adequate overall coverage for VAP and CRBSI. A wide range of dual therapy regimens would have conferred greater than 90% adequate coverage, with excess coverage estimated to be as high as +56%, +42% and +18% at 12 h, 24 h and 48 h, respectively. We did not detect a decrease in mortality associated with early adequate treatment, and so could not estimate potential downstream benefits. CONCLUSIONS: WISCA-derived empiric antimicrobial regimens can be calculated for patients with intensive care unit (ICU)-acquired infections, and have the potential to reduce time to adequate treatment. Prospective research must confirm whether implementation of WISCA prescribing aids facilitate timely adequate treatment and improved ICU outcomes.

Effect of selective decontamination on antimicrobial resistance in intensive care units: a systematic review and meta-analysis.

BACKGROUND: Many meta-analyses have shown reductions in infection rates and mortality associated with the use of selective digestive decontamination (SDD) or selective oropharyngeal decontamination (SOD) in intensive care units (ICUs). These interventions have not been widely implemented because of concerns that their use could lead to the development of antimicrobial resistance in pathogens. We aimed to assess the effect of SDD and SOD on antimicrobial resistance rates in patients in ICUs. METHODS: We did a systematic review of the effect of SDD and SOD on the rates of colonisation or infection with antimicrobial-resistant pathogens in patients who were critically ill. We searched for studies using Medline, Embase, and Cochrane databases, with no limits by language, date of publication, study design, or study quality. We included all studies of selective decontamination that involved prophylactic application of topical non-absorbable antimicrobials to the stomach or oropharynx of patients in ICUs, with or without additional systemic antimicrobials. We excluded studies of interventions that used only antiseptic or biocide agents such as chlorhexidine, unless antimicrobials were also included in the regimen. We used the Mantel-Haenszel model with random effects to calculate pooled odds ratios. FINDINGS: We analysed 64 unique studies of SDD and SOD in ICUs, of which 47 were randomised controlled trials and 35 included data for the detection of antimicrobial resistance. When comparing data for patients in intervention groups (those who received SDD or SOD) versus data for those in control groups (who received no intervention), we identified no difference in the prevalence of colonisation or infection with Gram-positive antimicrobial-resistant pathogens of interest, including meticillin-resistant Staphylococcus aureus (odds ratio 1·46, 95% CI 0·90-2·37) and vancomycin-resistant enterococci (0·63, 0·39-1·02). Among Gram-negative bacilli, we detected no difference in aminoglycoside-resistance (0·73, 0·51-1·05) or fluoroquinolone-resistance (0·52, 0·16-1·68), but we did detect a reduction in polymyxin-resistant Gram-negative bacilli (0·58, 0·46-0·72) and third-generation cephalosporin-resistant Gram-negative bacilli (0·33, 0·20-0·52) in recipients of selective decontamination compared with those who received no intervention. INTERPRETATION: We detected no relation between the use of SDD or SOD and the development of antimicrobial-resistance in pathogens in patients in the ICU, suggesting that the perceived risk of long-term harm related to selective decontamination cannot be justified by available data. However, our study indicates that the effect of decontamination on ICU-level antimicrobial resistance rates is understudied. We recommend that future research includes a non-crossover, cluster randomised controlled trial to assess long-term ICU-level changes in resistance rates. FUNDING: None.