Collaborative hunting in artificial agents with deep reinforcement learning.

Collaborative hunting, in which predators play different and complementary roles to capture prey, has been traditionally believed to be an advanced hunting strategy requiring large brains that involve high-level cognition. However, recent findings that collaborative hunting has also been documented in smaller-brained vertebrates have placed this previous belief under strain. Here, using computational multi-agent simulations based on deep reinforcement learning, we demonstrate that decisions underlying collaborative hunts do not necessarily rely on sophisticated cognitive processes. We found that apparently elaborate coordination can be achieved through a relatively simple decision process of mapping between states and actions related to distance-dependent internal representations formed by prior experience. Furthermore, we confirmed that this decision rule of predators is robust against unknown prey controlled by humans. Our computational ecological results emphasize that collaborative hunting can emerge in various intra- and inter-specific interactions in nature, and provide insights into the evolution of sociality.

From wolves to ants, many animals are known to be able to hunt as a team. This strategy may yield several advantages: going after bigger preys together, for example, can often result in individuals spending less energy and accessing larger food portions than when hunting alone. However, it remains unclear whether this behavior relies on complex cognitive processes, such as the ability for an animal to represent and anticipate the actions of its teammates. It is often thought that 'collaborative hunting' may require such skills, as this form of group hunting involves animals taking on distinct, tightly coordinated roles ­ as opposed to simply engaging in the same actions simultaneously. To better understand whether high-level cognitive skills are required for collaborative hunting, Tsutsui et al. used a type of artificial intelligence known as deep reinforcement learning. This allowed them to develop a computational model in which a small number of 'agents' had the opportunity to 'learn' whether and how to work together to catch a 'prey' under various conditions. To do so, the agents were only equipped with the ability to link distinct stimuli together, such as an event and a reward; this is similar to associative learning, a cognitive process which is widespread amongst animal species. The model showed that the challenge of capturing the prey when hunting alone, and the reward of sharing food after a successful hunt drove the agents to learn how to work together, with previous experiences shaping decisions made during subsequent hunts. Importantly, the predators started to exhibit the ability to take on distinct, complementary roles reminiscent of those observed during collaborative hunting, such as one agent chasing the prey while another ambushes it. Overall, the work by Tsutsui et al. challenges the traditional view that only organisms equipped with high-level cognitive processes can show refined collaborative approaches to hunting, opening the possibility that these behaviors may be more widespread than originally thought ­ including between animals of different species.

Lymphopenia after palliative radiotherapy for vertebral metastases.

Lymphopenia is a well-known side effect of radiotherapy and has been shown to have a negative impact on patient outcomes. However, the extent of lymphopenia caused by palliative radiotherapy and its effect on patient prognosis has not been clarified. The aim of this study was to determine the incidence and severity of lymphopenia after palliative radiotherapy for vertebral metastases and to determine their effects on patients' survival outcomes. We conducted a retrospective analysis for patients who underwent palliative radiotherapy for vertebral metastases and could be followed up for 12 weeks. Lymphocyte counts were documented at baseline and throughout the 12-week period following the start of radiotherapy and their medians and interquartile ranges (IQRs) were recorded. Exploratory analyses were performed to identify predictive factors for lymphopenia and its impact on overall survival (OS). A total of 282 cases that met the inclusion criteria were analyzed. The median baseline lymphocyte count was 1.26 × 103/µl (IQR: 0.89-1.72 × 103/µl). Peak lymphopenia occurred at a median of 26 days (IQR: 15-45 days) with a median nadir of 0.52 × 103/µl (IQR: 0.31-0.81 × 103/µl). Long-term analysis of patients surviving for 1 year showed that lymphopenia persisted at 1 year after radiotherapy. The main irradiation site, radiation field length and pretreatment lymphocyte count were significantly related to grade 3 or higher lymphopenia. Lymphopenia was identified as a significant predictor of OS by multivariate Cox regression analysis. This study demonstrated the incidence of lymphopenia after palliative radiotherapy for vertebral metastases and its effect on patients' OS.

R-Cut: Enhancing Explainability in Vision Transformers with Relationship Weighted Out and Cut.

Transformer-based models have gained popularity in the field of natural language processing (NLP) and are extensively utilized in computer vision tasks and multi-modal models such as GPT4. This paper presents a novel method to enhance the explainability of transformer-based image classification models. Our method aims to improve trust in classification results and empower users to gain a deeper understanding of the model for downstream tasks by providing visualizations of class-specific maps. We introduce two modules: the "Relationship Weighted Out" and the "Cut" modules. The "Relationship Weighted Out" module focuses on extracting class-specific information from intermediate layers, enabling us to highlight relevant features. Additionally, the "Cut" module performs fine-grained feature decomposition, taking into account factors such as position, texture, and color. By integrating these modules, we generate dense class-specific visual explainability maps. We validate our method with extensive qualitative and quantitative experiments on the ImageNet dataset. Furthermore, we conduct a large number of experiments on the LRN dataset, which is specifically designed for automatic driving danger alerts, to evaluate the explainability of our method in scenarios with complex backgrounds. The results demonstrate a significant improvement over previous methods. Moreover, we conduct ablation experiments to validate the effectiveness of each module. Through these experiments, we are able to confirm the respective contributions of each module, thus solidifying the overall effectiveness of our proposed approach.

An evaluation of rehabilitation students' learning goals in their first year: a text mining approach.

Introduction: Qualitative information in the form of written reflection reports is vital for evaluating students' progress in education. As a pilot study, we used text mining, which analyzes qualitative information with quantitative features, to investigate how rehabilitation students' goals change during their first year at university. Methods: We recruited 109 first-year students (66 physical therapy and 43 occupational therapy students) enrolled in a university rehabilitation course. These students completed an open-ended questionnaire about their learning goals at the time of admission and at 6 and 12 months after admission to the university. Text mining was used to objectively interpret the descriptive text data from all three-time points to extract frequently occurring nouns at once. Then, hierarchical cluster analysis was performed to generate clusters. The number of students who mentioned at least one noun in each cluster was counted and the percentages of students in each cluster were compared for the three periods using Cochran's Q test. Results: The 31 nouns that appeared 10 or more times in the 427 sentences were classified into three clusters: "Socializing," "Practical Training," and "Classroom Learning." The percentage of students in all three clusters showed significant differences across the time periods (p < 0.001 for "Socializing"; p < 0.01 for "Practical Training" and "Classroom Learning"). Conclusion: These findings suggest that the students' learning goals changed during their first year of education. This objective analytical method will enable researchers to examine transitional trends in students' reflections and capture their psychological changes, making it a useful tool in educational research.

Anti-MDA5 Antibody-positive Clinically Amyopathic Dermatomyositis with Rapidly Progressing Interstitial Lung Disease Successfully Treated by Initiation of Combined Immunosuppressive Therapy Plus Plasma Exchange and Subsequently Switching Tacrolimus to Tofacitinib.

A 36-year-old man with inverse Gottron's sign was admitted for clinically amyopathic dermatomyositis (CADM) with rapidly progressive interstitial lung disease (RP-ILD). Early addition of plasma exchange (PE) to triple therapy improved severe respiratory failure and transiently decreased serum ferritin levels and anti-melanoma differentiation-associated gene 5 antibody (anti-MDA5 Ab) titers. Furthermore, switching from tacrolimus to tofacitinib resulted in disease remission. Recognition of the inverse Gottron's sign may allow for the earlier diagnosis of anti-MDA5 Ab-positive dermatomyositis, and early addition of PE to triple therapy and administration of tofacitinib in refractory cases may be effective for anti-MDA5 Ab-positive CADM with RP-ILD under life-threatening conditions.

Estimating Counterfactual Treatment Outcomes Over Time in Complex Multiagent Scenarios.

Evaluation of intervention in a multiagent system, for example, when humans should intervene in autonomous driving systems and when a player should pass to teammates for a good shot, is challenging in various engineering and scientific fields. Estimating the individual treatment effect (ITE) using counterfactual long-term prediction is practical to evaluate such interventions. However, most of the conventional frameworks did not consider the time-varying complex structure of multiagent relationships and covariate counterfactual prediction. This may lead to erroneous assessments of ITE and difficulty in interpretation. Here, we propose an interpretable, counterfactual recurrent network in multiagent systems to estimate the effect of the intervention. Our model leverages graph variational recurrent neural networks (GVRNNs) and theory-based computation with domain knowledge for the ITE estimation framework based on long-term prediction of multiagent covariates and outcomes, which can confirm the circumstances under which the intervention is effective. On simulated models of an automated vehicle and biological agents with time-varying confounders, we show that our methods achieved lower estimation errors in counterfactual covariates and the most effective treatment timing than the baselines. Furthermore, using real basketball data, our methods performed realistic counterfactual predictions and evaluated the counterfactual passes in shot scenarios.

Decentralized policy learning with partial observation and mechanical constraints for multiperson modeling.

Extracting the rules of real-world multi-agent behaviors is a current challenge in various scientific and engineering fields. Biological agents independently have limited observation and mechanical constraints; however, most of the conventional data-driven models ignore such assumptions, resulting in lack of biological plausibility and model interpretability for behavioral analyses. Here we propose sequential generative models with partial observation and mechanical constraints in a decentralized manner, which can model agents' cognition and body dynamics, and predict biologically plausible behaviors. We formulate this as a decentralized multi-agent imitation-learning problem, leveraging binary partial observation and decentralized policy models based on hierarchical variational recurrent neural networks with physical and biomechanical penalties. Using real-world basketball and soccer datasets, we show the effectiveness of our method in terms of the constraint violations, long-term trajectory prediction, and partial observation. Our approach can be used as a multi-agent simulator to generate realistic trajectories using real-world data.

Understanding hereditary hemorrhagic telangiectasia: From genetic anomalies to systemic manifestations, quality of life, and epistaxis management-Exploring the otolaryngologist's integral role.

Hereditary hemorrhagic telangiectasia (HHT), also known as Osler-Rendu-Weber syndrome, is a rare autosomal dominant disorder characterized by vascular malformations. This comprehensive review aimed to provide an overview and summarize various aspects of HHT, including the genetic abnormalities, complications associated with visceral arteriovenous malformations (AVMs), prognosis of HHT, quality of life (QOL), and treatment of epistaxis. In addition, this review highlights the challenges in diagnosing HHT and emphasizes the critical role of otolaryngologists in the early detection of HHT. Otolaryngologists can refer patients with refractory epistaxis for AVM screening to expedite intervention. Mutation of the genes involved in the transforming growth factor-ß signaling pathway leads to the incidence of HHT, resulting in the formation of abnormal blood vessel formation. These vascular malformations commonly manifest as telangiectasia on the skin and mucous membranes; however, epistaxis remains the hallmark symptom of HHT. The impact of HHT goes beyond the visible symptoms and often includes the formation of life-threatening visceral AVMs in the lungs, liver, and brain. The prognosis of patients with HHT is closely related to the development of these complications, necessitating timely diagnosis and intervention. Refractory epistaxis diminishes the QOL of patients with HHT. The management of epistaxis ranges from conservative measures to advanced interventions such as prevention, conservative treatments, ablation, surgical procedures, and the administration of anti-angiogenic agents. However, effective management requires a multidisciplinary approach. The diagnosis of HHT remains challenging due to its variable presentation and lack of awareness among physicians. This review highlights the importance of reducing the duration between symptom onset and diagnosis. Otolaryngologists who are experienced in the management of refractory epistaxis can aid in identifying potential cases of HHT. They can facilitate the initiation of screening for visceral AVMs via prompt recognition of the signs and symptoms of HHT, contributing to improved patient outcomes. Early detection and intervention through screening can extend the life expectancy of patients with HHT to levels comparable with that of the general population. In conclusion, this review provides insight into various aspects of HHT and emphasizes the importance of timely diagnosis and intervention in the mitigation of the potentially life-threatening complications associated with this disorder. Otolaryngologists play a critical role in this process, serving as gatekeepers to the identification of cases of HHT and implementation of appropriate screening and management pathways, thereby improving the life expectancy and QOL of patients.

Glasgow prognostic score for assessing the efficacy of whole-brain radiation therapy in cases of recursive partitioning analysis class 2 and class 3 multiple brain metastases: a retrospective study.

PURPOSE: Whole-brain radiotherapy (WBRT) may not be beneficial for patients with brain metastases (BMs). The Glasgow Prognostic Score (GPS) is a suggested prognostic factor for malignancies. However, GPS has never been assessed in patients with BMs who have undergone WBRT. The purpose of this study was to determine whether GPS can be used to identify subgroups of patients with BMs who have a poor prognosis, such as recursive partitioning analysis (RPA) Class 2 and Class 3, and who will not receive clinical prognostic benefits from WBRT. MATERIALS AND METHODS: A total of 180 Japanese patients with BMs were treated with WBRT between May 2008 and October 2015. We examined GPS, age, Karnofsky Performance Status (KPS), RPA, graded prognostic assessment (GPA), number of lesions, tumor size, history of brain surgery, presence of clinical symptoms, and radiation doses. RESULTS: The overall median survival time (MST) was 6.1 months. seventeen patients (9.4%) were alive more than 2 years after WBRT. In univariate analysis, KPS ≤ 70 (p = 0.0066), GPA class 0-2 (p = 0.0008), > 3 BMs (p = 0.012), > 4 BMs (p = 0.02), patients who received ≥ 3 Gy per fraction (p = 0.0068), GPS ≥ 1 (p = 0.0003), and GPS ≥ 2 (p = 0.0009) were found to significantly decrease the MST. Patients who had brain surgery before WBRT (p = 0.036) had a longer survival. On multivariate analysis, GPS ≥ 1 (p = 0.008) was found to significantly decrease MST. CONCLUSION: Our results suggest that GPS ≥ 1 indicates a poor prognosis in patients undergoing WBRT for intermediate and poor prognosis BMs.

The Impact of Pathological Grade Group 3 on Relapse-free Survival After Salvage Radiotherapy for Postoperative Prostate Cancer.

BACKGROUND/AIM: This retrospective study aimed to investigate the outcomes of relapse-free survival (RFS) after salvage radiation therapy (SRT) to the prostate bed for postoperative biochemical recurrence of prostate cancer. PATIENTS AND METHODS: A total of 87 patients were analyzed. There were 27, 32, and 24 patients with pathological grade groups of 1-2, 3, and 4-5, respectively. SRT doses of 64, 66 or 70 Gy were administered to 24, 3 and 60 patients, respectively. The Kaplan-Meier method was used to estimate time-to-event outcomes. The multiple imputations method was used to impute missing values, and Cox proportional-hazards models were applied for multivariate analyses. RESULTS: The median follow-up period for patients overall was 58.6 months. The 5-year RFS rates of the whole cohort was 59.4% and those for pathological grade groups 1-2, 3 and 4-5 were 88.9%, 37.7% and 39.5%, respectively. In multivariate analyses, higher pathological grade group [4-5 vs. 3 vs. 1-2: hazard radio (HR)=8.65, p<0.01], negative surgical resection margin (positive vs. negative: HR=0.41, p=0.02) and higher pre-salvage treatment serum prostate-specific antigen (cutoff value 0.31 ng/ml: HR=3.50, p<0.01) were significantly associated with poorer RFS. The cumulative incidences of grade 2 or more late rectal bleeding and late hematuria were 4.9% and 8.7%, respectively, at 5 years and 4.9% and 15.7%, respectively, at 8 years. These toxicities occurred only in the 70 Gy-treated arm. CONCLUSION: Our study revealed that pathological grade group 3 prostate cancer patients experienced moderately unfavorable RFS after SRT. Higher radiation doses might increase late toxicities without improving RFS.

L-DIG: A GAN-Based Method for LiDAR Point Cloud Processing under Snow Driving Conditions.

LiDAR point clouds are significantly impacted by snow in driving scenarios, introducing scattered noise points and phantom objects, thereby compromising the perception capabilities of autonomous driving systems. Current effective methods for removing snow from point clouds largely rely on outlier filters, which mechanically eliminate isolated points. This research proposes a novel translation model for LiDAR point clouds, the 'L-DIG' (LiDAR depth images GAN), built upon refined generative adversarial networks (GANs). This model not only has the capacity to reduce snow noise from point clouds, but it also can artificially synthesize snow points onto clear data. The model is trained using depth image representations of point clouds derived from unpaired datasets, complemented by customized loss functions for depth images to ensure scale and structure consistencies. To amplify the efficacy of snow capture, particularly in the region surrounding the ego vehicle, we have developed a pixel-attention discriminator that operates without downsampling convolutional layers. Concurrently, the other discriminator equipped with two-step downsampling convolutional layers has been engineered to effectively handle snow clusters. This dual-discriminator approach ensures robust and comprehensive performance in tackling diverse snow conditions. The proposed model displays a superior ability to capture snow and object features within LiDAR point clouds. A 3D clustering algorithm is employed to adaptively evaluate different levels of snow conditions, including scattered snowfall and snow swirls. Experimental findings demonstrate an evident de-snowing effect, and the ability to synthesize snow effects.

Pre-treatment Evaluation of Patients Eligible for Whole Brain Radiation Therapy: The Risk of Hippocampal Metastases in a Retrospective Study of 248 Cases at a Single Institution.

Whole brain radiation therapy (WBRT) is effective for multiple brain metastases (BMs) but may impair neurocognitive function (NCF). The incidence of hippocampal metastasis (HM) is low, and the factors associated with the occurrence of HM remain unclear. This study aimed to assess the occurrence of limbic system metastasis (LSM), including HM, and to analyze the risk of HM. We retrospectively analyzed 248 patients who underwent three-dimensional conformal radiation therapy for BMs between May 2008 and October 2015. Gadolinium-enhanced brain MRI or CT scans were used for diagnosis. Statistical analysis involved assessing clinical factors, including age, gender, primary tumor, number of BMs, and maximum metastasis diameter, in relation to the presence of HMs using logistic regression and receiver operating characteristic (ROC) curve analysis. The median age at treatment was 62 years (range: 11-83 years). Primary lesion sites included the lung (n = 150; 60.5%), breast (n = 45; 18.1%), gastrointestinal tract (n = 18; 7.3%), and bone and soft tissue (n = 2; 0.8%). Histological cancer types included adenocarcinoma (n = 113; 45.6%), squamous cell carcinoma (n = 26; 10.5%), small cell carcinoma (n = 28; 11.3%), invasive ductal carcinoma (n = 35; 14.1%), sarcoma (n = 3; 1.2%), and others (n = 43; 17.3%). MRI or CT scans of the 248 patients were analyzed, indicating a total count of 2,163 brain metastases (median: five metastases per patient). HMs were identified in 18 (7.3%) patients. The most common location for LSMs was the cingulum/cingulate gyrus in 26 (10.5%) patients. In univariate and multivariate analyses, patients with 15 or fewer BMs had a significantly lower incidence of HMs (odds ratio (OR), 0.018 (95% confidence interval (CI), 0.030-0.24)) (p < 0.0001). A maximal tumor size of less than 2 cm significantly increased the incidence of HMs (OR, 13.8 (95%CI, 1.80-105.3)) (p = 0.0003). The presence of cingulum/cingulate gyrus metastases also demonstrated a significant increase in the incidence of HMs (OR, 9.42 (95%CI, 3.30-26.84)) (p < 0.0001). The present study has uncovered a novel association between a high number of metastases in the cingulate gyrus and the development of HMs. Patients with BMs eligible for WBRT with metastases in the cingulate gyrus may be at risk of developing HM.

Controllable Unsupervised Snow Synthesis by Latent Style Space Manipulation.

In the field of intelligent vehicle technology, there is a high dependence on images captured under challenging conditions to develop robust perception algorithms. However, acquiring these images can be both time-consuming and dangerous. To address this issue, unpaired image-to-image translation models offer a solution by synthesizing samples of the desired domain, thus eliminating the reliance on ground truth supervision. However, the current methods predominantly focus on single projections rather than multiple solutions, not to mention controlling the direction of generation, which creates a scope for enhancement. In this study, we propose a generative adversarial network (GAN)-based model, which incorporates both a style encoder and a content encoder, specifically designed to extract relevant information from an image. Further, we employ a decoder to reconstruct an image using these encoded features, while ensuring that the generated output remains within a permissible range by applying a self-regression module to constrain the style latent space. By modifying the hyperparameters, we can generate controllable outputs with specific style codes. We evaluate the performance of our model by generating snow scenes on the Cityscapes and the EuroCity Persons datasets. The results reveal the effectiveness of our proposed methodology, thereby reinforcing the benefits of our approach in the ongoing evolution of intelligent vehicle technology.

Acute hematologic toxicity of radiation therapy - a comprehensive analysis and predictive nomogram.

To investigate radiation-induced cytopenia and establish predictive nomograms for hematological toxicity, we reviewed 3786 patients aged 18 or older who received radiation monotherapy between 2010 and 2021 for non-hematologic malignancies. We collected data on patient background, treatment content and hematologic toxicities for 12 weeks after the start of radiotherapy. The patients were randomly divided into training and test groups in 7:3 ratio. In the training group, we conducted ordered logistic regression analysis to identify predictive factors for neutropenia, lymphocytopenia, anemia and thrombocytopenia. Nomograms to predict Grade 2-4 cytopenia were generated and validated in the test group. Grade 3 or higher hematologic toxicities were observed in 9.7, 44.6, 8.3 and 3.1% of patients with neutropenia, lymphocytopenia, anemia and thrombocytopenia, respectively. We identified six factors for neutropenia grade, nine for lymphocytopenia grade and six for anemia grade with statistical significance. In the analysis of thrombocytopenia, the statistical model did not converge because of a small number of events. Nomograms were generated using factors with high predictive power. In evaluating the nomograms, we found high area under the receiver operating characteristic curve values (neutropenia; 0.75-0.85, lymphopenia; 0.89-0.91 and anemia; 0.85-0.86) in predicting Grade 2-4 cytopenia in the test group. We established predictive nomograms for neutropenia, leukocytopenia and anemia and demonstrated high reproducibility when validated in an independent cohort of patients.

Factors Associated with Improvement in Activities of Daily Living during Hospitalization: A Retrospective Study of Older Patients with Hip Fractures.

BACKGROUND: In this study, we aimed to examine the changes in delirium during hospitalization of patients and its association with behavioral and psychological symptoms of dementia (BPSD), as well as improvements in activities of daily living (ADL). METHODS: A longitudinal, retrospective cohort study was conducted involving 83 older adults (≥65 years) with hip fractures. We collected Mini-Mental State Examination (MMSE) and Functional Independence Measure-motor domain (m-FIM) assessment results from the medical charts at two time points: baseline (first week of hospitalization) and pre-discharge (final week before discharge). Additionally, we collected data on delirium and BPSD at three points: baseline, week 2 post-admission, and pre-discharge. We performed univariate logistic regression analysis using changes in m-FIM scores as the dependent variable and MMSE and m-FIM scores at baseline and pre-discharge, along with delirium and BPSD subtypes at baseline, week 2 post-admission, and pre-discharge, as the explanatory variables. Finally, we performed a multivariate logistic regression analysis incorporating the significant variables from the univariate analysis to identify factors associated with ADL improvement during hospitalization. RESULTS: We observed significant correlations between ADL improvement during hospitalization and baseline m-FIM and MMSE scores, hypoactive delirium state, and BPSD subtype pre-discharge. Notably, all participants with hypoactive symptoms before discharge exhibited some subtype of delirium and BPSD at baseline. CONCLUSION: Besides ADL ability and cognitive function at admission, the presence of hypoactive delirium and BPSD subtype before discharge may hinder ADL improvement during hospitalization.

Chronic encephalomyelitis virus exhibits cellular tropism and evades pDCs by binding to sialylated integrins as the cell surface receptors.

Theiler's murine encephalomyelitis virus (TMEV) causes a chronic demyelinating disease similar to multiple sclerosis in mice. Although sialic acids have been shown to be essential for TMEV attachment to the host, the surface receptor has not been identified. While type I interferons play a pivotal role in the elimination of the chronic infectious Daniel (DA) strain, the role of plasmacytoid dendritic cells (pDCs) is controversial. We herein found that TMEV binds to conventional DCs but not to pDCs. A glycomics analysis showed that the sialylated N-glycan fractions were lower in pDCs than in conventional DCs, indicating that pDCs are not susceptible to TMEV infection due to the low levels of sialic acid. TMEV capsid proteins contain an integrin recognition motif, and dot blot assays showed that the integrin proteins bind to TMEV and that the viral binding was reduced in the desialylated αX ß2 . αX ß2 protein suppressed TMEV replication in vivo, and TMEV co-localized with integrin αM at the cell membrane and TLR 3 in the cytoplasm, suggesting that αM serves as the viral attachment and entry. These results show that the chronic encephalomyelitis virus utilizes sialylated integrins as cell surface receptors, leading to cellular tropism to evade pDC activation.

Paranasal Schwannomas: A Comprehensive Study of 2 Cases.

Paranasal sinus tumors are a heterogeneous group of neoplasms (with paranasal schwannomas being a rare subtype) that are often present with non-specific symptoms, such as nasal obstruction and epistaxis. Thus, early diagnosis is crucial for optimal management. This study presents 2 cases of paranasal schwannomas, detailing their clinical presentation, diagnostic methods, and treatment approaches. Both patients underwent endoscopic sinus surgery with successful tumor excision and had no significant complications or recurrences during follow-up. Diagnosis was based on a combination of clinical examination, radiological imaging (computed tomography and magnetic resonance imaging), and histopathological confirmation with immunohistochemical staining. Treatment consisted primarily of endonasal resection, with consideration of frontal craniotomy if necessary. This study aims to contribute to the understanding of paranasal schwannomas and emphasizes the importance of early detection and treatment to improve patient outcomes.

A Rare Case of Pseudo-Malignant Paranasal Extranodal Rosai-Dorfman Disease.

Rosai-Dorfman disease is a very rare disease characterized by histiocytic accumulation in the head and neck region and lymph node enlargement. We report a rare pseudo-malignant paranasal extranodal Rosai-Dorfman disease. A 69-year-old-man presented nasal bleeding and nasal obstruction. Paranasal mass was detected in the left nasal cavity and computed tomography (CT) findings are the sphenoid sinus, maxillary sinus, and ethmoid sinus were involved with inconstant bone thickening, however, no bone destruction was detected. Magnetic resonance imaging scans show iso-intensity signal in T1-weighed image and T2-weighed image. Positron emission tomography/CT fluorodeoxyglucose (FDG) uptake in posterior ethmoid sinus and sphenoid sinus, bilateral cervical lymph node, clavicle, and sternum. Based on the above results, we considered malignant lymphoma and performed a biopsy. After pathological examination, a diagnosis of Rosai-Dorfman disease was established.