Eva B. Aamodt, Martin Røvang, Mona K. Beyer
et al.
BackgroundMeasures of white matter hyperintensities (WMHs) represent a crucial part of post-stroke outcome prediction. Automatic WMH segmentation has proven particularly challenging in stroke cases. Using an improved method for WMH segmentation that incorporates stroke lesions, we set out to explore factors associated with higher WMH burden, as well as the association between WMH burden and post-stroke dependency across two different countries that may demonstrate significant variation in radiological presentation.MethodsA total of 384 acute ischemic stroke (AIS) survivors from the Norwegian Cognitive Impairment After Stroke (Nor-COAST; NO) study and the Houston Methodist Registry of Neurological Endpoint Assessments among Patients with Ischemic and Hemorrhagic Stroke (REINAH; US) database were analyzed. MRI and clinical data were collected upon acute care hospital admission. WMHs were measured automatically using the nnU-Net methodology, taking into account the acute stroke lesion.ResultsNo significant difference in WMH percentage was found between sites. Factors associated with higher WMH burden included only age in NO, while in US, very high age (≥ 85), smoking, and being underweight were key factors. The two sites showed significant differences in demographics and clinical characteristics: the US cohort exhibited greater racial heterogeneity, higher body mass index (BMI) with more extremely obese patients, higher National Institutes of Health Stroke Scale (NIHSS) scores, and more thrombectomies, whereas the NO cohort exhibited more tobacco use, hypercholesterolemia, and longer stay at the hospital. Post-stroke dependency was initially associated with higher WMH percentage overall but only remained significant after adjusment in Norwegians aged ≥85, while in the US, dependency was driven by stroke severity and treatment after adjustment.ConclusionCohorts from the US and Norway exhibit no significant difference in WMH burden, but differ in the factors associated with WMHs.
Background Physical activity volume (PAV) has been linked to a wide range of health outcomes; however, its association with incident nervous system diseases remains incompletely understood. This study explored their relationship using data from UK Biobank. Methods A retrospective cohort study was conducted involving 278,306 participants from the UK Biobank. PAV was quantified as metabolic equivalent of task (MET) minutes per week, derived from self-reported physical activity levels, and categorized into three groups. Incident cases of nervous system diseases were identified through ICD-10 codes obtained from hospital inpatient records, death registries, and self-reports. Cox proportional hazards models were employed to estimate hazard ratios (HRs) and 95% confidence intervals (CIs), adjusting for a series of covariates. Restricted cubic splines were applied to assess potential non-linear associations. Results Women and individuals aged 60 years or older exhibited higher incidence rates of nervous system diseases. After multivariable adjustment, higher PAV was significantly associated with a lower risk of overall nervous system diseases (HR < 1). A non-linear dose–response relationship was observed, with the lowest risk occurring at a PAV level of 1,356 MET-min/week. Subgroup analyses indicated that elevated PAV conferred protective effects against several specific conditions. Conversely, higher PAV was associated with an increased risk of meningitis. Conclusion Increased levels of physical activity are associated with a reduced risk of numerous nervous system diseases, with optimal protection observed at approximately 1,356 MET-min/week. These findings support the promotion of moderate-to-vigorous physical activity as a preventive strategy for neurological disorders, particularly among high-risk populations.
Hepato-pancreato-biliary (HPB) disorders represent a global public health challenge due to their high morbidity and mortality. Although large language models (LLMs) have shown promising performance in general medical question-answering tasks, the current evaluation benchmarks are mostly derived from standardized examinations or manually designed questions, lacking HPB coverage and clinical cases. To address these issues, we systematically eatablish an HPB disease evaluation benchmark comprising 3,535 closed-ended multiple-choice questions and 337 open-ended real diagnosis cases, which encompasses all the 33 main categories and 465 subcategories of HPB diseases defined in the International Statistical Classification of Diseases, 10th Revision (ICD-10). The multiple-choice questions are curated from public datasets and synthesized data, and the clinical cases are collected from prestigious medical journals, case-sharing platforms, and collaborating hospitals. By evalauting commercial and open-source general and medical LLMs on our established benchmark, namely ClinBench-HBP, we find that while commercial LLMs perform competently on medical exam questions, they exhibit substantial performance degradation on HPB diagnosis tasks, especially on complex, inpatient clinical cases. Those medical LLMs also show limited generalizability to HPB diseases. Our results reveal the critical limitations of current LLMs in the domain of HPB diseases, underscoring the imperative need for future medical LLMs to handle real, complex clinical diagnostics rather than simple medical exam questions. The benchmark will be released at https://clinbench-hpb.github.io.
In this study, we explored the use of spectrograms to represent handwriting signals for assessing neurodegenerative diseases, including 42 healthy controls (CTL), 35 subjects with Parkinson's Disease (PD), 21 with Alzheimer's Disease (AD), and 15 with Parkinson's Disease Mimics (PDM). We applied CNN and CNN-BLSTM models for binary classification using both multi-channel fixed-size and frame-based spectrograms. Our results showed that handwriting tasks and spectrogram channel combinations significantly impacted classification performance. The highest F1-score (89.8%) was achieved for AD vs. CTL, while PD vs. CTL reached 74.5%, and PD vs. PDM scored 77.97%. CNN consistently outperformed CNN-BLSTM. Different sliding window lengths were tested for constructing frame-based spectrograms. A 1-second window worked best for AD, longer windows improved PD classification, and window length had little effect on PD vs. PDM.
Lateral ventricle (LV) shape analysis holds promise as a biomarker for neurological diseases; however, challenges remain due to substantial shape variability across individuals and segmentation difficulties arising from limited MRI resolution. We introduce LV-Net, a novel framework for producing individualized 3D LV meshes from brain MRI by deforming an anatomy-aware joint LV-hippocampus template mesh. By incorporating anatomical relationships embedded within the joint template, LV-Net reduces boundary segmentation artifacts and improves reconstruction robustness. In addition, by classifying the vertices of the template mesh based on their anatomical adjacency, our method enhances point correspondence across subjects, leading to more accurate LV shape statistics. We demonstrate that LV-Net achieves superior reconstruction accuracy, even in the presence of segmentation imperfections, and delivers more reliable shape descriptors across diverse datasets. Finally, we apply LV-Net to Alzheimer's disease analysis, identifying LV subregions that show significantly associations with the disease relative to cognitively normal controls. The codes for LV shape modeling are available at https://github.com/PWonjung/LV_Shape_Modeling.
Umakanta Maharana, Sarthak Verma, Avarna Agarwal
et al.
Large language models (LLMs) offer a promising pre-screening tool, improving early disease detection and providing enhanced healthcare access for underprivileged communities. The early diagnosis of various diseases continues to be a significant challenge in healthcare, primarily due to the nonspecific nature of early symptoms, the shortage of expert medical practitioners, and the need for prolonged clinical evaluations, all of which can delay treatment and adversely affect patient outcomes. With impressive accuracy in prediction across a range of diseases, LLMs have the potential to revolutionize clinical pre-screening and decision-making for various medical conditions. In this work, we study the diagnostic capability of LLMs for Rheumatoid Arthritis (RA) with real world patients data. Patient data was collected alongside diagnoses from medical experts, and the performance of LLMs was evaluated in comparison to expert diagnoses for RA disease prediction. We notice an interesting pattern in disease diagnosis and find an unexpected \textit{misalignment between prediction and explanation}. We conduct a series of multi-round analyses using different LLM agents. The best-performing model accurately predicts rheumatoid arthritis (RA) diseases approximately 95\% of the time. However, when medical experts evaluated the reasoning generated by the model, they found that nearly 68\% of the reasoning was incorrect. This study highlights a clear misalignment between LLMs high prediction accuracy and its flawed reasoning, raising important questions about relying on LLM explanations in clinical settings. \textbf{LLMs provide incorrect reasoning to arrive at the correct answer for RA disease diagnosis.}
Samantha Min Er Yew, Xiaofeng Lei, Jocelyn Hui Lin Goh
et al.
Background: RETFound, a self-supervised, retina-specific foundation model (FM), showed potential in downstream applications. However, its comparative performance with traditional deep learning (DL) models remains incompletely understood. This study aimed to evaluate RETFound against three ImageNet-pretrained supervised DL models (ResNet50, ViT-base, SwinV2) in detecting ocular and systemic diseases. Methods: We fine-tuned/trained RETFound and three DL models on full datasets, 50%, 20%, and fixed sample sizes (400, 200, 100 images, with half comprising disease cases; for each DR severity class, 100 and 50 cases were used. Fine-tuned models were tested internally using the SEED (53,090 images) and APTOS-2019 (3,672 images) datasets and externally validated on population-based (BES, CIEMS, SP2, UKBB) and open-source datasets (ODIR-5k, PAPILA, GAMMA, IDRiD, MESSIDOR-2). Model performance was compared using area under the receiver operating characteristic curve (AUC) and Z-tests with Bonferroni correction (P<0.05/3). Interpretation: Traditional DL models are mostly comparable to RETFound for ocular disease detection with large datasets. However, RETFound is superior in systemic disease detection with smaller datasets. These findings offer valuable insights into the respective merits and limitation of traditional models and FMs.
ObjectiveThis study aimed to analyze the clinical characteristics and prognosis of patients with autoimmune encephalitis (PWAE) who experienced seizures during the acute phase.MethodsClinical data were collected from 84 patients diagnosed with AE at the General Hospital of Ningxia Medical University between January 2015 and January 2023. Patients were divided into seizure and non-seizure groups. Clinical characteristics of both groups were compared, including differences between anti-NMDAR and anti-LGI1 encephalitis within the seizure group. Due to the limited sample size and to avoid overfitting, we focused on univariate logistic regression analysis to identify individual prognostic factors.ResultsA total of 84 patients were enrolled, with 76.19% (64/84) in the seizure group and 23.81% (20/84) in the non-seizure group. The seizure group had a longer hospital stay (p = 0.013), higher rates of impaired consciousness (p = 0.001), and more frequent intensive care unit (ICU) admission (p = 0.011). They also had higher peripheral blood neutrophil-to-lymphocyte ratio (NLR), leukocyte count, and uric acid levels (p = 0.038, p = 0.006, p = 0.020), and were more likely to show slow-wave rhythms on electroencephalography (EEG) (p = 0.031). At 2-year follow-up, there was no significant difference in prognosis between the seizure and non-seizure groups (p = 0.653), with 35.94% (23/64) of the seizure group having a poor prognosis. Status epilepticus (SE), complications, endotracheal intubation, mRS score at discharge, APE2, and RITE2 scores increased the risk of poor prognosis (OR > 1), while intensive care and albumin reduced the risk (OR < 1).ConclusionSeizures are common in the early stages of AE, with faciobrachial dystonic seizures (FBDS) characteristic of anti-LGI1 encephalitis and SE and super-refractory status epilepticus (Sup-RSE) frequently observed in anti-NMDAR encephalitis. Seizure semiology across AE subtypes lacks specificity, and no symptoms clearly distinguish immune-mediated from non-immune causes. While seizures are linked to AE severity, particularly in anti-NMDAR encephalitis, they do not appear to impact overall prognosis. SE, complications, endotracheal intubation, modified Rankin Scale (mRS) score at discharge, Antibody-Prevalence in Epilepsy and Encephalopathy (APE2) score, Response to Immunotherapy in Epilepsy and Encephalopathy (RITE2) score, intensive care, and albumin were identified as significant prognostic factors.
Barbara Carpita, Benedetta Nardi, Chiara Bonelli
et al.
IntroductionDue to their similar behavioral presentation, it can sometimes be challenging to distinguish between a social anxiety disorder (SAD) and the social avoidance that is frequently described in autism spectrum disorder (ASD). Moreover, a growing body of evidences is reporting that a significant proportion of subjects with ASD also meet the requirements for SAD and, vice versa, subjects with SAD tend to exhibit a higher prevalence of autistic traits.AimIn this framework, the current study aims to evaluate prevalence and correlates of autistic traits in a sample of adult subjects diagnosed with SAD and healthy controls (HC), also evaluating which autism spectrum dimensions may statistically predict higher SAD symptoms.Methods56 subjects with a clinical diagnosis of SAD and 56 gender and age matched HC were recruited from the Psychiatric Clinic of the University of Pisa. Subjects were assessed with the SCID-5, the Social Anxiety Spectrum – Short Version (SHY- SV) and the Adult Autism Subthreshold Spectrum (AdAS Spectrum).ResultsSAD group scored significantly higher in all AdAS Spectrum and SHY-SV domains and total score compared to the HC group with no significant gender difference. SHY-SV total and domain scores, were strongly and positively and strongly correlated with all AdAS Spectrum domains and total score. AdAS Spectrum total score and Childhood/Adolescence, Non-Verbal Communication, Empathy and Restricted interests and Rumination domain scores score were significant predictors of higher SHY-SV score.ConclusionOur results confirm the link between SAD and autistic traits also in adult population, describing not only high levels of autistic traits in SAD adults, but also significant correlations between many core features of the two disorders and a predictive role of autistic traits on higher SAD symptoms.
Background: Aim of this study was to establish acceptable cut off values of psoas muscle
area (PMA) for evaluation of low muscle mass in postmenopausal osteoporotic women.
Methods and Materials: Ninety-five women with postmenopausal osteoporosis who had
underwent lumbar spine MRI and dual energy x ray absorptiometry were retrieved
retrospectively. Psoas muscle cross-sectional area (CSA) and index were measured at L3-
level and cut off values of PMA were investigated for age groups. Results: Of the ninety-five
women with postmenopausal osteoporosis (63.96 ± 9.19 years), the rate of sarcopenia was
16.7% (n= 10) in the ≤ 65 age group, 22.2% (n= 10) in the > 65 age group and 19% (n= 20) in
total. The cut off values of PMA were determined as 468.55 mm2 and 479.20 mm2 for ≤ 65
and > 65 years of age groups, respectively. The evaluation of intra-observer reliability
resulted almost perfect with intraclass correlation coefficient ranging from 0.997 to 0.999
(p<0.001). The inter-observer reliability was also almost perfect with intraclass correlation
coefficients ranging from 0.994 to 0.999(p<0.001). Conclusions: This study has provided cut
off values for MRI defined PMA in postmenopausal osteoporotic women. Further
longitudinal studies are required to confirm whether these cut offs are successful in
predicting mortality and other adverse outcomes.
Cardiovascular diseases (CVDs) are notably prevalent among patients with obstructive sleep apnea (OSA), posing unique challenges in predicting CVD progression due to the intricate interactions of comorbidities. Traditional models typically lack the necessary dynamic and longitudinal scope to accurately forecast CVD trajectories in OSA patients. This study introduces a novel multi-level phenotypic model to analyze the progression and interplay of these conditions over time, utilizing data from the Wisconsin Sleep Cohort, which includes 1,123 participants followed for decades. Our methodology comprises three advanced steps: (1) Conducting feature importance analysis through tree-based models to underscore critical predictive variables like total cholesterol, low-density lipoprotein (LDL), and diabetes. (2) Developing a logistic mixed-effects model (LGMM) to track longitudinal transitions and pinpoint significant factors, which displayed a diagnostic accuracy of 0.9556. (3) Implementing t-distributed Stochastic Neighbor Embedding (t-SNE) alongside Gaussian Mixture Models (GMM) to segment patient data into distinct phenotypic clusters that reflect varied risk profiles and disease progression pathways. This phenotypic clustering revealed two main groups, with one showing a markedly increased risk of major adverse cardiovascular events (MACEs), underscored by the significant predictive role of nocturnal hypoxia and sympathetic nervous system activity from sleep data. Analysis of transitions and trajectories with t-SNE and GMM highlighted different progression rates within the cohort, with one cluster progressing more slowly towards severe CVD states than the other. This study offers a comprehensive understanding of the dynamic relationship between CVD and OSA, providing valuable tools for predicting disease onset and tailoring treatment approaches.
Asish Bera, Debotosh Bhattacharjee, Ondrej Krejcar
Crop yield production could be enhanced for agricultural growth if various plant nutrition deficiencies, and diseases are identified and detected at early stages. The deep learning methods have proven its superior performances in the automated detection of plant diseases and nutrition deficiencies from visual symptoms in leaves. This article proposes a new deep learning method for plant nutrition deficiencies and disease classification using a graph convolutional network (GNN), added upon a base convolutional neural network (CNN). Sometimes, a global feature descriptor might fail to capture the vital region of a diseased leaf, which causes inaccurate classification of disease. To address this issue, regional feature learning is crucial for a holistic feature aggregation. In this work, region-based feature summarization at multi-scales is explored using spatial pyramidal pooling for discriminative feature representation. A GCN is developed to capacitate learning of finer details for classifying plant diseases and insufficiency of nutrients. The proposed method, called Plant Nutrition Deficiency and Disease Network (PND-Net), is evaluated on two public datasets for nutrition deficiency, and two for disease classification using four CNNs. The best classification performances are: (a) 90.00% Banana and 90.54% Coffee nutrition deficiency; and (b) 96.18% Potato diseases and 84.30% on PlantDoc datasets using Xception backbone. Furthermore, additional experiments have been carried out for generalization, and the proposed method has achieved state-of-the-art performances on two public datasets, namely the Breast Cancer Histopathology Image Classification (BreakHis 40X: 95.50%, and BreakHis 100X: 96.79% accuracy) and Single cells in Pap smear images for cervical cancer classification (SIPaKMeD: 99.18% accuracy). Also, PND-Net achieves improved performances using five-fold cross validation.
Charina C. Lüder, Tanja Michael, Johanna Lass-Hennemann
et al.
ABSTRACTBackground: Refugees with exposure to multiple traumatic events are at high risk for developing posttraumatic stress disorder (PTSD) and depression. Narrative exposure therapy (NET) is an effective treatment for the core symptoms of PTSD, but it does not reliably reduce depressive symptoms. Endurance exercise on the other hand was consistently found to be effective in treating depression making it a promising adjunct to NET. Up to date, no studies exist investigating the combination of NET and endurance exercise in a sample of refugees with PTSD and comorbid depression.Objectives: In the proposed randomized controlled trial, we aim to investigate whether a combination of NET and moderate-intensity aerobic exercise training (MAET) enhances treatment outcome for refugees with PTSD and comorbid depressive symptoms. We expect a greater improvement in psychopathology in participants who receive the combined treatment.Methods and analysis: 68 refugees and asylum seekers with PTSD and clinically relevant depressive symptoms will be recruited in the proposed study. Participants will be randomly assigned to receive either NET only (NET-group) or NET plus MAET (NET+-group). All participants will receive 10 NET sessions. Participants in the NET+-group will additionally take part in MAET. Primary (PTSD, depression) and secondary (general mental distress, agoraphobia and somatoform complaints, sleep quality) outcome measures will be assessed before treatment, after treatment, and at six-month follow-up. The hypotheses will be tested with multiple 2 × 3 mixed ANOVA's.Trial registration: German Clinical Trials Register identifier: DRKS00022145.
Objective: Bipolar disorder (BD) progresses in episodes and includes periods of remission between episodes. Although symptoms are decreased during remission, patients' functioning levels may be lower than the pre-disease period. This study aims to evaluate the relationship between the level of functioning and exercise addiction, temperament and personality traits in patients with BD in remission.
Methods:In this study, BD patients who have been in remission for three months and more were evaluated. Young Mani Rating Scale and Hamilton Depression Rating Scale were used to determine the status of remission. Bipolar Disorder Functioning Scale was used for functioning, Eysenck Personality Survey-Revised Short Form for personality, TEMPS-A temperament scale for temperament, and the Exercise Addiction Scale was used for exercise addiction.
Results:92 BD patients in remission were included in the study. A positive relationship was found between functioning level and hyperthymic temperament and extraversion, and a negative relationship was found between depressive temperament, cyclothymic temperament, irritable temperament, anxious temperament, neuroticism, and psychoticism. Although these patients had a certain degree of exercise dependence, it did not have a relationship with functioning.
Conclusion: Functionality may be affected by personality and temperament characteristics during remission periods in BD. Psychosocial interventions in these areas may be beneficial in the treatment of these patients. [PBS 2023; 13(2.000): 71-77]
AimThis retrospective study aimed to investigate the independent clinical variables associated with the onset of acute cerebral ischemic stroke (AIS) in patients with stable chronic obstructive pulmonary disease (COPD).MethodA total of 244 patients with COPD who had not experienced a relapse within 6 months were included in this retrospective study. Of these, 94 patients hospitalized with AIS were enrolled in the study group, and the remaining 150 were enrolled in the control group. Clinical data and laboratory parameters were collected for both groups within 24 h after hospitalization, and the data of the two groups were statistically analyzed.ResultsThe levels of age, white blood cell (WBC), neutrophil (NEUT), glucose (GLU), prothrombin time (PT), albumin (ALB), and red blood cell distribution width (RDW) were different in the two groups (P < 0.01). Logistic regression analysis showed that age, WBC, RDW, PT, and GLU were independent risk factors for the occurrence of AIS in patients with stable COPD. Age and RDW were selected as new predictors, and the receiver operating characteristic curves (ROC) were plotted accordingly. The areas under the ROC curves of age, RDW, and age + RDW were 0.7122, 0.7184, and 0.7852, respectively. The sensitivity was 60.5, 59.6, and 70.2%, and the specificity was 72.4, 86.0, and 60.0%, respectively.ConclusionThe combination of RDW and age in patients with stable COPD might be a potential predictor for the onset of AIS.