Efraín Tatis, Natalia Ramos Terrades, María Alejandra Gabaldón
et al.
Introduction: Fibrillary glomerulonephritis (FGN) has a poor prognosis and lacks standardized treatment. In this study, we describe clinicopathological characteristics, clinical course, and management of patients with FGN at our center. Materials and methods: A retrospective cohort study of 13 patients diagnosed with FGN by positive DNAJB9 at kidney biopsy between 2019 and 2024. Demographic data, comorbidities, laboratory values, histopathological findings, and treatments were collected, Complete remission was defined as proteinuria <500 mg/g with normal renal function; partial remission as >50% reduction in proteinuria to <2000 mg/g with stable renal function; and no remission as absence of improvement, progression to end-stage kidney disease (ESKD), or death. Results: The mean age was 61.4 years, 69,2% male, median proteinuria of 2318 mg/g (38% nephrotic range) and serum creatinine 1.64 mg/dL, Histopathological findings: 29% global glomerulosclerosis. The most frequent pattern was membranoproliferative (61.6%), followed by mesangial (30.8%) and membranous in 1 patient (7.7%), Hypoalbuminemia (<3 g/dL) was associated with worse prognosis (HR 6.52, p = 0.04); proteinuria >3500 mg/g and creatinine showed non-significant trends. Immunosuppression was given in 84.6% of patients: rituximab (RTX) (61.5%), RTX + mycophenolate (7.7%), sequential RTX followed by ocrelizumab and obinutuzumab (7.7%), and corticosteroids alone (7.7%). Partial remission was achieved in 30.8%, all within the exclusive RTX group, A total of 38.4% progressed to ESKD or death. Conclusions: In our study, FGN showed poor prognosis and partial response to rituximab. FGN with hypoalbuminemia had worse outcomes. Prospective studies and larger cohorts are needed to validate these findings and optimize its management. Resumen: Introducción: La glomerulonefritis fibrilar (GNF) tiene un mal pronóstico y carece de tratamiento estandarizado. Materiales y métodos: En este estudio analizamos las características clínicas y patológicas, evolución y manejo de 13 pacientes diagnosticados con GNF mediante biopsia renal positiva para DNAJB9 entre 2019 y 2024 en nuestro centro. Se recogieron datos demográficos, comorbilidades, valores de laboratorio, hallazgos histopatológicos y tratamientos. Se definió remisión completa como proteinuria <500 mg/g con función renal normal; remisión parcial como reducción >50% de proteinuria a <2000 mg/g con función renal estable; y sin remisión como ausencia de mejora, progresión a enfermedad renal crónica (ERC) estadio 5 o muerte. Resultados: La media de edad fue 61,4 años, 69,2% hombres, con una proteinuria mediana de 2318 mg/g (38% en rango nefrótico) y creatinina sérica mediana de 1,64 mg/dL. Resultados de biopsia renal: el 29% de los glomérulos estaba esclerosado. El patrón histológico más frecuente fue el membranoproliferativo (61,6%), seguido del mesangial (30,8%) y membranosa (7,7%). La hipoalbuminemia (<3 g/dL) se asoció a peor pronóstico (HR 6,52, p = 0,04). La proteinuria >3500 mg/g y creatinina mostraron tendencias no significativas. El 84,6% recibió inmunosupresión: rituximab (RTX) (61,5%), RTX + micofenolato (7,7%), RTX seguido de ocrelizumab y obinutuzumab (7,7%), y corticoides en monoterapia (7,7%). Se consiguió remisión parcial en 30,8%, todos tratados con RTX en monoterapia. Un 38,4% progresó a ERC estadio 5 o falleció. Conclusión: La GNF mostró mal pronóstico y respuesta parcial a RTX. La GNF con hipoalbuminemia presentó peor pronóstico. Se necesitan estudios prospectivos con mayores cohortes para confirmar estos hallazgos y optimizar su manejo.
Background:
A common disorder in men that has been linked to the natural aging process is benign prostatic hyperplasia (BPH). In men aged 70 and older, the frequency is approximately 40%.
Aims:
In this study, we compared the results of bipolar transurethral resection of the prostate (TURP) with our experience using holmium laser enucleation of the prostate (HoLEP).
Settings and Design:
The research was prospective and randomized. It was conducted from April 2024 to January 2025 to treat BPH.
Methodology:
Fifty participants in total had taken part in the study. Patients were randomized using the odd-even method into two groups; Group A was administered with HoLEP, and Group B with bipolar TURP. Both groups were distributed equally, with 25 patients in each.
Statistical Analysis:
For statistical analysis, an independent t-test was employed. Any P value below 0.05 was considered statistically significant.
Results:
Operative time was 70.74 ± 16.54 min in Group A, whereas 52.46 ± 12.24 min in Group B. The operative time and capsular perforation were found to be statistically significant between the groups with a P < 0.001 and 0.04, respectively. The International Prostate Symptom Score scores were 5.65 ± 2.23 in Group A and 5.43 ± 2.02 in Group B, respectively. Qmax was found to be 20.32 ± 8.7 in Group A, while 18.66 ± 5.45 in Group B.
Conclusion:
In the study, both HoLEP and TURP tend to be equally effective in terms of symptomatic relief. A steep learning curve has been seen along with less morbidity, which is associated with HoLEP.
Early detection of highly pathogenic avian influenza (HPAI) and endemic poultry diseases is critical for global food security. While computer vision models excel at classifying diseases from fecal imaging, deploying these systems at scale is bottlenecked by farm data privacy concerns and institutional data silos. Furthermore, existing open-source agricultural datasets frequently suffer from severe, undocumented data contamination. In this paper, we introduce $\textbf{FecalFed}$, a privacy-preserving federated learning framework for poultry disease classification. We first curate and release $\texttt{poultry-fecal-fl}$, a rigorously deduplicated dataset of 8,770 unique images across four disease classes, revealing and eliminating a 46.89$\%$ duplication rate in popular public repositories. To simulate realistic agricultural environments, we evaluate FecalFed under highly heterogeneous, non-IID conditions (Dirichlet $α=0.5$). While isolated single-farm training collapses under this data heterogeneity, yielding only 64.86$\%$ accuracy, our federated approach recovers performance without centralizing sensitive data. Specifically, utilizing server-side adaptive optimization (FedAdam) with a Swin-Small architecture achieves 90.31$\%$ accuracy, closely approaching the centralized upper bound of 95.10\%. Furthermore, we demonstrate that an edge-optimized Swin-Tiny model maintains highly competitive performance at 89.74$\%$, establishing a highly efficient, privacy-first blueprint for on-farm avian disease monitoring.
Motivation: Predicting gene-disease associations (GDAs) is the problem to determine which gene is associated with a disease. GDA prediction can be framed as a ranking problem where genes are ranked for a query disease, based on features such as phenotypic similarity. By describing phenotypes using phenotype ontologies, ontology-based semantic similarity measures can be used. However, traditional semantic similarity measures use only the ontology taxonomy. Recent methods based on ontology embeddings compare phenotypes in latent space; these methods can use all ontology axioms as well as a supervised signal, but are inherently transductive, i.e., query entities must already be known at the time of learning embeddings, and therefore these methods do not generalize to novel diseases (sets of phenotypes) at inference time. Results: We developed INDIGENA, an inductive disease-gene association method for ranking genes based on a set of phenotypes. Our method first uses a graph projection to map axioms from phenotype ontologies to a graph structure, and then uses graph embeddings to create latent representations of phenotypes. We use an explicit aggregation strategy to combine phenotype embeddings into representations of genes or diseases, allowing us to generalize to novel sets of phenotypes. We also develop a method to make the phenotype embeddings and the similarity measure task-specific by including a supervised signal from known gene-disease associations. We apply our method to mouse models of human disease and demonstrate that we can significantly improve over the inductive semantic similarity baseline measures, and reach a performance similar to transductive methods for predicting gene-disease associations while being more general. Availability and Implementation: https://github.com/bio-ontology-research-group/indigena
Masakuni Sakaguchi, Toshiya Maebayashi, Yuta Sekino
et al.
Abstract Introduction In recent years, a chemoradiotherapy has been developed as a radical treatment for stage II–III muscle-invasive bladder cancer (MIBC) that can preserve the bladder for patients who cannot tolerate radical cystectomy (RC) or who do not wish to undergo RC. However, most of the studies were conducted on younger patients with MIBC, and it is not clear if it is effective for elderly patients with MIBC. In this study, we reviewed the effects and adverse events after radical radiotherapy in elderly patients with MIBC to determine if radiotherapy has been/can be equally recommended for younger patients with MIBC. Methods We extracted full research reports in English comparing treatment results between different age groups and reports targeting elderly patients with MIBC. A keyword search of the PubMed database was conducted in the decade ending on December 8, 2021. Studies reporting post-treatment overall survival (OS), relapse-free/progression-free/disease-free survival (RFS/PFS/DFS), disease-specific/cancer-specific survival (DSS/CSS), and complete response (CR) rate, adverse events (AEs), and quality of life (QOL) in elderly patients with MIBC were searched. Thirty-nine full articles, including those with comparisons by age group or treatments for elderly patients, were retrieved. Results OS was significant or tended to be poor in elderly patients. There were no differences in PFS and CSS between younger and elderly patients. No differences in the rates of grade 3 morbidities between younger and elderly patients were also observed. Conclusion The lack of a difference in PFS/CSS and toxicities between elderly and younger MIBC patients indicated that curative chemoradiotherapy is effective for not only younger but also elderly patients. With advances in treatment, further prospective studies are needed to optimize the management of MIBC in elderly patients.
Ocular disease affects billions of individuals unevenly worldwide. It continues to increase in prevalence with trends of growing populations of diabetic people, increasing life expectancies, decreasing ophthalmologist availability, and rising costs of care. We present EyeAI, a system designed to provide artificial intelligence-assisted detection of ocular diseases, thereby enhancing global health. EyeAI utilizes a convolutional neural network model trained on 1,920 retinal fundus images to automatically diagnose the presence of ocular disease based on a retinal fundus image input through a publicly accessible web-based application. EyeAI performs a binary classification to determine the presence of any of 45 distinct ocular diseases, including diabetic retinopathy, media haze, and optic disc cupping, with an accuracy of 80%, an AUROC of 0.698, and an F1-score of 0.8876. EyeAI addresses barriers to traditional ophthalmologic care by facilitating low-cost, remote, and real-time diagnoses, particularly for equitable access to care in underserved areas and for supporting physicians through a secondary diagnostic opinion. Results demonstrate the potential of EyeAI as a scalable, efficient, and accessible diagnostic tool. Future work will focus on expanding the training dataset to enhance the accuracy of the model further and improve its diagnostic capabilities.
Bias in recommender systems not only distorts user experience but also perpetuates and amplifies existing societal stereotypes, particularly in sectors like fashion e-commerce. This study employs a dynamic modeling approach to scrutinize the mechanisms of bias activation and reinforcement within Fashion Recommender Systems (FRS). By leveraging system dynamics modeling and experimental simulations, we dissect the temporal evolution of bias and its multifaceted impacts on system performance. Our analysis reveals that inductive biases exert a more substantial influence on system outcomes than user biases, suggesting critical areas for intervention. We demonstrate that while current debiasing strategies, including data rebalancing and algorithmic regularization, are effective to an extent, they require further enhancement to comprehensively mitigate biases. This research underscores the necessity for advancing these strategies and extending system boundaries to incorporate broader contextual factors such as user demographics and item diversity, aiming to foster inclusivity and fairness in FRS. The findings advocate for a proactive approach in recommender system design to counteract bias propagation and ensure equitable user experiences.
Oskar Triebe, Fletcher Passow, Simon Wittner
et al.
The reliability of local power grid infrastructure is challenged by sustainable energy developments increasing electric load uncertainty. Transmission System Operators (TSOs) need load forecasts of higher spatial resolution, extending current forecasting operations from zonal aggregates to individual nodes. However, nodal loads are less accurate to forecast and require a large number of individual forecasts, which are hard to manage for the human experts assessing risks in the control room's daily operations (operator). In collaboration with a TSO, we design a multi-level system that meets the needs of operators for hourly day-ahead load forecasting. Utilizing a uniquely extensive dataset of zonal and nodal net loads, we experimentally evaluate our system components. First, we develop an interpretable and scalable forecasting model that allows for TSOs to gradually extend zonal operations to include nodal forecasts. Second, we evaluate solutions to address the heterogeneity and volatility of nodal load, subject to a trade-off. Third, our system is manageable with a fully parallelized single-model forecasting workflow. Our results show accuracy and interpretability improvements for zonal forecasts, and substantial improvements for nodal forecasts. In practice, our multi-level forecasting system allows operators to adjust forecasts with unprecedented confidence and accuracy, and to diagnose otherwise opaque errors precisely.
Ananya Joshi, George Khoury, Christodoulas Floudas
Alzheimer's disease is characterized by dangerous amyloid plaques formed by deposits of the protein Beta-Amyloid aggregates in the brain. The specific amino acid sequence that is responsible for the aggregates of Beta-Amyloid is lys-leu-val-phe-phe (KLVFF). KLVFF aggregation inhibitors, which we design in this paper, prevent KLVFF from binding with itself to form oligomers or fibrils (and eventually plaques) that cause neuronal death. Our binder-blocker peptides are designed such that, on one side, they bind strongly to KLVFF, and on the other side, they disrupt critical interactions, thus preventing aggregation. Our methods use optimization techniques and molecular simulations and identify 10 candidate sequences for trial of the 3.2 million possible sequences. This approach for inhibitor identification can be generalized to other diseases characterized by protein aggregation, such as Parkinson's, Huntington's, and prion diseases.
Francisco Maduell, José Jesús Broseta, Joaquim Casals
et al.
Introduction: The sodium gradient during hemodialysis sessions is one of the key factors in sodium balance in patients with dialysis-dependent chronic kidney disease; however, until the appearance of the new monitors with sodium modules, the differences between prescribed and measured sodium have been understudied. The present study aimed to compare the impact on the measured conductivity and the initial and final plasma sodium after changing the 5008 Cordiax to the new 6008 Cordiax monitor. Material and methods: 106 patients on hemodialysis were included. Each patient underwent 2 dialysis sessions in which only the monitor was varied. The variables collected were dialysate, sodium and bicarbonate prescribed, real conductivity, initial and final plasma sodium measured, and the calculated sodium gradient (ΔPNa). Results: The change of dialysis monitor showed small but statistically significant differences in the initial (138.14 mmol/L with 5008 vs. 138.81 mmol/L with 6008) and final plasma sodium (139.58 mmol/L vs. 140.97 mmol/L), as well as in the actual conductivity obtained (13.97 vs. 14.1 mS/cm). The ΔPNa also increased significantly. Conclusion: The change from 5008 to 6008 monitor is associated with increased conductivity, leading the patient to end the sessions with higher plasma sodium and ΔPNa. Knowing and confirming this change will allow us to individualize the sodium prescription and avoid possible undesirable effects. It could be the preliminary study to explore the new sodium biosensor incorporated into the new generation of monitors. Resumen: Introducción: El gradiente de sodio durante las sesiones es uno de los factores clave en el balance de este ion en los pacientes en hemodiálisis; sin embargo, hasta la aparición de los nuevos monitores con módulos de sodio, las diferencias entre el sodio prescrito y el medido han sido poco estudiadas. El objetivo del presente estudio fue comparar el impacto del cambio del monitor 5008 Cordiax al nuevo monitor 6008 Cordiax sobre los resultados de la conductividad real medida, del sodio plasmático inicial y final. Material y métodos: Se incluyeron 106 pacientes en hemodiálisis. Cada paciente recibió 2 sesiones de diálisis en la que solo se varió el monitor. Las variables recogidas fueron: el concentrado, sodio y bicarbonato prescritos, conductividad real, sodio plasmático inicial y final medidos por dialisancia iónica y se calculó el cambio de la concentración de sodio plasmático durante el tratamiento o delta de sodio (ΔPNa). Resultados: El cambio de monitor de diálisis mostró pequeñas diferencias, aunque significativas, en el sodio plasmático inicial (138.14 mmol/L con 5008 vs 138.81 mmol/L con 6008) y final (139.58 mmol/L vs 140.97 mmol/L), así como en la conductividad real obtenida (13,97 vs 14,10 mS/cm). El ΔPNa también aumento significativamente. Conclusión: El cambio de monitor 5008 a 6008 se asocia a un aumento en la conductividad, un sodio plasmático más elevado y un incremento en el ΔPNa. El conocer y confirmar este cambio permitirá individualizar la prescripción de sodio, evitar posibles efectos indeseables y podría ser el estudio preliminar para explorar el nuevo biosensor de control de sodio incorporado en la nueva generación de monitores.
Raghav Singhal, Mukund Sudarshan, Anish Mahishi
et al.
Early detection of many life-threatening diseases (e.g., prostate and breast cancer) within at-risk population can improve clinical outcomes and reduce cost of care. While numerous disease-specific "screening" tests that are closer to Point-of-Care (POC) are in use for this task, their low specificity results in unnecessary biopsies, leading to avoidable patient trauma and wasteful healthcare spending. On the other hand, despite the high accuracy of Magnetic Resonance (MR) imaging in disease diagnosis, it is not used as a POC disease identification tool because of poor accessibility. The root cause of poor accessibility of MR stems from the requirement to reconstruct high-fidelity images, as it necessitates a lengthy and complex process of acquiring large quantities of high-quality k-space measurements. In this study we explore the feasibility of an ML-augmented MR pipeline that directly infers the disease sidestepping the image reconstruction process. We hypothesise that the disease classification task can be solved using a very small tailored subset of k-space data, compared to image reconstruction. Towards that end, we propose a method that performs two tasks: 1) identifies a subset of the k-space that maximizes disease identification accuracy, and 2) infers the disease directly using the identified k-space subset, bypassing the image reconstruction step. We validate our hypothesis by measuring the performance of the proposed system across multiple diseases and anatomies. We show that comparable performance to image-based classifiers, trained on images reconstructed with full k-space data, can be achieved using small quantities of data: 8% of the data for detecting multiple abnormalities in prostate and brain scans, and 5% of the data for knee abnormalities. To better understand the proposed approach and instigate future research, we provide an extensive analysis and release code.
Francis G. T. Kamba, Leonard C. Eze, Jean Claude Kamgang
et al.
We propose and analyze an epidemiological model for vector borne diseases that integrates a multi-stage vector population and several host sub-populations which may be characterized by a variety of compartmental model types: subpopulations all include Susceptible and Infected compartments, but may or may not include Exposed and/or Recovered compartments. The model was originally designed to evaluate the effectiveness of various prophylactic measures in malaria-endemic areas, but can be applied as well to other vector-borne diseases. This model is expressed as a system of several differential equations, where the number of equations depends on the particular assumptions of the model. We compute the basic reproduction number $\mathcal R_0$, and show that if $\mathcal R_0\leqslant 1$, the disease free equilibrium (DFE) is globally asymptotically stable (GAS) on the nonnegative orthant. If $\mathcal R_0>1$, the system admits a unique endemic equilibrium (EE) that is GAS. We analyze the sensitivity of $R_0$ and the EE to different system parameters, and based on this analysis we discuss the relative effectiveness of different control measures.
Family history is considered a risk factor for many diseases because it implicitly captures shared genetic, environmental and lifestyle factors. Finland's nationwide electronic health record (EHR) system spanning multiple generations presents new opportunities for studying a connected network of medical histories for entire families. In this work we present a graph-based deep learning approach for learning explainable, supervised representations of how each family member's longitudinal medical history influences a patient's disease risk. We demonstrate that this approach is beneficial for predicting 10-year disease onset for 5 complex disease phenotypes, compared to clinically-inspired and deep learning baselines for Finland's nationwide EHR system comprising 7 million individuals with up to third-degree relatives. Through the use of graph explainability techniques, we illustrate that a graph-based approach enables more personalized modeling of family information and disease risk by identifying important relatives and features for prediction.
Gauri Naik, Nandini Narvekar, Dimple Agarwal
et al.
Eye diseases have posed significant challenges for decades, but advancements in technology have opened new avenues for their detection and treatment. Machine learning and deep learning algorithms have become instrumental in this domain, particularly when combined with Optical Coherent Technology (OCT) imaging. We propose a novel method for efficient detection of eye diseases from OCT images. Our technique enables the classification of patients into disease free (normal eyes) or affected by specific conditions such as Choroidal Neovascularization (CNV), Diabetic Macular Edema (DME), or Drusen. In this work, we introduce an end to end web application that utilizes machine learning and deep learning techniques for efficient eye disease prediction. The application allows patients to submit their raw OCT scanned images, which undergo segmentation using a trained custom UNet model. The segmented images are then fed into an ensemble model, comprising InceptionV3 and Xception networks, enhanced with a self attention layer. This self attention approach leverages the feature maps of individual models to achieve improved classification accuracy. The ensemble model's output is aggregated to predict and classify various eye diseases. Extensive experimentation and optimization have been conducted to ensure the application's efficiency and optimal performance. Our results demonstrate the effectiveness of the proposed approach in accurate eye disease prediction. The developed web application holds significant potential for early detection and timely intervention, thereby contributing to improved eye healthcare outcomes.
This technical report describes our system for track 1, 2 and 4 of the VoxCeleb Speaker Recognition Challenge 2022 (VoxSRC-22). By combining several ResNet variants, our submission for track 1 attained a minDCF of 0:090 with EER 1:401%. By further incorporating three fine-tuned pre-trained models, our submission for track 2 achieved a minDCF of 0:072 with EER 1:119%. For track 4, our system consisted of voice activity detection (VAD), speaker embedding extraction, agglomerative hierarchical clustering (AHC) followed by a re-clustering step based on a Bayesian hidden Markov model and overlapped speech detection and handling. Our submission for track 4 achieved a diarisation error rate (DER) of 4.86%. The submissions all ranked the 2nd places for the corresponding tracks.
Eman Yahia Alqaissi, Fahd Saleh Alotaibi, Muhammad Sher Ramzan
Controlling infectious diseases is a major health priority because they can spread and infect humans, thus evolving into epidemics or pandemics. Therefore, early detection of infectious diseases is a significant need, and many researchers have developed models to diagnose them in the early stages. This paper reviewed research articles for recent machine-learning (ML) algorithms applied to infectious disease diagnosis. We searched the Web of Science, ScienceDirect, PubMed, Springer, and IEEE databases from 2015 to 2022, identified the pros and cons of the reviewed ML models, and discussed the possible recommendations to advance the studies in this field. We found that most of the articles used small datasets, and few of them used real-time data. Our results demonstrated that a suitable ML technique depends on the nature of the dataset and the desired goal.