Disease Classification Research Articles

Emergency hand surgery in the course of time : Clinical and economic aspects of inpatient and outpatient cases at auniversity hospital in the last decade

Hand surgery departments nowadays face the challenge of operating profitably, training new specialists, managing increasing case numbers and implementing the required shift towards outpatient care in clinical practice. The aim of this study was to describe the demographic and economic indicators of auniversity hospital and to examine the development of inpatient and outpatient case numbers over the past decade. Aretrospective data analysis was carried out for patients treated for hand injuries in the period from 2013 to 2023. The main diagnoses of cases were identified and descriptively analyzed using the International Classification of Diseases (ICD). In order to demonstrate acorrelation, alinear regression model was calculated and ap-value <0.05 was considered significant. In total, 2918 cases were included over adecade starting in 2013. Out of these 776 cases were treated on an outpatient basis and two thirds of the patients were male. The patients were generally healthy with an average American Society of Anesthesiologists (ASA) status of 1.6 ± 0.6 and with apatient clinical complexity level (PCCL) of 0 in 79% of cases. Patient numbers increased continuously from 161 cases in 2013 to 393 cases in 2022 (p < 0.001). The most common main diagnosis was infectious hand diseases (L02.4-M65.14). This resulted in the most frequently assigned diagnosis-related group (DRG) being I32E (319 cases). For inpatient cases the average incision-to-suture time was 51:51 min and for outpatient cases 26:03 min. The revenue was on average 4372.90 € per inpatient case and 300.77 € per outpatient case. The rate of examinations by the German Medical Service of Health Funds (Medizinischer Dienst) was 19% over the entire observational period. The increasing case numbers in the outpatient and inpatient sectors indicate acentralization and consolidation of hand surgical expertise in the region. There are significant revenue differences between outpatient and inpatient cases.

A novel stacked ensemble classifier for Parkinson’s disease classification using voice features

A novel stacked ensemble classifier for Parkinson’s disease classification using voice features

An intelligent magnetic resonance imagining-based multistage Alzheimer's disease classification using swish-convolutional neural networks.

Alzheimer's disease (AD) refers to a neurological disorder that causes damage to brain cells and results in decreasing cognitive abilities and memory. In brain scans, this degeneration can be seen in different ways. The disease can be classified into four stages: Non-demented (ND), moderate demented (MoD), mild demented (MiD), and very mild demented (VMD). To prepare the raw dataset for analysis, the collected magnetic resonance imaging (MRI) images are subjected to several pre-processing techniques in order to improve the performance accuracy of the proposed model. Medical images generally have poor contrast and get affected by noise, which ends up with inaccurate diagnosis. For the different phases of AD to be detected, a clear image is necessary. To address this issue, the influence of the artefacts must be reduced, enhance the contrast, and reduce the loss of information. A novel framework for image enhancement is suggested to increase the accuracy in the detection and identification of AD. In this study, the raw MRI dataset from the Alzheimer's disease neuroimaging initiative (ADNI) database is subjected to skull stripping, contrast enhancement, and image filtering followed by data augmentation to balance the dataset with four types of Alzheimer's classes. The pre-processed data are subjected to five different pre-trained models such as AlexNet, ResNet, VGG 16, EfficientNet, and Inceptionv3 achieving a testing accuracy rate of 91.2%, 88.21%, 92.34%, 93.45%, and 85.12%, respectively. These pre-trained models are compared with the proposed CNN (convolutional neural network) model designed with Adam optimizer and Flatten Swish activation function which reaches the highest accuracy of 96.5% with a learning rate of 0.000001. The five pre-trained CNNmodels along with the proposed swish-based AD-CNN were tested using various performance metrics to evaluate the model efficiency in classifying and identifying the AD classes. From the result analysis, it is evident that the proposed AD-CNN model outperforms all the other models.

Accuracy and Performance Characteristics of Administrative Codes for the Diagnosis of Autoinflammatory Syndromes

Objective The aim of this study was to evaluate and validate the accuracy and performance characteristics of administrative codes in diagnosing autoinflammatory syndromes (AISs). Methods We identified potential AIS patients from the electronic medical records at the University of Iowa Hospital and Clinics and the Stead Family Children's Hospital using a screening filter based on the 10th edition of the International Classification of Diseases (ICD-10) codes and interleukin-1 antagonists. Diagnostic criteria for adult-onset Still disease, systemic juvenile idiopathic arthritis, Behçet disease (BD), familial Mediterranean fever (FMF), cryopyrin-associated periodic syndrome (CAPS), and SAPHO (synovitis, acne, pustulosis, hyperostosis, and osteitis) syndrome and chronic nonbacterial osteomyelitis (SAPHO-CNO) were reviewed for each patient. Patients who did not meet the diagnostic criteria were categorized as non-AIS. In this cross-sectional study, we calculated the sensitivity, specificity, positive predictive value, negative predictive value, and area under the receiver operating characteristic curve for the ICD codes in diagnosing AIS. Results Out of the 502 patients with potential AIS, 338 patients (67%) had a true AIS diagnosis. Sensitivity ranged from 80% (SAPHO-CNO) to 100% (BD and FMF), and positive predictive value ranged from 15% (FMF) to 80% (SAPHO-CNO). Specificity ranged from 81% (FMF) to 99% (CAPS and SAPHO-CNO), whereas negative predictive value ranged from 98% (adult-onset Still disease) to 100% (systemic juvenile idiopathic arthritis, BD, FMF, and CAPS). All ICD codes or code combinations for the diagnosis of specific AIS subtypes showed high accuracy with areas under the receiver operating characteristic curve ≥0.89. Conclusions This study validated the accuracy of administrative codes for diagnosing AIS, supporting their use in constructing AIS cohorts for clinical outcomes research.

Prenatal homelessness, food insecurity, and unemployment and adverse infant outcomes in a California cohort, 2007-2020.

Characterize the relationship between infant outcomes and prenatal homelessness, food insecurity and unemployment. California live births between 22- and 44-weeks' gestation comprised 6,089,327 pregnancies (2007-2020). Data were collected from linked Vital Statistics and hospital discharge records. Prenatal homelessness, food insecurity, and unemployment were classified as health-related social needs (HRSN) using International Classification of Disease codes in delivery records. Risk ratios for preterm birth, low birthweight, small for gestational age, neonatal intensive care unit admission, emergency department admission, rehospitalization, and death were estimated using log-linear Poisson regression adjusted for birthing person race, payer, and education. 65.7 per 100,000 births had HRSN. These infants had a higher risk of preterm birth (aRR 2.7), low birthweight (aRR 2.7), SGA (aRR 1.5), NICU admission (aRR 3.5), and death (aRR 3.0). HRSN increase the risk of infant morbidity and mortality but remain underreported in administrative records, making definitive conclusions difficult.

Evolving molecular classification of aggressive B-cell lymphoma.

This review aims to provide an overview of the latest developments in the classification and molecular understanding of aggressive B-cell lymphomas, specifically focusing on diffuse large B-cell lymphoma (DLBCL) and high-grade B-cell lymphoma (HGBL). Advances in molecular techniques have led to novel ways to classify these lymphomas based on clinical, histological, transcriptional, and genetic properties. While these methods have predominantly focused on the malignant compartment, recent studies emphasize the value of profiling the tumour microenvironment for a more comprehensive disease classification. Additionally, the integration of liquid biopsies represents a promising advancement, offering less invasive and dynamic insights into tumour characteristics and treatment response. Although molecular profiles are not yet routinely used to guide therapy, emerging data highlight their potential to predict responses to novel treatments. It is our belief that integrating molecular profiling and liquid biopsies into clinical practice and research now will pave the way for more personalized and effective therapies in the future.

Reoperations After Operatively and Non-Operatively Treated Periprosthetic Knee Fractures: A Nationwide Study on 1,931 Fractures After Primary Total Knee Arthroplasty.

Periprosthetic knee fracture (PPKF) following total knee arthroplasty (TKA) can be difficult to treat. A PPKF can be treated both operatively and non-operatively, and the treatment varies between fracture sites. This study aimed to assess the risk of reoperation according to the fracture site and treatment of the PPKF. This study is a nationwide cohort study using register data from 1997 to 2022. Cruciate-retaining or posterior-stabilized primary TKA were identified from the Danish Knee Arthroplasty Register (DKR). Subsequent PPKFs, fracture treatments, and later reoperations were identified through the International Classification of Diseases (ICD-10) diagnosis and procedure codes in the Danish National Patient Register. Additionally, indications for revision TKA (rTKA) in the DKR were used to identify PPKFs and the reason for reoperation. We included 1,931 PPKFs (1,494 femoral, 207 patellae, and 230 tibial) with an overall two-year reoperation risk of 20% (95% CI [confidence interval]: 18 to 23). Femoral PPKFs had a two-year reoperation risk of 16% (12 to 22) after non-operative treatment and 21% (18 to 24) after operative treatment. Patellar PPKFs had a two-year reoperation risk in five to 17 after non-operative treatment and 46% (30 to 69) after operative treatment. Tibial PPKFs had a two-year reoperation risk of 17% (11 to 27) after non-operative treatment, and 36% (25 to 53) after operative treatment. The overall two-year reoperation risk was 20% (18 to 23). Operative-treated PPKFs had a higher risk of reoperation across all fracture locations. The highest reoperation risk was found after operatively treated patella and tibial PPKFs (46 and 36%, respectively), and the lowest reoperation risk was found after non-operative treated patella PPKFs.

Superiority of Frailty Over Age in Predicting Outcomes Among Clostridium Difficile Patients: Evidence from National Data.

Clostridium difficile infection (CDI) has become a significant healthcare issue with increasing morbidity and mortality in the US and Europe. Frailty, characterized by reduced physical reserves and resistance to stressors, is linked to poor outcomes but its impact on CDI patients remains underexplored. This study seeks to address this gap through a nationwide analysis. Using the National Readmission Database from 2016 to 2020, we employed the International Classification of Diseases, 10th revision, Clinical Modifications codes to identify adult patients admitted with CDI. We further stratified CDI hospitalizations based on frailty. Utilizing a regression model, we assessed the impact of frailty on CDI outcomes. We included 144,611 CDI patients of whom 98,167 (67.88%) were frail. Multivariate analysis showed that frail CDI patients had significantly higher mortality (adjusted odds ratio (aOR) 4.87), acute kidney injury requiring dialysis (aOR 9.50), septic shock (aOR 14.23), and intensive care unit admission (aOR 6.80). CDI-specific complications were more likely in frail patients, including toxic megacolon (aOR 10.22), intestinal perforation (aOR 2.30), need for colectomy (aOR 3.90) and CDI recurrence (aOR 3.65). Resource utilization, indicated by hospitalization charges, length of stay, and 30-day readmission rates, was greater among frail patients. Our study underscores the significant association between frailty and various critical endpoints of CDI, including its incidence, mortality, and recurrence. Additionally, frailty independently predicts higher resource utilization and elevated 30-day readmission. Recognizing frailty as a determinant of CDI outcomes can aid clinicians in risk stratification and guide tailored interventions for this population.

Hospitalization for SARS-CoV-2 and the risk of self-harm readmission: a French nationwide retrospective cohort study.

The impact of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection on the risk of self-harming behaviours warrants further investigation. Here, we hypothesized that people with a history of hospitalization for self-harm may be particularly at risk of readmission in case of SARS-CoV-2 hospitalization. We conducted a retrospective analysis based on the French national hospitalization database. We identified all patients hospitalized for deliberate self-harm (10th edition of the International Classification of Diseases codes X60-X84) between March 2020 and March 2021. To study the effect of SARS-CoV-2 hospitalization on the risk of readmission for self-harm at 1-year of the inclusion, we performed a multivariable Fine and Gray model considering hospital death as a competing event. A total of 61,782 individuals were hospitalized for self-harm. During the 1-year follow-up, 9,403 (15.22%) were readmitted for self-harm. Between inclusion and self-harm readmission or the end of follow-up, 1,214 (1.96% of the study cohort) were hospitalized with SARS-CoV-2 (mean age 60years, 52.9% women) while 60,568 were not (mean age 45years, 57% women). Multivariate models revealed that the factors independently associated with self-harm readmission were: hospitalization with SARS-CoV-2 (adjusted hazard ratio (aHR)=3.04 [2.73-3.37]), psychiatric disorders (aHR=1.61 [1.53-1.69]), self-harm history (aHR=2.00 [1.88-2.04]), intensive care and age above 80. In hospitalized people with a personal history of self-harm, infection with SARS-CoV-2 increased the risk of readmission of self-harm, with an effect that seemed to add to the effect of a history of mental disorders, with an equally significant magnitude. Infection may be a significantly stressful condition that precipitates self-harming acts in vulnerable individuals. Clinicians should pay particular attention to the emergence of suicidal ideation in these patients in the aftermath of SARS-CoV-2 infection.

Risk factors for severe retinopathy of prematurity stratified by birth weight and gestational age in privately insured infants.

To evaluate risk factors associated with severe retinopathy of prematurity ROP (sROP) in two separate cohorts of infants. We performed a retrospective study of the Merative MarketScan Commercial Database between 2003 and 2022. Infants with ROP were stratified into two cohorts: group A (defined as infants with both birth weight [BW] <1000 g and gestational age [GA] <29 weeks) and group B (defined as infants with either BW ≥1000 g or GA ≥29 weeks). Infants with sROP were defined as those who received treatment for ROP. Outpatient and inpatient claims using International Classification of Diseases and Current Procedural Terminology codes were used to identify neonatal comorbidities. Multivariable logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for sROP requiring treatment with respect to various comorbidities. A total of 425 of 8,789 infants with ROP developed sROP (4.8%). We identified 2,726 infants in group A, of whom 387 (14.2%) required a procedure, and 6,063 in group B, of whom 38 (0.6%) required a procedure. In group A, intraventricular hemorrhage (OR = 1.38; 95% CI, 1.09-1.74) and patent ductus arteriosus ligation (OR = 1.65; 95% CI, 1.25-2.16) were the comorbidities that significantly increased the odds of sROP on multivariable analysis. In group B, infection (OR = 1.96; 95% CI, 1.01-3.80) was the only comorbidity that significantly increased the adjusted odds of sROP. Risk factors for sROP may differ between smaller and larger infants. Infection may serve as an important risk factor for ROP progression amongst larger infants due to its connection with poor postnatal growth.

First Trimester Tetracycline Exposure and Risk of Major Congenital Malformations.

Safety data on tetracycline antibiotic use during pregnancy are limited. To investigate the association between first trimester exposure to tetracyclines and the risk of major congenital malformations (MCMs). This cohort study used data from nationwide registers for singleton infants live-born in Sweden from July 1, 2006, to December 31, 2018, with follow-up through December 2019. Tetracycline-exposed and unexposed infants were matched on propensity scores (ratio 1:10). Data analysis was performed from June 2023 to May 2024. First trimester exposure to tetracyclines determined from maternal prescription fills, compared with no exposure. The primary outcome was any MCM diagnosed in the first year of life; secondary outcomes were 12 major malformation organ system subgroups and 16 individual malformations selected on the basis of prior safety signals and/or fulfillment of prespecified statistical power criteria. Malformations were identified by International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes aligned to the European Surveillance of Congenital Anomalies (EUROCAT) classification. Log binomial regression was used to estimate relative risks (RRs) with 95% CIs. From a source cohort of 1 245 889 eligible infants (640 892 male [51.4%]), a propensity score-matched analytical cohort of 69 656 infants (35 903 male [51.5%]) was generated. Of 6340 infants exposed to tetracyclines during the first trimester (3325 male [52.4%]), 252 received a diagnosis of any MCM (39.75 cases per 1000 infants; 95% CI, 35.14-44.93 cases per 1000 infants) compared with 2454 of 63 316 unexposed infants (38.76 cases per 1000 infants; 95% CI, 37.27-40.30 cases per 1000 infants). Tetracycline exposure was not associated with any MCM (RR, 1.03; 95% CI, 0.90-1.16). The RRs for specific tetracycline substances were 1.07 (95% CI, 0.93-1.23) for doxycycline, 0.83 (95% CI, 0.60-1.15) for lymecycline, and 0.78 (95% CI, 0.32-1.92) for tetracycline-oxytetracycline. There was no increased risk for 10 of 12 malformation subgroups or for any of the 16 individual malformations analyzed. The higher RRs in the main analysis for nervous system anomalies (1.92; 95% CI, 0.98-3.78) and eye anomalies (1.76; 95% CI, 1.07-2.91) were attenuated in a sensitivity analysis with follow-up extended to age 3 years (nervous system anomalies, RR, 1.08; 95% CI, 0.52-2.24; eye anomalies, RR, 1.42; 95% CI, 0.88-2.29). In this cohort study, first trimester tetracycline exposure was not associated with increased risks of any MCMs. Additional studies are needed to rule out potential risks owing to power limitations for several MCM subgroups and individual malformations.

Impact of hospital readmissions following hospitalisation with delirium on 12-month mortality: a quaternary Australian hospital experience.

Inpatient delirium and unplanned hospital readmissions are associated with increased mortality. This study aimed to determine the effect of 28-day unplanned hospital readmissions on 12-month mortality post-discharge in patients with delirium during index hospitalisation. Retrospective longitudinal cohort study of adults aged 65 or above with delirium during hospitalisation at a Victorian quaternary hospital was performed. Delirium was identified by the inclusion of ICD-10 (International Classification of Diseases, 10th revision) codes in the hospital medical discharge summary. Descriptive statistics was obtained for baseline characteristics. Cox proportional hazards model was developed to identify independent predictors of 12-month post-discharge mortality. One thousand six hundred thirty-four patients with delirium during in-patient admission were included. The overall 12-month mortality rate was 35% (572 patients). Of the 1,425 patients who survived their index admission, 11.2% had an unplanned 28-day readmission. In Cox regression analysis, unplanned readmission (hazard ratio (HR) 2.3, 95% confidence internal (CI) 1.7-2.9), older age (HR 1.38, CI 1.11-1.72), Charlson Comorbidity Index (HR 1.21, CI 1.17-1.27), and discharge to nursing home (HR 1.58, CI 1.23-2.02) were independent predictors of 12-month mortality. Readmitted patients with 12-month mortality were older, with higher rates of dementia, polypharmacy, and nursing home residence compared to readmitted patients who did not reach this endpoint. Unplanned hospital readmission within 28days of discharge is an independent predictor of 12-month mortality post in-hospital delirium admission. Admissions complicated by delirium and readmission episodes should instigate discussions regarding prognostication and goals of care. Greater research is required to minimise hospital readmission rates following discharge in this cohort.

Combined graph convolutional networks with a multi-connection pattern to identify tremor-dominant Parkinson's disease and essential tremor with resting tremor.

Essential tremor with resting tremor (rET) and tremor-dominant Parkinson's disease (tPD) share many similar clinical symptoms, leading to frequent misdiagnoses. Functional connectivity (FC) matrix analysis derived from resting-state functional MRI (Rs-fMRI) offers a promising approach for early diagnosis and for exploring FC network pathogenesis in rET and tPD. However, methods relying solely on a single connection pattern may overlook the complementary roles of different connectivity patterns, resulting in reduced diagnostic differentiation. Therefore, we propose a multi-pattern connection Graph Convolutional Network (MCGCN) method to integrate information from various connection modes, distinguishing between rET and healthy controls (HC), tPD and HC, and rET and tPD. We constructed FC matrices using three different connectivity modes for each subject and used these as inputs to the MCGCN model for disease classification. The classification performance of the model was evaluated for each connectivity mode. Subsequently, gradient-weighted class activation mapping (Grad-CAM) was used to identify the most discriminative brain regions. The important brain regions identified were primarily distributed within cerebellar-motor and non-motor cortical networks. Compared with single-pattern GCN, our proposed MCGCN model demonstrated superior classification accuracy, underscoring the advantages of integrating multiple connectivity modes. Specifically, the model achieved an average accuracy of 88.0% for distinguishing rET from HC, 88.8% for rET from tPD, and 89.6% for tPD from HC. Our findings indicate that combining graph convolutional networks with multi-connection patterns can not only effectively discriminate between tPD, rET, and HC but also enhance our understanding of the functional network mechanisms underlying rET and tPD.

Robust Multi-Class Classification for Real-Time Agricultural Applications Using Efficient and Adaptive Deep Learning

Plant diseases severely affect agricultural productivity, necessitating accurate and rapid detection methods. This research presents a robust, multi-class plant disease classification framework using adaptive deep learning. We utilize pre-trained convolutional neural networks (CNNs), specifically Xception, InceptionResNetV2, InceptionV3, ResNet50, and the proposed EfficientNetB3-based Adaptive Augmented Deep Learning (EfficientNetB3-AADL) model. Our approach leverages transfer learning combined with extensive data augmentation and trimming techniques to enhance model performance and mitigate overfitting. The EfficientNetB3-AADL architecture incorporates convolutional and max pooling layers, regularization strategies, and a dense feature learning layer, optimized to classify 52 disease categories from a publicly available leaf image dataset. The model’s performance is extensively evaluated using metrics such as accuracy, precision, recall, and F1 score. Notably, EfficientNetB3-AADL achieves superior accuracy over 98%, outperforming other CNN models. The proposed methodology highlights the efficacy of compound scaling and adaptive data augmentation in ensuring robust and efficient disease classification, suitable for real-time agricultural applications. This advancement supports sustainable farming by offering a scalable, computationally efficient solution for early and accurate disease detection in diverse crop species.

DeepEMPR: coffee leaf disease detection with deep learning and enhanced multivariance product representation

Plant diseases threaten agricultural sustainability by reducing crop yields. Rapid and accurate disease identification is crucial for effective management. Recent advancements in artificial intelligence (AI) have facilitated the development of automated systems for disease detection. This study focuses on enhancing the classification of diseases and estimating their severity in coffee leaf images. To do so, we propose a novel approach as the preprocessing step for the classification in which enhanced multivariance product representation (EMPR) is used to decompose the considered image into components, a new image is constructed using some of those components, and the contrast of the new image is enhanced by applying high-dimensional model representation (HDMR) to highlight the diseased parts of the leaves. Popular convolutional neural network (CNN) architectures, including AlexNet, VGG16, and ResNet50, are evaluated. Results show that VGG16 achieves the highest classification accuracy of approximately 96%, while all models perform well in predicting disease severity levels, with accuracies exceeding 85%. Notably, the ResNet50 model achieves accuracy levels surpassing 90%. This research contributes to the advancement of automated crop health management systems.

Novel statistically equivalent signature-based hybrid feature selection and ensemble deep learning LSTM and GRU for chronic kidney disease classification

Chronic kidney disease (CKD) involves numerous variables, but only a few significantly impact the classification task. The statistically equivalent signature (SES) method, inspired by constraint-based learning of Bayesian networks, is employed to identify essential features in CKD. Unlike conventional feature selection methods, which typically focus on a single set of features with the highest predictive potential, the SES method can identify multiple predictive feature subsets with similar performance. However, most feature selection (FS) classifiers perform suboptimally with strongly correlated data. The FS approach faces challenges in identifying crucial features and selecting the most effective classifier, particularly in high-dimensional data. This study proposes using the Least Absolute Shrinkage and Selection Operator (LASSO) in conjunction with the SES method for feature selection in CKD identification. Following this, an ensemble deep-learning model combining long short-term memory (LSTM) and gated recurrent unit (GRU) networks is proposed for CKD classification. The features selected by the hybrid feature selection method are fed into the ensemble deep-learning model. The model’s performance is evaluated using accuracy, precision, recall, and F1 score metrics. The experimental results are compared with individual classifiers, including decision tree (DT), Random Forest (RF), logistic regression (LR), and support vector machine (SVM). The findings indicate a 2% improvement in classification accuracy when using the proposed hybrid feature selection method combined with the LSTM and GRU ensemble deep-learning model. Further analysis reveals that certain features, such as HEMO, POT, bacteria, and coronary artery disease, contribute minimally to the classification task. Future research could explore additional feature selection methods, including dynamic feature selection that adapts to evolving datasets and incorporates clinical knowledge to enhance CKD classification accuracy further.

2024/00264 Harnessing Transdisciplinary Knowledge: Integrated Deep Learning Techniques for Accurate Tomato Leaf Disease Classification

The study proposes a transdisciplinary approach integrating knowledge from fields such ascomputer science, botany, and data science to classifying leaf diseases. We integrated two deep-learningmodels that combine the strengths of the Inception network and the ResNet architecture to address thechallenge of accurately classifying tomato leaf diseases. The Inception network’s ability to quickly pick upvisual features on multiple scales is used to pull out fine-grained details that are needed to tell the differencebetween small changes in the shape of tomato leaves and disease symptoms. The ResNet architecture isgood at learning deep representations and getting around the vanishing gradient problem. This lets themodel learn the high-level concepts and complicated connections between different tomato leaf diseasepatterns. The integration of these two powerful deep-learning techniques results in a robust and highlyperformant tomato leaf classification model. Extensive tests on a 10-class dataset of tomato leaves, with 9disease categories and 1 healthy class, show that the proposed model works better than others, with a testset accuracy of 98.07%. The findings of this research contribute to the advancement of automated andefficient tomato leaf disease detection systems, which can aid in the early identification and management oftomato diseases, leading to improved crop yields and quality.

Enhancing image-based diagnosis of gastrointestinal tract diseases through deep learning with EfficientNet and advanced data augmentation techniques.

The early detection and diagnosis of gastrointestinal tract diseases, such as ulcerative colitis, polyps, and esophagitis, are crucial for timely treatment. Traditional imaging techniques often rely on manual interpretation, which is subject to variability and may lack precision. Current methodologies leverage conventional deep learning models that, while effective to an extent, often suffer from overfitting and generalization issues on medical image datasets due to the intricate and subtle variations in disease manifestations. These models typically do not fully utilize the potential of transfer learning or advanced data augmentation, leading to less-than-optimal performance, especially in diverse real-world scenarios where data variability is high. This study introduces a robust model using the EfficientNetB5 architecture combined with a sophisticated data augmentation strategy. The model is tailored for the high variability and intricate details present in gastrointestinal tract disease images. By integrating transfer learning with maximal pooling and extensive regularization, the model aims to enhance diagnostic accuracy and reduce overfitting. The proposed model achieved a test accuracy of 98.89%, surpassing traditional methods by incorporating advanced regularization and augmentation techniques. The application of horizontal flipping and dynamic scaling during training significantly improved the model's ability to generalize, evidenced by a low-test loss of 0.230 and high precision metrics across all classes. The proposed deep learning framework demonstrates superior performance in the automated classification of gastrointestinal diseases from image data. By addressing key limitations of existing models through innovative techniques, this study contributes to the enhancement of diagnostic processes in medical imaging, potentially leading to more accurate and timely disease interventions.

Abstract 4120332: RISING TRENDS IN ISCHEMIC HEART DISEASE RELATED MORTALITY AMONG OLDER ADULTS WITH SLEEP APNEA IN THE UNITED STATES FROM 1999 TO 2021

Introduction: Sleep apnea (SA) is often underrecognized and undertreated despite its high prevalence in the adult population and its association with adverse cardiovascular outcomes. There are limited estimates of national trends on cardiovascular mortality in older patients with sleep apnea. We aimed to assess the sex and race-related trends of ischemic heart disease (IHD) mortality in the older adults with SA using a large population-based database. Methods: We utilized the Centers for Disease Control and Prevention Wide-Ranging, Online Data for Epidemiologic Research (CDC WONDER) database which provides information from death certificates of all US residents according to the International Classification of Diseases, Tenth Revision (ICD-10). The demographic and mortality data were obtained for the United States population &gt;65 years from 1999 to 2021. Ischemic heart disease (ICD-10 codes I20-I25) was listed as the underlying cause of death, and SA (G47.3) as a contributing cause of death. Age adjusted mortality rates (AAMRs) per 1,000,000 population were calculated by standardizing deaths to the year 2000 US population. We used Jointpoint Regression Program to analyze temporal trends in mortality from 2000 to 2021. Average annual percentage change (AAPC) with 95% CI were calculated to examine trends in AAMR over time. Results: Overall, AAMR of IHD mortality for patients with SA increased from 7.9 per 1,000,000 (95% CI, 6.9-8.8) in 1999 to 53.4 per 1,000,000 (95% CI, 51.4-55.4) in 2021 with an AAPC of 9.1% per year (95% CI, 8.8-9.5). Men had consistently higher AAMR than women throughout the study period (overall AAMR men: 45.51 (95% CI, 44.8-46.2); women: 12.5 (95% CI, 12.2-12.8). Both the groups had a similar increasing trend in AAMR, with men having a steeper increase. [AAPC men: 9.3% (95% CI, 8.5-10.8) versus AAPC women: 8.6%, 95% CI, 8.1-9.7]. Non Hispanic (NH) White population had the greatest AAMR throughout the study period, followed by NH Black and Hispanic or Latino. The NH White population had the largest increase in AAMR from 1999 to 2021 (AAPC 9.4%, 95% CI:8.9-10.1). Conclusion: In the United States, there has been a general increase in IHD mortality related to sleep apnea over the last two decades. This rising trend as noted in our analysis is concerning and underscores the need for more robust cardiovascular surveillance in these patients.

Abstract 4136289: Aortic Valve Calcium as a Predictor of Chronic Kidney Disease in a Multi-Ethnic Cohort: The MESA Study

Background: Aortic valve calcium (AVC) is associated with an increased risk of cardiovascular disease, non-cardiovascular disease such as dementia, and all-cause mortality. Traditional atherosclerotic cardiovascular disease risk factors are associated with both AVC and chronic kidney disease (CKD), but whether there is an association between AVC and CKD is unknown. Objectives: To ascertain whether AVC quantified by cardiac CT scanning is independently associated with the long-term risk of incident CKD among individuals without a previous history of cardiovascular disease. Methods: We examined 6,346 Multi-Ethnic Study of Atherosclerosis (MESA) participants who underwent cardiac CT scanning at Visit 1 (2000-02) and had an eGFR of ≥ 60 mL/min/1.73 m 2 . AVC was quantified using the Agatston method and categorized as 0, 1-99, and ≥100. Incident CKD was defined as an eGFR &lt; 60 mL/min/1.73 m 2 accompanied with an at least 40% decline in eGFR from baseline, and/or a diagnosis of CKD and indicators of end stage renal disease extracted from hospital records using the International Classification of Disease (ICD) codes. We performed Kaplan-Meier survival curve analyses along with multivariable Cox proportional hazard regression models, adjusted for age, gender, race/ethnicity, highest level of education and traditional cardiovascular risk factors along with coronary artery calcium (CAC), lipoprotein (a) (Lp[a]), and the APOE-ε4 genotype to examine the association between AVC (categorical and log-transformed) and incident CKD. Results: Participants had a mean age 62.2±10.1 years, 53% were women, and AVC &gt;0 was present in 795 (12%) participants. During a median follow-up time of 16.9 years, 982 (15%) participants developed incident CKD. AVC examined as a continuous variable was associated with a significantly increased risk of developing CKD (per log-unit [AVC+1] HR 1.06 [95% CI: 1.02-1.10]; p = 0.005). There was a stepwise increased risk for CKD with higher AVC levels ( Figure ). Similarly, in the multivariable adjusted Cox models, participants with AVC ≥100 had a higher risk of incident CKD, compared with the AVC=0 group (HR 1.48 [95% CI: 1.15-1.89]; p = 0.002). The observed associations remained after further adjusting for CAC score ( p = 0.024), Lp (a) ( p = 0.004), and the APOE-ε4 genotype ( p = 0.004). Conclusions: In a multi-ethnic cohort of participants free of CKD at baseline, AVC was independently associated with a higher risk of incident CKD.