Hasil untuk "Nutritional diseases. Deficiency diseases"

Yajun Qiao, Huimin Zheng, Ruiying Cheng et al.

Abstract Osteoporosis is the most common bone metabolic disease, but the altitude environment increases the incidence of osteoporosis. Gut microbiota is a key potential target for osteoporosis. However, it is not clear how plateau environment (hypoxia/hypothermia) interferes with the development of osteoporosis by affecting gut microbiota. Therefore, the aim of this paper is to explain that hypoxia and hypothermia environment is involved in bone metabolism regulation by affecting gut microbiota, which may be one of the pathways for the early development of osteoporosis. This paper reviews a large number of clinical and basic studies to systematically evaluate the pathway by which gut microbiota is involved in regulating bone metabolism, and further discuss the potential effects of hypoxia/hypothermia on gut microbiota in regulating bone metabolism. This review summarizes that gut microbiota was mainly involved in the regulation of bone metabolism through immune, hormone and metabolite levels, while hypothermia/hypoxia affected bone metabolism mainly through the effects of microbial immune response and short-chain fatty acids (SCFAs) secretion. In addition, our interpretation of Tibetan dietary patterns reveals a new potential complementary therapy for osteoporosis intervention. Although the initial results are exciting, more trials are needed to understand the interactions between diet, gut microbiota, and bone metabolism.

Emilio J. Medrano-Sanchez, Ciel A. Gutierrez-Berrocal, Luciana C. Gonzales-Aguilar et al.

This literature review examined the relationship between energy drink consumption and cardiovascular health in young people. Following PRISMA 2020, we searched Scopus for articles published from 2020 to 2025 and included 33 original studies after screening 133 records. Evidence from observational, clinical, and experimental research was synthesized into six themes: youth consumption; direct cardiovascular outcomes; composition and toxicity; animal or cellular experiments; perceptions and habits; and occupational or sociodemographic factors. Across studies, habitual intake was linked to acute blood-pressure rises, arrhythmias, endothelial dysfunction, and metabolic disturbances, sometimes within 24 h of a single can. Risks were amplified by high caffeine and taurine doses and by co-use with alcohol or intense exercise. Adolescents and young adults were most vulnerable, due to heightened sympathetic responses, frequent use under academic or work stress, and limited risk perception. Authors highlighted five actions: longitudinal research; tighter ingredient monitoring and transparent labeling; consumer education; protection of vulnerable groups; and clinical guidance for responsible use. These results were observed across regions and study designs. Overall, the findings indicate that unregulated energy-drink consumption is a preventable cardiovascular risk in youth, justifying the use of coordinated public-health measures, including curriculum-based education, marketing restrictions, ingredient oversight, and clinical screening to mitigate harm.

Sofia I Olmedo, Claudia R Valeggia, Cecilia Palavecino et al.

The lifestyles and worldviews of indigenous communities have long been deeply intertwined with natural resources, particularly water. These vital resources are now severely threatened by systemic social marginalization and the enduring impacts of colonization, further violating the human right to water access. Our primary objective was to assess the domains and correlates of water insecurity in a Pilagá community in Formosa, Argentina. This sequential exploratory mixed-methods cross-sectional study, conducted in 2023, involved data collection from Pilagá households representing 59 family clusters, covering all family units in the community. We used a prevalidated Household Water Insecurity Experience survey. Qualitative data were gathered through semistructured interviews and participant observation. The average age of participants was 36.8 ± 12.7 y, with most being women, who primarily handled the task of fetching water. Water insecurity was prevalent, affecting 62% of households, most of which depended on well pumps. The most serious concern associated with water was the lack of long-term stability. Through an ecologic model, we identified multiple interrelated contextual variables, revealing that shifts in one area (geographic, capitalistic market, water policies, and infrastructure policies) had ripple effects across others. Key correlates included water sources, cultural perceptions of water, resource distribution, and social dynamics around water. The Pilagá community confronts pervasive water insecurity within a challenging and evolving socioecologic landscape.

Mashudu Nemukula, Siphesihle Mkhwanazi, Tumelo Jessica Mapheto et al.

Background/Objectives: Type 2 diabetes mellitus (T2DM) is a major global public health challenge with a significant impact on human life. The current study aims to provide a comprehensive analysis of the magnitude of dyslipidemia and the factors associated with elevated LDL-C levels among Black South Africans with T2DM. Methods: This was a cross-sectional study conducted in a tertiary hospital. Blood samples for glycated hemoglobin (HbA1c) and lipid profile were collected from the study participants and analyzed using Siemens Atellica™ analyzer. The data was entered into Microsoft excel and analyzed using SPSS version 24. Bivariate and multivariate logistic regression was employed to identify variables significantly associated with the outcomes, with a <i>p</i>-value  ≤  0.05 and a 95% confidence interval. Results: A total of 194 study participants with T2DM were recruited in the study. The overall prevalence of dyslipidemia was 90.72%. Of those with dyslipidemia, 40.9% had an isolated dyslipidemia, 39.7% had a combined dyslipidemia and 19.3% had atherogenic dyslipidemia. Significant factors associated with elevated levels of LDL-C included age, non-adherence to treatment (NAT) and duration. However, after multivariate analysis, NAT was found to be an independent associated factor with elevated levels of LDL-C (AOR: 4.596; 95% CI: 0.177–2.874; <i>p =</i> 0.027). Conclusions: Our study found that dyslipidemia is highly prevalent among Black South African patients with T2DM at a tertiary hospital, despite the use of lipid-lowering therapy. NAT was significantly associated with elevated levels of LDL-C. However, it is important to note that the study employed a cross-sectional design, conducted at a single hospital, which may impair the generalizability of the findings.

Paola F. Antonietti, Mattia Corti, Sergio Gómez et al.

This work presents a structure-preserving, high-order, unconditionally stable numerical method for approximating the solution to the Fisher-Kolmogorov equation on polytopic meshes, with a particular focus on its application in simulating misfolded protein spreading in neurodegenerative diseases. The model problem is reformulated using an entropy variable to guarantee solution positivity, boundedness, and satisfaction of a discrete entropy-stability inequality at the numerical level. The scheme combines a local discontinuous Galerkin method on polytopal meshes for the space discretization with a $ν$-step backward differentiation formula for the time integration. Implementation details are discussed, including a detailed derivation of the linear systems arising from Newton's iteration. The accuracy and robustness of the proposed method are demonstrated through extensive numerical tests. Finally, the method's practical performance is demonstrated through simulations of $α$-synuclein propagation in a two-dimensional brain geometry segmented from MRI data, providing a relevant computational framework for modeling synucleopathies (such as Parkinson's disease) and, more generally, neurodegenerative diseases.

Minh Le, Khoi Ton

Chronic diseases are long-term, manageable, yet typically incurable conditions, highlighting the need for effective preventive strategies. Machine learning has been widely used to assess individual risk for chronic diseases. However, many models rely on medical test data (e.g. blood results, glucose levels), which limits their utility for proactive self-assessment. Additionally, to gain public trust, machine learning models should be explainable and transparent. Although some research on self-assessment machine learning models includes explainability, their explanations are not validated against established medical literature, reducing confidence in their reliability. To address these issues, we develop deep learning models that predict the risk of developing 13 chronic diseases using only personal and lifestyle factors, enabling accessible, self-directed preventive care. Importantly, we use SHAP-based explainability to identify the most influential model features and validate them against established medical literature. Our results show a strong alignment between the models' most influential features and established medical literature, reinforcing the models' trustworthiness. Critically, we find that this observation holds across 13 distinct diseases, indicating that this machine learning approach can be broadly trusted for chronic disease prediction. This work lays the foundation for developing trustworthy machine learning tools for self-directed preventive care. Future research can explore other approaches for models' trustworthiness and discuss how the models can be used ethically and responsibly.

You Zhou, Yingli Xie, Laijing Du et al.

Abstract Background The metabolic score for insulin resistance (METS-IR) has been validated as a novel, simple, and reliable surrogate marker for insulin resistance; however, its utility for evaluating the prognosis of heart failure with preserved ejection fraction (HFpEF) remains to be elucidated. Therefore, we aimed to analyze the association between METS-IR and the long-term prognosis of HFpEF. Methods We enrolled a total of 4,702 participants with HFpEF in this study. The participants were divided into three groups according to METS-IR tertiles: (Ln [2 × fasting plasma glucose + fasting triglycerides] × body mass index) / (Ln [high-density lipoprotein cholesterol]). The occurrence of primary endpoints, including all-cause mortality and cardiovascular (CV) death, was documented. Results There were 3,248 participants with HFpEF (mean age, 65.7 ± 13.8 years; male, 59.0%) in total who were included in the final analysis. The incidence of primary outcomes from the lowest to the highest METS-IR tertiles were 46.92, 86.01, and 124.04 per 1000 person-years for all-cause death and 26.75, 49.01, and 64.62 per 1000 person-years for CV death. The multivariate Cox hazards regression analysis revealed hazard ratios for all-cause and CV deaths of 2.48 (95% CI 2.10–2.93; P < 0.001) and 2.29 (95% CI 1.83–2.87; P < 0.001) when the highest and lowest METS-IR tertiles were compared, respectively. In addition, the predictive efficacy of METS-IR remained significant across various comorbidity subgroups (all P < 0.05). Further, adding the METS-IR to the baseline risk model for all-cause death improved the C-statistic value (0.690 for the baseline model vs. 0.729 for the baseline model + METS-IR, P < 0.01), the integrated discrimination improvement value (0.061, P < 0.01), the net reclassification improvement value (0.491, P < 0.01), and the clinical net benefit. Conclusions An elevated METS-IR, which is associated with an increased mortality risk, is a potential valuable prognostic marker for individuals with HFpEF.

Juliette Ortholand, Nicolas Gensollen, Stanley Durrleman et al.

Introduction: Heterogeneity of the progression of neurodegenerative diseases is one of the main challenges faced in developing effective therapies. With the increasing number of large clinical databases, disease progression models have led to a better understanding of this heterogeneity. Nevertheless, these diseases may have no clear onset and biological underlying processes may start before the first symptoms. Such an ill-defined disease reference time is an issue for current joint models, which have proven their effectiveness by combining longitudinal and survival data. Objective In this work, we propose a joint non-linear mixed effect model with a latent disease age, to overcome this need for a precise reference time. Method: To do so, we utilized an existing longitudinal model with a latent disease age as a longitudinal sub-model and associated it with a survival sub-model that estimates a Weibull distribution from the latent disease age. We then validated our model on different simulated scenarios. Finally, we benchmarked our model with a state-of-the-art joint model and reference survival and longitudinal models on simulated and real data in the context of Amyotrophic Lateral Sclerosis (ALS). Results: On real data, our model got significantly better results than the state-of-the-art joint model for absolute bias (4.21(4.41) versus 4.24(4.14)(p-value=1.4e-17)), and mean cumulative AUC for right censored events (0.67(0.07) versus 0.61(0.09)(p-value=1.7e-03)). Conclusion: We showed that our approach is better suited than the state-of-the-art in the context where the reference time is not reliable. This work opens up the perspective to design predictive and personalized therapeutic strategies.

Scott Pezanowski, Etien Luc Koua, Joseph C Okeibunor et al.

Objectives: Our research adopts computational techniques to analyze disease outbreaks weekly over a large geographic area while maintaining local-level analysis by incorporating relevant high-spatial resolution cultural and environmental datasets. The abundance of data about disease outbreaks gives scientists an excellent opportunity to uncover patterns in disease spread and make future predictions. However, data over a sizeable geographic area quickly outpace human cognition. Our study area covers a significant portion of the African continent (about 17,885,000 km2). The data size makes computational analysis vital to assist human decision-makers. Methods: We first applied global and local spatial autocorrelation for malaria, cholera, meningitis, and yellow fever case counts. We then used machine learning to predict the weekly presence of these diseases in the second-level administrative district. Lastly, we used machine learning feature importance methods on the variables that affect spread. Results: Our spatial autocorrelation results show that geographic nearness is critical but varies in effect and space. Moreover, we identified many interesting hot and cold spots and spatial outliers. The machine learning model infers a binary class of cases or none with the best F1 score of 0.96 for malaria. Machine learning feature importance uncovered critical cultural and environmental factors affecting outbreaks and variations between diseases. Conclusions: Our study shows that data analytics and machine learning are vital to understanding and monitoring disease outbreaks locally across vast areas. The speed at which these methods produce insights can be critical during epidemics and emergencies.

Melanie S. Hulshoff, Isabel N. Schellinger, Xingbo Xu et al.

Abstract Background The prevalence of diabetes mellitus has risen considerably and currently affects more than 422 million people worldwide. Cardiovascular diseases including myocardial infarction and heart failure represent the major cause of death in type 2 diabetes (T2D). Diabetes patients exhibit accelerated aortic stiffening which is an independent predictor of cardiovascular disease and mortality. We recently showed that aortic stiffness precedes hypertension in a mouse model of diabetes (db/db mice), making aortic stiffness an early contributor to cardiovascular disease development. Elucidating how aortic stiffening develops is a pressing need in order to halt the pathophysiological process at an early time point. Methods To assess EndMT occurrence, we performed co-immunofluorescence staining of an endothelial marker (CD31) with mesenchymal markers (α-SMA/S100A4) in aortic sections from db/db mice. Moreover, we performed qRT-PCR to analyze mRNA expression of EndMT transcription factors in aortic sections of db/db mice and diabetic patients. To identify the underlying mechanism by which EndMT contributes to aortic stiffening, we used aortas from db/db mice and diabetic patients in combination with high glucose-treated human umbilical vein endothelial cells (HUVECs) as an in vitro model of diabetes-associated EndMT. Results We demonstrate robust CD31/α-SMA and CD31/S100A4 co-localization in aortic sections of db/db mice which was almost absent in control mice. Moreover, we demonstrate a significant upregulation of EndMT transcription factors in aortic sections of db/db mice and diabetic patients. As underlying regulator, we identified miR-132-3p as the most significantly downregulated miR in the micronome of db/db mice and high glucose-treated HUVECs. Indeed, miR-132-3p was also significantly downregulated in aortic tissue from diabetic patients. We identified Kruppel-like factor 7 (KLF7) as a target of miR-132-3p and show a significant upregulation of KLF7 in aortic sections of db/db mice and diabetic patients as well as in high glucose-treated HUVECs. We further demonstrate that miR-132-3p overexpression and KLF7 downregulation ameliorates EndMT in high glucose-treated HUVECs. Conclusions We demonstrate for the first time that EndMT contributes to aortic stiffening in T2D. We identified miR-132-3p and KLF7 as novel EndMT regulators in this context. Altogether, this gives us new insights in the development of aortic stiffening in T2D.

Qianxia Jiang, Bethany Forseth, Lauren Fitzpatrick et al.

Abstract Background Most neighborhood food and activity related environment research in children has been cross-sectional. A better understanding of prospective associations between these neighborhood environment factors and children’s weight status can provide stronger evidence for informing interventions and policy. This study examined associations of baseline and changes in neighborhood healthy food access and walkability with changes in children’s weight status over 5 years. Methods Height, weight, and home address were obtained for 4,493 children (> 75% were Black or Latinx) from primary care visits within a large pediatric health system. Eligible participants were those who had measures collected during two time periods (2012–2014 [Time 1] and 2017–2019 [Time 2]). Data were integrated with census tract-level healthy food access and walkability data. Children who moved residences between the time periods were considered ‘movers’ (N = 1052; 23.4%). Mixed-effects models, accounting for nesting of children within census tracts, were conducted to model associations of baseline and changes in the neighborhood environment variables with Time 2 weight status (BMIz and overweight or obese vs. healthy weight). Models adjusted for weight status and child and neighborhood sociodemographics at baseline. Results Children living in a neighborhood with [ample] healthy food access at Time 1 had a lower BMIz at Time 2, regardless of mover status. A decrease in healthy food access was not significantly associated with children’s weight status at Time 2. Baseline walkability and improvements in walkability were associated with a lower BMIz at Time 2, regardless of mover status. Conclusions Findings provide evidence that residing in a neighborhood with healthy food access and walkability may support a healthy weight trajectory in children. Findings on changes in the neighborhood environment suggested that improved walkability in the neighborhood may support children’s healthy weight. The greater and more consistent findings among movers may be due to movers experiencing greater changes in neighborhood features than the changes that typically occur within a neighborhood over a short period of time. Future research is needed to investigate more robust environmental changes to neighborhoods.

Pengcheng Xu, Jinpu Cai, Yulin Gao et al.

DNA methylation is a crucial regulator of gene transcription and has been linked to various diseases, including autoimmune diseases and cancers. However, diagnostics based on DNA methylation face challenges due to large feature sets and small sample sizes, resulting in overfitting and suboptimal performance. To address these issues, we propose MIRACLE, a novel interpretable neural network that leverages autoencoder-based multi-task learning to integrate multiple datasets and jointly identify common patterns in DNA methylation. MIRACLE's architecture reflects the relationships between methylation sites, genes, and pathways, ensuring biological interpretability and meaningfulness. The network comprises an encoder and a decoder, with a bottleneck layer representing pathway information as the basic unit of heredity. Customized defined MaskedLinear Layer is constrained by site-gene-pathway graph adjacency matrix information, which provides explainability and expresses the site-gene-pathway hierarchical structure explicitly. And from the embedding, there are different multi-task classifiers to predict diseases. Tested on six datasets, including rheumatoid arthritis, systemic lupus erythematosus, multiple sclerosis, inflammatory bowel disease, psoriasis, and type 1 diabetes, MIRACLE demonstrates robust performance in identifying common functions of DNA methylation across different phenotypes, with higher accuracy in prediction dieseases than baseline methods. By incorporating biological prior knowledge, MIRACLE offers a meaningful and interpretable framework for DNA methylation data analysis in the context of autoimmune diseases.

Dibya Nath, G. M. Shahariar

Gastrointestinal diseases pose significant healthcare chall-enges as they manifest in diverse ways and can lead to potential complications. Ensuring precise and timely classification of these diseases is pivotal in guiding treatment choices and enhancing patient outcomes. This paper introduces a novel approach on classifying gastrointestinal diseases by leveraging cost-sensitive pre-trained deep convolutional neural network (CNN) architectures with supervised contrastive learning. Our approach enables the network to learn representations that capture vital disease-related features, while also considering the relationships of similarity between samples. To tackle the challenges posed by imbalanced datasets and the cost-sensitive nature of misclassification errors in healthcare, we incorporate cost-sensitive learning. By assigning distinct costs to misclassifications based on the disease class, we prioritize accurate classification of critical conditions. Furthermore, we enhance the interpretability of our model by integrating gradient-based techniques from explainable artificial intelligence (AI). This inclusion provides valuable insights into the decision-making process of the network, aiding in understanding the features that contribute to disease classification. To assess the effectiveness of our proposed approach, we perform extensive experiments on a comprehensive gastrointestinal disease dataset, such as the Hyper-Kvasir dataset. Through thorough comparisons with existing works, we demonstrate the strong classification accuracy, robustness and interpretability of our model. We have made the implementation of our proposed approach publicly available at https://github.com/dibya404/Gastrointestinal-Disease-Classification-through-Explainable-and-Cost-Sensitive-DNN-with-SCL

Weijie Sun, Sunil Vasu Kalmady, Amir Salimi et al.

Electrocardiogram (ECG) abnormalities are linked to cardiovascular diseases, but may also occur in other non-cardiovascular conditions such as mental, neurological, metabolic and infectious conditions. However, most of the recent success of deep learning (DL) based diagnostic predictions in selected patient cohorts have been limited to a small set of cardiac diseases. In this study, we use a population-based dataset of >250,000 patients with >1000 medical conditions and >2 million ECGs to identify a wide range of diseases that could be accurately diagnosed from the patient's first in-hospital ECG. Our DL models uncovered 128 diseases and 68 disease categories with strong discriminative performance.

Kantip Kiratiratanapruk, Pitchayagan Temniranrat, Wasin Sinthupinyo et al.

Fast, accurate and affordable rice disease detection method is required to assist rice farmers tackling equipment and expertise shortages problems. In this paper, we focused on the solution using computer vision technique to detect rice diseases from rice field photograph images. Dealing with images took in real-usage situation by general farmers is quite challenging due to various environmental factors, and rice leaf object size variation is one major factor caused performance gradation. To solve this problem, we presented a technique combining a CNN object detection with image tiling technique, based on automatically estimated width size of rice leaves in the images as a size reference for dividing the original input image. A model to estimate leaf width was created by small size CNN such as 18 layer ResNet architecture model. A new divided tiled sub-image set with uniformly sized object was generated and used as input for training a rice disease prediction model. Our technique was evaluated on 4,960 images of eight different types of rice leaf diseases, including blast, blight, brown spot, narrow brown spot, orange, red stripe, rice grassy stunt virus, and streak disease. The mean absolute percentage error (MAPE) for leaf width prediction task evaluated on all eight classes was 11.18% in the experiment, indicating that the leaf width prediction model performed well. The mean average precision (mAP) of the prediction performance on YOLOv4 architecture was enhanced from 87.56% to 91.14% when trained and tested with the tiled dataset. According to our study, the proposed image tiling technique improved rice disease detection efficiency.

Guan Wang, Yusuke Kikuchi, Jinglin Yi et al.

Retinal vascular diseases affect the well-being of human body and sometimes provide vital signs of otherwise undetected bodily damage. Recently, deep learning techniques have been successfully applied for detection of diabetic retinopathy (DR). The main obstacle of applying deep learning techniques to detect most other retinal vascular diseases is the limited amount of data available. In this paper, we propose a transfer learning technique that aims to utilize the feature similarities for detecting retinal vascular diseases. We choose the well-studied DR detection as a source task and identify the early detection of retinopathy of prematurity (ROP) as the target task. Our experimental results demonstrate that our DR-pretrained approach dominates in all metrics the conventional ImageNet-pretrained transfer learning approach, currently adopted in medical image analysis. Moreover, our approach is more robust with respect to the stochasticity in the training process and with respect to reduced training samples. This study suggests the potential of our proposed transfer learning approach for a broad range of retinal vascular diseases or pathologies, where data is limited.

Poornima Singh Thakur, Pritee Khanna, Tanuja Sheorey et al.

Plant diseases are the primary cause of crop losses globally, with an impact on the world economy. To deal with these issues, smart agriculture solutions are evolving that combine the Internet of Things and machine learning for early disease detection and control. Many such systems use vision-based machine learning methods for real-time disease detection and diagnosis. With the advancement in deep learning techniques, new methods have emerged that employ convolutional neural networks for plant disease detection and identification. Another trend in vision-based deep learning is the use of vision transformers, which have proved to be powerful models for classification and other problems. However, vision transformers have rarely been investigated for plant pathology applications. In this study, a Vision Transformer enabled Convolutional Neural Network model called "PlantXViT" is proposed for plant disease identification. The proposed model combines the capabilities of traditional convolutional neural networks with the Vision Transformers to efficiently identify a large number of plant diseases for several crops. The proposed model has a lightweight structure with only 0.8 million trainable parameters, which makes it suitable for IoT-based smart agriculture services. The performance of PlantXViT is evaluated on five publicly available datasets. The proposed PlantXViT network performs better than five state-of-the-art methods on all five datasets. The average accuracy for recognising plant diseases is shown to exceed 93.55%, 92.59%, and 98.33% on Apple, Maize, and Rice datasets, respectively, even under challenging background conditions. The efficiency in terms of explainability of the proposed model is evaluated using gradient-weighted class activation maps and Local Interpretable Model Agnostic Explanation.

Md Mozaharul Mottalib, Jessica C Jones-Smith, Bethany Sheridan et al.

Obesity is a major health problem, increasing the risk of various major chronic diseases, such as diabetes, cancer, and stroke. While the role of obesity identified by cross-sectional BMI recordings has been heavily studied, the role of BMI trajectories is much less explored. In this study, we use a machine-learning approach to subtype individuals' risk of developing 18 major chronic diseases by using their BMI trajectories extracted from a large and geographically diverse EHR dataset capturing the health status of around two million individuals for a period of six years. We define nine new interpretable and evidence-based variables based on the BMI trajectories to cluster the patients into subgroups using the k-means clustering method. We thoroughly review each cluster's characteristics in terms of demographic, socioeconomic, and physiological measurement variables to specify the distinct properties of the patients in the clusters. In our experiments, the direct relationship of obesity with diabetes, hypertension, Alzheimer's, and dementia has been re-established and distinct clusters with specific characteristics for several of the chronic diseases have been found to be conforming or complementary to the existing body of knowledge.

Lie Ju, Zhen Yu, Lin Wang et al.

In the real world, medical datasets often exhibit a long-tailed data distribution (i.e., a few classes occupy the majority of the data, while most classes have only a limited number of samples), which results in a challenging long-tailed learning scenario. Some recently published datasets in ophthalmology AI consist of more than 40 kinds of retinal diseases with complex abnormalities and variable morbidity. Nevertheless, more than 30 conditions are rarely seen in global patient cohorts. From a modeling perspective, most deep learning models trained on these datasets may lack the ability to generalize to rare diseases where only a few available samples are presented for training. In addition, there may be more than one disease for the presence of the retina, resulting in a challenging label co-occurrence scenario, also known as \textit{multi-label}, which can cause problems when some re-sampling strategies are applied during training. To address the above two major challenges, this paper presents a novel method that enables the deep neural network to learn from a long-tailed fundus database for various retinal disease recognition. Firstly, we exploit the prior knowledge in ophthalmology to improve the feature representation using a hierarchy-aware pre-training. Secondly, we adopt an instance-wise class-balanced sampling strategy to address the label co-occurrence issue under the long-tailed medical dataset scenario. Thirdly, we introduce a novel hybrid knowledge distillation to train a less biased representation and classifier. We conducted extensive experiments on four databases, including two public datasets and two in-house databases with more than one million fundus images. The experimental results demonstrate the superiority of our proposed methods with recognition accuracy outperforming the state-of-the-art competitors, especially for these rare diseases.

Ivan K. Chinn