Diagnosis Of Diabetes Research Articles

Enhanced multi‐key privacy‐preserving distributed deep learning protocol with application to diabetic retinopathy diagnosis

SummaryIn this work, privacy‐preserving distributed deep learning (PPDDL) is re‐visited with a specific application to diagnosing long‐term illness like diabetic retinopathy. In order to protect the privacy of participants datasets, a multi‐key PPDDL solution is proposed which is robust against collusion attacks and is also post‐quantum robust. Additionally, the PPDDL solution provides robust network security in terms of integrity of transmitted ciphertexts and keys, forward secrecy, and prevention of man‐in‐the‐middle attacks and is extensively verified using Verifpal. Proposed solution is evaluated on retina image datasets to detect diabetic retinopathy, with deep learning accuracy results of 96.30%, 96.21% and 96.20% for DDL, DDL + SINGLE and DDL + MULTI scenarios respectively. Results from our simulation indicate that accuracy of the PPDDL is maintained while protecting the privacy of the datasets of participants. Our proposed solution is also efficient in terms of the communication and run‐time costs.

Validation of GenProb-T1D and its clinical utility for differentiating types of diabetes in a biobank from a US healthcare system.

Atypical diabetes with overlapping clinical features of type 1 (T1D) and type 2 (T2D) is common and challenging diagnostically and for implementing effective treatment. Here, we validate a recently reported genetic probability of type 1 diabetes (GenProb-T1D) from the UK Biobank (UKB) for differentiating type 1 diabetes and type 2 diabetes in a diabetes patient cohort from a healthcare system-based biobank in the USA. Among 3,363 diabetes patients, we confirmed the performance of GenProb-T1D in differentiating typical type 1 diabetes vs type 2 diabetes. Furthermore, for 359 atypical diabetes patients, those with GenProb-T1D higher than the pre-defined cutoff derived from the UKB had clinical presentations more consistent with that of typical type 1 diabetes. Similar findings were found in participants of European and non-European ancestries. This study provides necessary validation to translate GenProb-T1D into genetic testing in a multi-ancestry cohort. Measuring underlying genetic susceptibility of type 1 diabetes and type 2 diabetes can supplement current clinical tools for earlier and more accurate diagnoses of diabetes.

Heart Failure Risk Prediction in a Population With a High Burden of Diabetes: Evidence From the Strong Heart Study.

Despite the high burden of diabetes and cardiovascular risk factors in American Indian communities in the United States, prospective studies of heart failure (HF) in this population group are scarce, and the generalizability of previous HF risk scales may be limited. We developed a parsimonious HF risk prediction equation that accounts for relevant risk factors affecting American Indian communities, focusing on diabetes and kidney damage. A total of 3059 participants from the SHS (Strong Heart Study) (56±8 years of age, 58% women) were included. Five hundred seven developed HF. Progressively adjusted Cox proportional hazards models were used to identify risk factors for HF and HF subtypes. Predictors of risk at 5 and 10 years included older age (hazard ratio [HR], 1.79 [95% CI, 1.43-2.25]; HR, 1.68 [95% CI, 1.44-1.95]), smoking (HR, 2.26 [95% CI, 1.23-4.13]; HR, 2.08 [95% CI, 1.41-3.06]), macroalbuminuria (HR, 8.38 [95% CI, 4.44-15.83]; HR, 5.20 [95% CI, 3.42-7.9]), microalbuminuria (HR, 2.72 [95% CI, 1.51-4.90]; HR, 1.92 [95% CI, 1.33, 2.78]), and previous myocardial infarction (HR, 6.58 [95% CI, 2.54-17.03]; HR, 3.87 [95% CI, 2.29-6.54]), respectively. These predictors, together with diabetes diagnosis and glycated hemoglobin were significant at 10 and 28 years. High discrimination performance was achieved (C index, 0.81 [95% CI, 0.76-0.84]; C index, 0.78 [95% CI, 0.75-0.81]; and C index, 0.77 [95% CI, 0.74-0.78] at 5, 10, and up to 28 years of follow up, respectively). Some associations varied across HF subtypes, although diabetes, albuminuria, and previous myocardial infarction were associated with all subtypes. This prospective study of HF risk factors in American Indian communities identifies that smoking, body mass index, and indicators of diabetes control and kidney damage (glycated hemoglobin and albuminuria) are major determinants of HF. Our findings can improve HF risk assessment in populations with a high burden of diabetes.

Aboriginal young people's experiences of type 2 diabetes diagnosis, management and support: A qualitative study in the Kimberley region of Western Australia.

In Australia, Aboriginal and Torres Strait Islander young people in remote settings are most-affected by young onset type 2 diabetes (T2D). It is necessary to understand young people's experiences, including factors impacting on self-management, to improve models of care. A phenomenological methodology underpinned this qualitative study in Western Australia's Kimberley region. Two Aboriginal Community Controlled Health Services supported recruitment of seven Aboriginal young people aged 12-24 with T2D, who participated in interviews. A carer and health professional of one young person in each site were also interviewed and relevant medical record data reviewed to assist with triangulation of data. De-identified transcripts were inductively coded and a coding structure developed with oversight by a Kimberley Aboriginal researcher. Young people reported varied experiences and emotions relating to a T2D diagnosis. Most recounted this was upsetting and some reported current negative impact on emotional wellbeing. Challenges with understanding and managing diabetes were highlighted, particularly regarding healthy eating, physical activity and medication. Family are a prominent source of self-management support, with the intergenerational impact of diabetes being evident for each participant. Positive relationships with health professionals, entailing continuity of care, were valued. There are significant emotional and medical challenges for young people with T2D and their families. Recommendations from this work will contribute to the development of local resources and initiatives to improve diabetes-related support. SO WHAT?: Alongside broader efforts to support good health at the societal level, enhanced health education and family-oriented support structures including Aboriginal clinical staff for young people with T2D are needed.

The Impact of Hearing Loss on Diabetes Distress Among Adults With Type 2 Diabetes.

The purpose of this study was to assess the combined effects of type 2 diabetes (T2D) and hearing loss on the level of distress in individuals with diabetes. The 2021 National Health Interview Survey included 2633 adults (ages 18+) with T2D who reported perceived hearing loss, level of diabetes-related distress, household composition, and demographic characteristics. Logistic regressions evaluated these association between hearing loss and diabetes distress controlling for age, income, region of residence, marital status, rurality, educational attainment, insurance coverage, time since diabetes diagnosis, and household composition. Interaction effects examined differential associations between demographic groups. About 3.5% of adults with T2D in the sample reported hearing loss, and 70% reported diabetes-related distress. Results indicated that diabetes distress was significantly more likely among individuals with hearing loss (OR = 2.08) relative to their hearing counterparts. Additionally, females (OR = 1.50), low-income earners (OR = 1.84), middle-income earners (OR = 1.41), non-Hispanic Blacks (OR = 1.58), and Hispanics (OR = 1.75) with diabetes had higher odds of reporting diabetes distress. Interaction analyses indicated that non-Hispanic Blacks and Hispanics with hearing loss had a 37% and 19%, respectively, higher likelihood of diabetes distress relative to non-Hispanic Whites with hearing loss. Diabetes distress was more likely among individuals with hearing loss. These findings suggest the communication challenges of individuals with hearing loss may explain the observed differences in distress.

Characterization of Ptau181 Among a Diverse Community-Based Cohort: A HABS-HD Study.

Examination of Alzheimer's disease (AD) related biomarkers among diverse communities has remained limited. The aim of this study was to expand on prior work to provide a characterization of ptau181 among a diverse community sample. Consideration was taken regarding the impact of comorbidities on ptau181 levels including medical. 3,228 (n = 770 African American [AA], n = 1,231 Hispanic, and n = 1,227 non-Hispanic white [NHW]) Health and Aging Brain Study- Health Disparities (HABS-HD) participants were included in this study. ANCOVAs were conducted to examine differences in ptau181 levels across race and ethnic groups. Violin plots were also generated stratified by APOEɛ4 carrier status, Amyloid PET positivity status, medical comorbidity (hypertension, dyslipidemia, chronic kidney disease [CKD], and diabetes) and by cognitive diagnosis. Ptau181 levels were found to differ between Hispanics and NHW after covarying for age, sex, and APOEɛ4 status. Amyloid PET positivity was associated with higher ptau181 levels across all groups. APOEɛ4 positivity status was only significantly associated with ptau181 levels among AAs. Across all race and ethnic groups, those with a diagnosis of CKD had higher levels of ptau181. When stratified by cognitive diagnosis, cognitively unimpaired Hispanics had higher ptau181 if they also had a diagnosis of CKD or diabetes. p-values ≤0.01. Differences in ptau181 levels were shown in a diverse community sample. Medical comorbidities had a differing effect on ptau181 levels particularly among Hispanics even without cognitive impairment. Findings support the need for future work to consider comorbid conditions when examining the utility of ptau181.

Associations of urinary biomarkers of phthalates, phenols, parabens, and organophosphate esters with glycemic traits in pregnancy: The Healthy Start Study

BackgroundCertain endocrine-disrupting chemicals (EDCs) are widespread in consumer products and may alter glucose metabolism. However, the impact of EDC exposures on glucose and insulin regulation during pregnancy is incompletely understood, despite potential adverse consequences for maternal and infant health. We estimated associations between 37 urinary biomarkers of EDCs and glucose-insulin traits among pregnant women. MethodsSeventeen phthalate or phthalate substitute metabolites, six environmental phenols, four parabens, and ten organophosphate ester metabolites were quantified in mid-pregnancy urine from 298 participants in the Healthy Start Study. Fasting blood glucose, insulin, and hemoglobin A1c were assessed concurrently, and Homeostasis Model Assessment 2-Insulin Resistance (HOMA2-IR) was calculated. Gestational diabetes diagnoses and screening results were obtained from medical records for a subset of participants. We estimated associations between each EDC and outcome separately using linear and robust Poisson regression models and analyzed EDC mixture effects. ResultsThe EDC mixture was positively associated with glucose, insulin, and HOMA2-IR, although overall associations were attenuated after adjustment for maternal BMI. Two mixture approaches identified di(2-ethylhexyl) phthalate (DEHP) metabolites as top contributors to the mixture's positive associations. In single-pollutant models, DEHP metabolites were positively associated with fasting glucose, fasting insulin, and HOMA2-IR even after adjustment for maternal BMI. For example, each interquartile range increase in log2-transformed mono(2-ethyl-5-oxohexyl) phthalate was associated with 2.4 mg/dL (95% confidence interval (CI): 1.1, 3.6) higher fasting glucose, 11.8% (95%CI: 3.6, 20.5) higher fasting insulin, and 12.3% (95%CI: 4.2, 21.1) higher HOMA2-IR. Few EDCs were associated with hemoglobin A1c or with a combined outcome of impaired glucose tolerance or gestational diabetes. DiscussionExposures to phthalates and particularly DEHP during pregnancy are associated with altered glucose-insulin regulation. Disruptions in maternal glucose metabolism during pregnancy may contribute to adverse pregnancy outcomes including gestational diabetes and fetal macrosomia, and associated long-term consequences for maternal and child health.

“Predicting diabetic kidney disease in youth with type 1 diabetes: Insights from genetic risk assessment”

ObjectiveDiabetic kidney disease (DKD) is influenced by multiple factors, yet its precise progression mechanisms remain largely unclear. This study aimed to create a clinical risk-scoring system based on genetic polymorphisms in the AFF3, CARS, CERS2, ERBB4, GLRA3, RAET1L, TMPO, and ZMIZ1 genes. MethodsThe study included a DKD group diagnosed with diabetic kidney disease before age 18 and a WDC group matched by age, gender, and age at diabetes diagnosis. Genetic data and clinical data from diabetes diagnosis to moderately increased albuminuria (MIA) detection were compared between the groups. ResultsAmong 43 DKD cases, 22 were girls and 21 were boys. At MIA diagnosis, mean body weight SDS was −0.24 ± 0.94, height SDS was 0.34 ± 1.15, and BMI SDS was −0.26 ± 0.94. Systolic blood pressure was at the 72nd percentile (2–99), and diastolic blood pressure was at the 74th percentile (33–99). Significant differences in rs267734, rs267738, and rs942263 polymorphisms were found between DKD and non-complication diabetic groups (13[30.2 %] vs 5[11.6 %], p = 0.034; 14[32.6 %] vs 5[11.6 %], p = 0.019; 26[60.5 %] vs 40[93 %], p < 0.001). ConclusionSeveral factors were identified as significant in DKD onset, including low follow-up weight SDS, elevated diastolic blood pressure, presence of rs267734, and absence of rs942263 polymorphisms. The model demonstrated a specificity of 81.4 % and a sensitivity of 74.4 %.

Needle-Shaped Biosensors for Precision Diagnoses: From Benchtop Development to In Vitro and In Vivo Applications.

To achieve the accurate recognition of biomarkers or pathological characteristics within tissues or cells, in situ detection using biosensor technology offers crucial insights into the nature, stage, and progression of diseases, paving the way for enhanced precision in diagnostic approaches and treatment strategies. The implementation of needle-shaped biosensors (N-biosensors) presents a highly promising method for conducting in situ measurements of clinical biomarkers in various organs, such as in the brain or spinal cord. Previous studies have highlighted the excellent performance of different N-biosensor designs in detecting biomarkers from clinical samples in vitro. Recent preclinical in vivo studies have also shown significant progress in the clinical translation of N-biosensor technology for in situ biomarker detection, enabling highly accurate diagnoses for cancer, diabetes, and infectious diseases. This article begins with an overview of current state-of-the-art benchtop N-biosensor designs, discusses their preclinical applications for sensitive diagnoses, and concludes by exploring the challenges and potential avenues for next-generation N-biosensor technology.

Relationship between types and levels of free fatty acids, peripheral insulin resistance, and oxidative stress in T2DM: A case-control study.

Free Fatty Acids (FFAs) are vital for energy homeostasis and the pathogenesis of a variety of diseases, including diabetes. For the first time, we presumed and investigated the types and levels of FFAs and their links to Insulin Resistance (IR) and Oxidative Stress (OS) in T2DM. A case-control study was conducted on 60 individuals with diabetes, 60 prediabetics with IFG, and 60 control groups. A Gas Chromatography Flame Ionization Detector (GC-FID) was used to estimate FFAs, which were then classified based on length and saturation. Indeed, antioxidant parameters such as TAC, MDA levels, PON-1, SOD-3, and CAT activity were assessed. Higher levels of LCFFA, SFFA, USFFA, and total FFA were found in people with diabetes and prediabetes. These levels were also linked to higher levels of HOMA-IR, BMI, FBS, HbA1C, and MDA, but lower levels of antioxidants. Furthermore, adjusting the above FFAs with age, sex, and antihypertensive medication increased T2DM development. SCFFA and ω3/6 fatty acids had a negative relationship with HOMA-IR, FBS, and insulin and a positive relationship with TAC. Adjusted SCFFA reduces T2DM risk. According to our models, total FFA is utilized to diagnose diabetes (AUC = 83.98, cut-off > 919 μM) and SCFFA for prediabetes (AUC = 82.32, cut-off < 39.56 μM). Total FFA (≥ 776 μM), LCFFA (≥ 613 μM), SFFA (≥ 471 μM), and USFFA (≥ 398 μM) all increase the risk of T2DM by increasing OS, BMI, and HOMA-IR. On the other hand, SCFFAs (≥ 38.7 μM) reduce the risk of T2DM by reducing BMI, HOMA-IR, and OS. SCFFAs and total FFAs can be used for the diagnosis of prediabetes and diabetes, respectively.

Genotype-first analysis in an unselected health system-based population reveals variable phenotypic severity of COL4A5 variants.

Our knowledge of X-linked Alport Syndrome [AS] comes mostly from selected cohorts with more severe disease. We examined the phenotypic spectrum of X-linked AS in males and females with a genotype-based approach using data from the Geisinger MyCode DiscovEHR study, an unselected health system-based cohort with exome sequencing and electronic health records. Patients with COL4A5 variants reported as pathogenic (P) or likely pathogenic (LP) in ClinVar, or protein-truncating variants (PTVs), were each matched with up to 5 controls without COL4A3/4/5 variants by sociodemographics, diabetes diagnosis, and year of first outpatient encounter. AS-related phenotypes included dipstick hematuria, bilateral sensorineural hearing loss (BSHL), proteinuria, decreased eGFR, and ESKD. Out of 170,856 patients, there were 29 hemizygous males (mean age 52.0 y [SD 20.0]) and 55 heterozygous females (mean age 59.3 y [SD 18.8]) with a COL4A5 P/LP variant, including 48 with the hypomorphic variant p.Gly624Asp. Overall, penetrance (having any AS phenotypic feature) was highest for non-p.Gly624Asp P/LP variants (males: 94%, females: 85%), intermediate for p.Gly624Asp (males: 77%, females: 69%), compared to controls (males: 32%; females: 50%). The proportion with ESKD was highest for males with P/LP variants (44%), intermediate for males with p.Gly624Asp (15%) and females with P/LP variants (10%), compared to controls (males: 3%, females 2%). Only 47% of individuals with COL4A5 had completed albuminuria screening, and a minority were taking renin-angiotensin aldosterone system (RAAS) inhibitors. Only 38% of males and 16% of females had a known diagnosis of Alport syndrome or thin basement membrane disease. In an unselected cohort, we show increased risks of AS-related phenotypes in men and women compared to matched controls, while showing a wider spectrum of severity than has been described previously and variability by genotype. Future studies are needed to determine whether early genetic diagnosis can improve outcomes in Alport Syndrome.

Efficient mixed-potential acetone sensor with yttria-stabilized zirconia and porous Co3O4 nanofoam sensing electrode for hazardous gas monitoring and breath analysis

For hazardous gas monitoring and non-invasive diagnosis of diabetes using breath analysis, porous foams assembled by Co3O4 nanoparticles were designed as sensing electrode materials to fabricate efficient yttria-stabilized zirconia (YSZ)-based acetone sensors. The sensitivity of the sensors was improved by varying the sintering temperature to regulate the morphology. Compared to other materials sintered at different temperatures, the porous Co3O4 nanofoams sintered at 800 °C exhibited the highest electrochemical catalytic activity during the electrochemical test. The response of the corresponding Co3O4-based sensor to 10 ppm acetone was –77.2 mV and it exhibited fast response and recovery times. Moreover, the fabricated sensor achieved a low detection limit of 0.05 ppm and a high sensitivity of –56 mV/decade in the acetone concentration range of 1–20 ppm. The sensor also exhibited excellent repeatability, acceptable selectivity, good O2/humidity resistance, and long-term stability during continuous measurements for over 30 days. Moreover, the fabricated sensor was used to determine the acetone concentration in the exhaled breaths of patients with diabetic ketosis. The results indicated that it could distinguish between healthy individuals and patients with diabetic ketosis, thereby proving its abilities to diagnose and monitor diabetic ketosis. Based on its excellent sensitivity and exhaled breath measurement results, the developed sensor has broad application prospects.

Health Care Use Among Cancer Patients With Diabetes, National Health and Nutrition Examination Survey, 2017-2020.

Diabetes is a common comorbidity among people with cancer. The objective of our study was to examine patterns of health care use among patients with cancer and either type 2 diabetes or prediabetes. We used data from the National Health and Nutrition Examination Survey (NHANES) for 2017-2020. The study population included US adults aged 18 years or older who were diagnosed with any cancer and type 2 diabetes or prediabetes (established by self-report and/or hemoglobin A1c measurement). We used Poisson and multivariate logistic regression models to determine the effect of comorbidity on health care use, defined as health care visits and overnight stays in a hospital. Of 905 cancer patients representing 27,180,715 people in the US, 24.4% had a type 2 diabetes diagnosis, and 25.8% had a prediabetes diagnosis. Patients with cancer and prediabetes had a significantly higher rate of health care visits (incidence rate ratio = 1.11; 95% CI, 1.01-1.22; P = .03) than patients with cancer only. We found no significant association between having cancer and type 2 diabetes and the number of health care visits or overnight hospital stays compared with patients with cancer only. More emphasis should be placed on optimal care coordination among people with cancer and other conditions, such as diabetes and prediabetes, to reduce the impact of comorbidity on health care use. Interventions integrated with technology to provide timely access to education on preventing or managing diabetes and prediabetes among cancer patients are warranted.

Effect of utilizing either a self-reported questionnaire or administrative data alone or in combination on the findings of a randomized controlled trial of the long-term effects of antenatal corticosteroids.

A combination of self-reported questionnaire and administrative data could potentially enhance ascertainment of outcomes and alleviate the limitations of both in follow up studies. However, it is uncertain how access to only one of these data sources to assess outcomes impact study findings. Therefore, this study aimed to determine whether the study findings would be altered if the outcomes were assessed by different data sources alone or in combination. At 50-year follow-up of participants in a randomized trial, we assessed the effect of antenatal betamethasone exposure on the diagnosis of diabetes, pre-diabetes, hyperlipidemia, hypertension, mental health disorders, and asthma using a self-reported questionnaire, administrative data, a combination of both, or any data source, with or without adjudication by an expert panel of five clinicians. Differences between relative risks derived from each data source were calculated using the Bland-Altman approach. There were 424 participants (46% of those eligible, aged 49 years, SD 1, 50% male). There were no differences in study outcomes between participants exposed to betamethasone and those exposed to placebo when the outcomes were assessed using different data sources. When compared to the study findings determined using adjudicated outcomes, the mean difference (limits of agreement) in relative risks derived from other data sources were: self-reported questionnaires 0.02 (-0.35 to 0.40), administrative data 0.06 (-0.32 to 0.44), both questionnaire and administrative data 0.01 (-0.41 to 0.43), and any data source, 0.01 (-0.08 to 0.10). Utilizing a self-reported questionnaire, administrative data, both questionnaire and administrative data, or any of these sources for assessing study outcomes had no impact on the study findings compared with when study outcomes were assessed using adjudicated outcomes.

How to ensure a type 2 diabetes diagnosis is made promptly

How to ensure a type 2 diabetes diagnosis is made promptly

Risk of healthcare visits from influenza in subjects with diabetes and impacts of early vaccination

IntroductionThe objective of this study was to determine the burden of influenza disease in patients with or without diabetes in a population of American adults to understand the benefits of...

Glucometrics utilisation in an urban teaching hospital in ireland: current practice and future aims.

Dysglycaemia in hospitalised patients is associated with poorer clinical outcomes, including cardiovascular events, longer hospital stays, and increased risk of mortality. Therefore, glucose monitoring is necessary to achieve best outcomes. This audit assesses use of point-of-care (POC) blood glucose (BG) testing in Tallaght University Hospital (TUH) over an 8-day period. It evaluates compliance with international and TUH glucose monitoring protocols and determines frequency of diabetes team consultations for inpatient adults. Data from an 8-day period (12/03/2023-19/03/2023) were extracted from the TUH COBAS-IT system and analysed. Invalid tests were excluded. Hyperglycaemia was defined as ≥ 10mmol/L and hypoglycaemia as ≤ 3.9mmol/L. Persistent hyperglycaemia was defined as two BG results of ≥ 10mmol/L. A chart review was conducted on adult patients with persistent hyperglycaemia to assess for HbA1C results, diabetes diagnosis, and diabetes consult. 3,530 valid tests were included and analysed. 674 individual patients had tests done. 1,165 tests (33.00%) were hyperglycaemic and 75 (2.12%) were hypoglycaemic. 68.25% of adults with persistent hyperglycaemia had an HbA1C test performed or documented within three months. 42.71% of inpatient adults with persistent hyperglycaemia and a known diabetes diagnosis received a consult from the diabetes team. Increased adherence to hospital protocols for testing HbA1C in adults with persistent hyperglycaemia could improve treatment and clinical outcomes. Increased diabetes team consultation could facilitate appropriate treatment and improve patient outcomes in persistently hyperglycaemic adult patient populations.

$$D_MD_RDF$$: diabetes mellitus and retinopathy detection framework using artificial intelligence and feature selection

AbstractDiabetes mellitus is one of the most common diseases affecting patients of different ages. Diabetes can be controlled if diagnosed as early as possible. One of the serious complications of diabetes affecting the retina is diabetic retinopathy. If not diagnosed early, it can lead to blindness. Our purpose is to propose a novel framework, named $$D_MD_RDF$$ D M D R D F , for early and accurate diagnosis of diabetes and diabetic retinopathy. The framework consists of two phases, one for diabetes mellitus detection (DMD) and the other for diabetic retinopathy detection (DRD). The novelty of DMD phase is concerned in two contributions. Firstly, a novel feature selection approach called Advanced Aquila Optimizer Feature Selection ($$A^2OFS$$ A 2 O F S ) is introduced to choose the most promising features for diagnosing diabetes. This approach extracts the required features from the results of laboratory tests while ignoring the useless features. Secondly, a novel classification approach (CA) using five modified machine learning (ML) algorithms is used. This modification of the ML algorithms is proposed to automatically select the parameters of these algorithms using Grid Search (GS) algorithm. The novelty of DRD phase lies in the modification of 7 CNNs using Aquila Optimizer for the classification of diabetic retinopathy. The reported results concerning the DMD datasets shows that AO reports best performance metrics in the feature selection process with the help of modified ML classifiers. The best achieved accuracy is 98.65% with the GS-ERTC model and max-absolute scaling on the “Early Stage Diabetes Risk Prediction Dataset” dataset. Also, from the reported results concerning the DRD datasets, the AOMobileNet is considered a suitable model for this problem as it outperforms the other modified CNN models with accuracy of 95.80% on the “The SUSTech-SYSU dataset” dataset.

Diabetic Retinopathy Diagnosis Using Supervised Classification

Diabetic Retinopathy Diagnosis Using Supervised Classification

Personalized diabetes diagnosis using machine learning and electronic health records

Diabetes mellitus (DM) poses a significant health challenge globally, necessitating accurate and timely diagnosis for effective management. Conventional diagnostic methods often struggle to address the multifaceted nature of diabetes and the requisite lifestyle adjustments. In this study, we propose a data-driven approach utilizing machine learning techniques to enhance diabetes diagnosis. By leveraging extensive patient attributes and medical records, machine learning algorithms can uncover intricate patterns and correlations. Our methodology, validated on the PIMA India dataset, demonstrates promising results. The random forest model achieved the highest accuracy of 87%, followed closely by gradient boost at 90%. Notably, XGBoost and CATBoost models attained a peak accuracy of 90.9%. These findings underscore the potential of machine learning in transforming diabetes diagnosis. Beyond improving diagnostic accuracy, our approach aims to guide individuals towards healthier lifestyles. Intelligent systems driven by machine learning hold promise for revolutionizing diabetes management, ultimately leading to better patient outcomes and more effective health care delivery.