Search

A medical disorder known as diabetic retinopathy (DR) affects people who suffer from diabetes. Many people are visually impaired due to DR. Primary cause of DR in patients is high blood sugar, and it affects blood vessels available in the retinal cell. The recent advancement in deep learning and computer vision methods, and their automation applications can recognize the presence of DR in retinal cells and vessel images. Authors have proposed an attention-based hybrid model to recognize diabetes in early stage to prevent harmful clauses. Proposed methodology uses DenseNetl21 architecture for convolution learning and then, the feature vector will be enhanced with channel and spatial attention model. The proposed architecture also simulates binary and multiclass classification to recognize the infection and the spreading of disease. Binary classification recognizes DR images either positive or negative, while multiclass classification represents an infection on a scale of 0-4. Simulation of the proposed methodology has achieved 98.57% and 99.01% accuracy for multiclass and binary classification, respectively. Simulation of the study also explored the impact of data augmentation to make the proposed model robust and generalized. Attention-based deep learning model has achieved remarkable accuracy to detect diabetic infection from retinal cellular images. Key words: deep learning, computer vision, diabetic retinopathy, disease detection, cellular images attention model, Introduction Early illness detection improves the effectiveness of therapy in the healthcare sector. Disease detection from cellular must need to prevent infection. Diabetes is an illness that increases when insufficient [...]

: Objective: To identify the optimal AUSDRISK threshold score to screen for pre?diabetes and diabetes. Methods: A total of 406 adult patients not diagnosed with diabetes were screened in General Practices (GP) between May and October 2019. All patients received a point of care (POC) HbA1c test. HbA1c test results were categorised into diabetes (?6.5% or ?48 mmol/mol), pre?diabetes (5.7?6.4% or 39?47 mmol/mol), or normal ( Results: Of these patients, 9 (2%) had undiagnosed diabetes and 60 (15%) had pre?diabetes. A Receiver Operator Characteristic (ROC) curve was constructed to predict the presence of pre?diabetes and diabetes; the area under the ROC curve was 0.72 (95%CI 0.65?0.78) indicating modest predictive ability. The optimal threshold cut point for AUSDRISK score was 17 (sensitivity 76%, specificity 61%, + likelihood ratio (LR) 1.96, ? likelihood ratio of 0.39) while the accepted cut point of 12 performed less well (sensitivity 94%, specificity 23%, +LR=1.22 ?LR+0.26). Conclusions: The AUSDRISK tool has the potential to be used as a screening tool for pre?diabetes/diabetes in GP practices. A cut point of ?17 would potentially identify 75% of all people at risk and three in 10 sent for further testing would be positive for prediabetes or diabetes. Implications for public health: Routine case?finding in high?risk patients will enable GPs to intervene early and prevent further public health burden from the sequelae of diabetes., Pre?diabetes is a condition where a person's fasting glucose is elevated but has not reached the diabetes threshold and it is a significant risk factor for the development of type [...]