scholarly journals Identification of Diabetic Retinopathy through Machine Learning

2021 ◽  
Vol 2021 ◽  
pp. 1-8
Author(s):  
Malik Bader Alazzam ◽  
Fawaz Alassery ◽  
Ahmed Almulihi
Keyword(s):  
Machine Learning ◽  
Diabetic Retinopathy ◽  
Diagnostic Accuracy ◽  
Cross Sectional Study ◽  
Disease Classification ◽  
Ophthalmological Examination ◽  
Restricted Boltzmann Machines ◽  
Cross Sectional ◽  
Diabetic Retinopathy Screening

A cross-sectional study of patients with suspected diabetic retinopathy (DR) who had an ophthalmological examination and a retinal scan is the focus of this research. Specialized retinal images were analyzed and classified using OPF and RBM models (restricted Boltzmann machines). Classification of retinographs was based on the presence or absence of disease-related retinopathy (DR). The RBM and OPF models extracted 500 and 1000 characteristics from the images for disease classification after the system training phase for the recognition of retinopathy and normality patterns. There were a total of fifteen different experiment series, each with a repetition rate of 30 cycles. The study included 73 diabetics (a total of 122 eyes), with 50.7% of them being men and 49.3% being women. The population was on the older side, at 59.7 years old on average. The RBM-1000 had the highest overall diagnostic accuracy (89.47) of any of the devices evaluated. The RBM-500 had a better autodetection system for DR signals in fundus images than the competition (100% sensitivity). In terms of specificity, RBM-1000 and OPF-1000 correctly identified all of the images that lacked DR signs. In particular, the RBM model of machine learning automatic disease detection performed well in terms of diagnostic accuracy, sensitivity, and application in diabetic retinopathy screening.

scholarly journals Diabetic Retinopathy Screening Ratio Is Improved When Using a Digital, Nonmydriatic Fundus Camera Onsite in a Diabetes Outpatient Clinic

2016 ◽  
Vol 2016 ◽  
pp. 1-10 ◽  
Author(s):  
Pia Roser ◽  
Hannes Kalscheuer ◽  
Jan B. Groener ◽  
Daniel Lehnhoff ◽  
Roman Klein ◽  
...  
Keyword(s):  
Diabetic Retinopathy ◽  
Outpatient Clinic ◽  
Microvascular Complications ◽  
Diabetic Polyneuropathy ◽  
Cross Sectional Study ◽  
Newly Diagnosed ◽  
Cross Sectional ◽  
Fundus Camera ◽  
Diabetic Retinopathy Screening ◽  
Single Field

Objective. To evaluate the effect of onsite screening with a nonmydriatic, digital fundus camera for diabetic retinopathy (DR) at a diabetes outpatient clinic.Research Design and Methods. This cross-sectional study included 502 patients, 112 with type 1 and 390 with type 2 diabetes. Patients attended screenings for microvascular complications, including diabetic nephropathy (DN), diabetic polyneuropathy (DP), and DR. Single-field retinal imaging with a digital, nonmydriatic fundus camera was used to assess DR. Prevalence and incidence of microvascular complications were analyzed and the ratio of newly diagnosed to preexisting complications for all entities was calculated in order to differentiate natural progress from missed DRs.Results. For both types of diabetes, prevalence of DR was 25.0% (n=126) and incidence 6.4% (n=32) (T1DM versus T2DM: prevalence: 35.7% versus 22.1%, incidence 5.4% versus 6.7%). 25.4% of all DRs were newly diagnosed. Furthermore, the ratio of newly diagnosed to preexisting DR was higher than those for DN (p=0.12) and DP (p=0.03) representing at least 13 patients with missed DR.Conclusions. The results indicate that implementing nonmydriatic, digital fundus imaging in a diabetes outpatient clinic can contribute to improved early diagnosis of diabetic retinopathy.

scholarly journals Usefulness of corneal esthesiometry for screening diabetic retinopathy

2003 ◽  
Vol 37 (5) ◽  
pp. 609-615 ◽  
Author(s):  
Lênio Souza Alvarenga ◽  
Elisabeth Nogueira Martins ◽  
Gustavo Teixeira Grottone ◽  
Paulo Henrique Ávila Morales ◽  
Augusto Paranhos Jr ◽  
...  
Keyword(s):  
Diabetic Retinopathy ◽  
Roc Curve ◽  
Operating Characteristic ◽  
Screening Program ◽  
Cross Sectional Study ◽  
Sectional Study ◽  
Cross Sectional ◽  
Diabetic Retinopathy Screening ◽  
The City ◽  
Retinopathy Screening

OBJECTIVE: To assess the usefulness of corneal esthesiometry for screening diabetic retinopathy. METHODS: A cross-sectional study was carried out comprising 575 patients attending a diabetic retinopathy-screening program in the city of São Paulo. Corneal esthesiometry was assessed with the Cochet-Bonnet esthesiometer. The presence of diabetic retinopathy was detected with indirect fundoscopy. The validity of corneal esthesiometry in identifying diabetic retinopathy was evaluated by the Receiver Operating Characteristic (ROC) curve. RESULTS: Sensitivity and specificity analyses of the corneal esthesiometry for detecting the stages of diabetic retinopathy using different cut-offs showed values less than 80%. The best indices (72.2% sensitivity and 57.4% specificity) were obtained for the identification of patients with proliferative diabetic retinopathy. CONCLUSIONS: In the study series, corneal esthesiometry was not a good indicator of diabetic retinopathy.

scholarly journals Comparison of true-colour wide-field confocal scanner imaging with standard fundus photography for diabetic retinopathy screening

2020 ◽  
pp. bjophthalmol-2019-315269 ◽  
Author(s):  
Abraham Olvera-Barrios ◽  
Tjebo FC Heeren ◽  
Konstantinos Balaskas ◽  
Ryan Chambers ◽  
Louis Bolter ◽  
...  
Keyword(s):  
Diabetic Retinopathy ◽  
Kappa Statistic ◽  
Cross Sectional Study ◽  
Fundus Photography ◽  
Diabetic Patients ◽  
Wide Field ◽  
Cross Sectional ◽  
Imaging Protocol ◽  
Diabetic Retinopathy Screening ◽  
Diabetic Eye Screening

BackgroundScreening of diabetic retinopathy (DR) reduces blindness by early identification of retinopathy. This study compares DR grades derived from a two-field imaging protocol from two imaging platforms, one providing a single 60-degree horizontal field of view (FOV) and the other, a standard 45-degree FOV.MethodsCross-sectional study which included 1257 diabetic patients aged ≥18 years attending their DR screening visit in the English National Diabetic Eye Screening Programme (NDESP). Patients with maculopathy (M1), preproliferative (R2) or proliferative DR (R3) were referred to an ophthalmologist. Patients with ungradable images (U) are examined in a slit-lamp biomicroscopy clinic. Image acquisition under mydriasis of two images per eye was carried out with the EIDON and with standard fundus cameras. Evaluation was performed by masked graders.ResultsAgreement after consensus with kappa statistic was 0.89 (quadratic weights (95% CI 0.87 to 0.92)) for NDESP severity grade, 0.88 (quadratic weights (95% CI 0.82 to 0.94)) for referable disease and 0.92 (linear weights (95% CI 0.88 to 0.95)) for maculopathy. The EIDON detected clinically relevant DR features outside the 45-degree fields in two patients (0.16%): one with intraretinal microvascular abnormalities (IRMAs) and one with neovascularisation. In eight patients (0.64%), the EIDON allowed DR feature visualisation inside the 45-degree fields that were not identified in the NDESP images: three patients (0.24%) with IRMA and five patients (0.40%) with maculopathy. The rates of ungradable encounters were 12 (0.95%) and 13 (1.03%) with the EIDON and NDESP images, respectively.ConclusionThe EIDON identifies a small number of additional patients with referable disease which are not detected with standard imaging. This is due to the EIDON finding disease outside the standard FOV and greater clarity finding disease within the standard FOV.

scholarly journals Verification of 3D-printed universal smartphone retinal imaging adapter against conventional fundus camera imaging for diabetic retinopathy screening

2020 ◽  
Vol 2 (3) ◽  
pp. 190-202
Author(s):  
Yuen Keat Gan ◽  
Amir Samsudin
Keyword(s):  
Diabetic Retinopathy ◽  
Retinal Imaging ◽  
Cross Sectional Study ◽  
Kappa Statistics ◽  
Cross Sectional ◽  
Fundus Camera ◽  
Diabetic Retinopathy Screening ◽  
Camera Imaging ◽  
Institutional Review ◽  
Kappa Agreement

Introduction: Screening for diabetic retinopathy (DR) is critical in preventing visual loss. However, current tools are expensive, bulky and sensitive, thus limiting screening coverage, especially in developing areas such as the interior of Borneo. Smartphone-assisted devices may provide an alternative and this study seeks to determine the level of agreement between a smartphone retinal imaging adapter (SRIA) against conventional ones. Materials and methods: This was a cross-sectional study with Institutional Review Board approval from the Medical Ethics Board of University of Malaya Medical Centre. A total of 284 eyes from 142 patients included underwent retinal imaging using a conventional fundus camera and the SRIA. The images were graded according to Early Treatment of Diabetic Retinopathy Study (ETDRS) classification. Agreement between both modalities was calculated using Cohen’s Kappa statistics. Results: The Kappa agreement between SRIA and conventional fundus imaging in grading individual ETDRS stages stood at 0.648 (p < 0.001), achieving up to 0.752 (p < 0.001) when differentiating between no DR, non-proliferative DR, and proliferative DR. Conclusion: DR grading SRIA and conventional fundus camera imaging were comparable. SRIA can be useful in eye screenings but still needs improvement.

scholarly journals Performance and limitation of machine learning algorithms for diabetic retinopathy screening: A meta-analysis (Preprint)

2020 ◽  
Author(s):  
Jo-Hsuan Wu ◽  
Tin-Yan Alvin Liu ◽  
Wan-Ting Hsu ◽  
Jennifer Hui-Chun Ho ◽  
Chien-Chang Lee
Keyword(s):  
Machine Learning ◽  
Diabetic Retinopathy ◽  
Diagnostic Accuracy ◽  
State Of The Art ◽  
Meta Analysis ◽  
External Validation ◽  
Screening Method ◽  
Machine Learning Algorithms ◽  
Diabetic Retinopathy Screening ◽  
Fundus Photographs

BACKGROUND Standardly diagnosed by human experts, the high prevalence of diabetic retinopathy (DR) warrants a more efficient screening method. Although machine learning (ML)-based automated DR diagnosis has gained attention due to recent approval of IDx-DR, performance of this tool has not be examined systematically, and the best ML technique for utilization in real-world setting has not been discussed. OBJECTIVE To examine systematically the overall diagnostic accuracy of ML in diagnosing DR of different categories based on color fundus photographs and to determine the state-of-the-art ML approach. METHODS Published studies in PubMed and EMBASE were searched from inception to June, 2020. Studies were screened for relevant outcomes, publication types, and data sufficiency, and a total of 60 (2.8%) out of 2128 studies were retrieved after study selection. Extraction of data was performed by 2 authors according to PRISMA, and the quality assessment was performed according to QUADUS-2. Meta-analysis of diagnostic accuracy was pooled using a bivariate random-effects model. The main outcomes included diagnostic accuracy, sensitivity, and specificity of ML in diagnosing DR based on color fundus photographs, as well as the performances of different major types of ML algorithms. RESULTS The primary meta-analysis included 60 color fundus photograph studies (445,175 interpretations). Overall, ML demonstrated high accuracy in diagnosing DR of various categories, with a pooled AUROC from 0.97 (95% CI: 0.96, 0.99) to 0.99 (95%CI: 0.98, 1.00). The performance of ML in detecting more-than-mild DR (mtmDR) was robust (Sen: 0.95, AUROC: 0.97), and by subgroup analyses, we observed that robust performance of ML was not limited to benchmark datasets (Sen: 0.92; AUROC: 0.96) but could be generalized to images collected in clinical practice (Sen: 0.97; AUROC: 097). Neural network was the most widely utilized method, and the subgroup analysis revealed a pooled AUROC of 0.98 (95% CI: 0.96, 0.99) for studies that utilized neural networks to diagnose mtmDR. CONCLUSIONS This meta-analysis demonstrated high diagnostic accuracy of ML algorithms in detecting diabetic retinopathy on color fundus photographs, suggesting that state-of-the-art, ML-based DR screening algorithms are likely ready for clinical applications. However, a significant portion of the earlier published studies had methodology flaws, such as the lack of external validation and presence of spectrum bias. The results of these studies should be interpreted with caution.

STUDY OF HISTOPATHOLOGY IN PATIENTS WITH HODGKIN AND NON - HODGKIN LYMPHOMA

2016 ◽  
pp. 59-65 ◽  
Author(s):  
Van Mao Nguyen
Keyword(s):  
Lymph Node ◽  
Hodgkin Lymphoma ◽  
Cross Sectional Study ◽  
University Hospital ◽  
Cross Sectional ◽  
Histopathological Classification ◽  
Non Hodgkin Lymphoma ◽  
Key Words Lymphoma ◽  
Malignant Grade

Background: Lymphoma is one of the most ten common cancers in the world as well as in Vietnam which has been ever increasing. It was divided into 2 main groups Hodgkin and non – Hodgkin lymphoma in which non-Hodgkin lymphoma appeared more frequency, worse prognosis and different therapy. Objectives: - To describe some common characteristics in patients with non – Hodgkin lymphoma; - To determine the proportion between Hodgkin and non- Hodgkin lymphoma, histopathological classification of classical Hodgkin by modified Rye 1966 and non-Hodgkin lymphoma by Working Formulation (WF) of US national oncology institute 1982. Materials and Method: This cross-sectional study was conducted on 65 patients with Hodgkin and non- Hodgkin lymphoma diagnosed definitely by histopathology at Hue Central Hospital and Hue University Hospital. Results:. The ratio of male/female for the non-Hodgkin lymphoma was 1.14/1, the most frequent range of age was 51-60 accounting for 35%, not common under 40 years. Non - Hodgkin lymphoma appeared at lymph node was the most common (51.7%), at the extranodal site was rather high 48.3%. The non - Hodgkin lymphoma proportion was predominant 92.3% comparing to the Hodgkin lymphoma only 7.7%; The most WF type was WF7 (53.3%), following the WF6 18,3% and WF5 11,7%; The intermediate malignancy grade of non- Hodgkin lymphoma was the highest proportion accouting for 85%, then the low and the high one 8.3% and 6.7% respectively. Conclusion: The histopathological classification and the malignant grade of lymphoma for Hodgkin and non - Hodgkin lymphoma played a practical role for the prognosis and the treatment orientation, also a fundamental one for the modern classification of non - Hodgkin lymphoma nowadays. Key words: lymphoma, Hodgkin lymphoma, non-Hodgkin lymphoma, classication, grade, histopathology, lymph node

scholarly journals Role of human epididymis protein 4 for detection of ovarian carcinoma in adnexal masses: A pilot study

2016 ◽  
Author(s):  
Nidhi Bansal ◽  
A. Suneja ◽  
K. Guleria ◽  
N. B. Vaid ◽  
K. Mishra ◽  
...  
Keyword(s):  
Diagnostic Accuracy ◽  
Cross Sectional Study ◽  
Ca 125 ◽  
Serum Samples ◽  
Cross Sectional ◽  
Adnexal Masses ◽  
Good Discriminator ◽  
Risk Of Malignancy ◽  
Risk Of Malignancy Index ◽  
Benign Tumours

Introduction: HE4 is a novel tumour biomarker used for early diagnosis of ovarian cancer. This study evaluated the diagnostic accuracy of HE4 alone and in combination with CA125, risk of malignancy index (RMI), risk of malignancy algorithm (ROMA). Methods: It was a cross sectional study conducted recruiting 88 women with adnexal masses who were planned for surgery. After baseline work up and ultrasound examination, serum samples were collected for estimation of CA 125 and HE4 levels. Serum HE4 levels were estimated using ELISA kit. RMI and ROMA score were calculated and diagnostic accuracy of HE4, CA 125, RMI, ROMA and their combination were compared. Cut off for HE4 and ROMA score were calculated using ROC curve. Results: Of 88 subjects, 66 were analyzed with 19 malignant (including 5 LMP) and 47 benign cases. The median value of HE4 among malignant cases was found to be significantly higher than among the benign cases. PPV and NPV of HE4 at a cut off 130.8 pMol/ml was 85.7% and 77.9% respectively. Highest PPV (88.9%) with acceptable NPV (80.7%) was found with ROMA followed by HE4 (PPV 85.7%; NPV 77.97%), RMI (PPV 76.92%; NPV 83%) and CA125 (PPV 52%; NPV 80.85%). Conclusion: HE4 levels were lower in Indian population both in malignant and benign tumours as compared to other studies. HE4 is a good discriminator and gives best accuracy when it is combined with CA125 in a logistic algorithm, ROMA.

Sign in / Sign up

Export Citation Format

Share Document