Elif Yusufoğlu
Research Subjects
Elif Yusufoğlu has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
- Sabiha Güngör Kobat
Department of Ophthalmology, Fırat University, Elazig, Turkey
5 shared publications Sabiha Güngör Kobat
4 shared publications- Sinem Keser
Department of Ophthalmology, Fethi Sekin City Hospital, Elazığ 23280, Turkey
3 shared publications - Abdulkadir Sengur
Department of Electrical-Electronic Engineering, Faculty of Technology, Firat University, 23000 Elazig, Turkey
2 shared publications - Fatih Cem Gül
Universal Eye Center, Ophthalmology Clinic, Elazig, Turkey
2 shared publications - Mehmet Kalayci
Department of Medical Biochemistry, Elazig Education and Research Hospital, Elazig, Turkey
2 shared publications
Latest Publications
- Open Access
- 22 May 2026
- Published by MDPI AG in Tomography
- Vol. 12 (6) , 76
DenseViT-OCT: A Hybrid CNN-Transformer Architecture with Multi-Scale Dense Feature Aggregation for Automated Epiretinal Membrane Severity Classification
Background/Objectives: Epiretinal membrane (ERM) is a common vitreoretinal disorder characterized by fibrocellular proliferation on the inner retinal surface, often leading to progressive visual impairment. Accurate grading of ERM severity using optical coherence tomography (OCT) is critical for treatment planning and surgical decision-making; however, manual grading is labor-intensive and subjective. This study aims to develop an automated and reliable deep learning-based method for ERM severity classification. Methods: We propose DenseViT-OCT, a hybrid deep learning model that integrates dense convolutional neural networks (CNN) and vision transformers (ViT). The model introduces three key modules: Multi-Scale Dense Feature Aggregation (MDFA) for capturing hierarchical features across multiple spatial scales, Adaptive Feature Calibration (AFC) for enhancing feature discrimination through channel and spatial attention, and Cross-Attention Feature Fusion (CAFF) for enabling bidirectional interaction between convolutional and transformer representations. The model was trained and evaluated on 2195 OCT B-scan images obtained from 397 patients. Results: DenseViT-OCT achieved an overall accuracy of 94.76% on the internal four-class test set, outperforming 19 benchmark models, including ConvNeXt, EfficientNet, ViT, and Swin Transformers. The model demonstrated balanced performance with a macro-averaged precision of 93.76%, recall of 93.22%, F1-score of 93.47%, Cohen’s kappa of 92.62%, and macro-Area Under the Curve (AUC) of 98.95%. Ablation experiments confirmed the contribution of the proposed MDFA, AFC, CAFF, and deep supervision components, with the full model consistently outperforming reduced variants and standalone DenseNet121 and ViT-B/16 backbones. In repeated experiments across five random seeds, DenseViT-OCT also achieved the best mean accuracy (0.9399 ± 0.0052). External validation on the public multicenter OCTDL dataset, performed as binary ERM-versus-normal classification because of label availability, yielded 90.76% accuracy and 97.61% AUC, indicating promising generalization beyond the development cohort. Conclusions: DenseViT-OCT provides a robust framework for automated ERM severity classification from OCT B-scans. The combination of local CNN features, global transformer context, and dedicated fusion modules improves classification performance and yields clinically meaningful error patterns. Although further stage-wise multicenter validation, volumetric OCT analysis, and prospective clinical assessment are required, the proposed method shows promise as a research-oriented decision-support framework for B-scan-level ERM assessment.
- 22 May 2026
- Published by MDPI AG in Tomography
- Vol. 12 (6) , 76
Open AccessDenseViT-OCT: A Hybrid CNN-Transformer Architecture with Multi-Scale Dense Feature Aggregation for Automated Epiretinal Membrane Severity Classification
Background/Objectives: Epiretinal membrane (ERM) is a common vitreoretinal disorder characterized by fibrocellular proliferation on the inner retinal surface, often leading to progressive visual impairment. Accurate grading of ERM severity using optical coherence tomography (OCT) is critical for treatment planning and surgical decision-making; however, manual grading is labor-intensive and subjective. This study aims to develop an automated and reliable deep learning-based method for ERM severity classification. Methods: We propose DenseViT-OCT, a hybrid deep learning model that integrates dense convolutional neural networks (CNN) and vision transformers (ViT). The model introduces three key modules: Multi-Scale Dense Feature Aggregation (MDFA) for capturing hierarchical features across multiple spatial scales, Adaptive Feature Calibration (AFC) for enhancing feature discrimination through channel and spatial attention, and Cross-Attention Feature Fusion (CAFF) for enabling bidirectional interaction between convolutional and transformer representations. The model was trained and evaluated on 2195 OCT B-scan images obtained from 397 patients. Results: DenseViT-OCT achieved an overall accuracy of 94.76% on the internal four-class test set, outperforming 19 benchmark models, including ConvNeXt, EfficientNet, ViT, and Swin Transformers. The model demonstrated balanced performance with a macro-averaged precision of 93.76%, recall of 93.22%, F1-score of 93.47%, Cohen’s kappa of 92.62%, and macro-Area Under the Curve (AUC) of 98.95%. Ablation experiments confirmed the contribution of the proposed MDFA, AFC, CAFF, and deep supervision components, with the full model consistently outperforming reduced variants and standalone DenseNet121 and ViT-B/16 backbones. In repeated experiments across five random seeds, DenseViT-OCT also achieved the best mean accuracy (0.9399 ± 0.0052). External validation on the public multicenter OCTDL dataset, performed as binary ERM-versus-normal classification because of label availability, yielded 90.76% accuracy and 97.61% AUC, indicating promising generalization beyond the development cohort. Conclusions: DenseViT-OCT provides a robust framework for automated ERM severity classification from OCT B-scans. The combination of local CNN features, global transformer context, and dedicated fusion modules improves classification performance and yields clinically meaningful error patterns. Although further stage-wise multicenter validation, volumetric OCT analysis, and prospective clinical assessment are required, the proposed method shows promise as a research-oriented decision-support framework for B-scan-level ERM assessment.
- Open Access
- 17 December 2024
- Published by MDPI AG in Diagnostics
- Vol. 14 (24) , 2836
A Comprehensive CNN Model for Age-Related Macular Degeneration Classification Using OCT: Integrating Inception Modules, SE Blocks, and ConvMixer
Background/Objectives: Age-related macular degeneration (AMD) is a significant cause of vision loss in older adults, often progressing without early noticeable symptoms. Deep learning (DL) models, particularly convolutional neural networks (CNNs), demonstrate potential in accurately diagnosing and classifying AMD using medical imaging technologies like optical coherence to-mography (OCT) scans. This study introduces a novel CNN-based DL method for AMD diagnosis, aiming to enhance computational efficiency and classification accuracy. Methods: The proposed method (PM) combines modified Inception modules, Depthwise Squeeze-and-Excitation Blocks, and ConvMixer architecture. Its effectiveness was evaluated on two datasets: a private dataset with 2316 images and the public Noor dataset. Key performance metrics, including accuracy, precision, recall, and F1 score, were calculated to assess the method’s diagnostic performance. Results: On the private dataset, the PM achieved outstanding performance: 97.98% accuracy, 97.95% precision, 97.77% recall, and 97.86% F1 score. When tested on the public Noor dataset, the method reached 100% across all evaluation metrics, outperforming existing DL approaches. Conclusions: These results highlight the promising role of AI-based systems in AMD diagnosis, of-fering advanced feature extraction capabilities that can potentially enable early detection and in-tervention, ultimately improving patient care and outcomes. While the proposed model demon-strates promising performance on the datasets tested, the study is limited by the size and diversity of the datasets. Future work will focus on external clinical validation to address these limita-tions.
- 17 December 2024
- Published by MDPI AG in Diagnostics
- Vol. 14 (24) , 2836
Open AccessA Comprehensive CNN Model for Age-Related Macular Degeneration Classification Using OCT: Integrating Inception Modules, SE Blocks, and ConvMixer
Background/Objectives: Age-related macular degeneration (AMD) is a significant cause of vision loss in older adults, often progressing without early noticeable symptoms. Deep learning (DL) models, particularly convolutional neural networks (CNNs), demonstrate potential in accurately diagnosing and classifying AMD using medical imaging technologies like optical coherence to-mography (OCT) scans. This study introduces a novel CNN-based DL method for AMD diagnosis, aiming to enhance computational efficiency and classification accuracy. Methods: The proposed method (PM) combines modified Inception modules, Depthwise Squeeze-and-Excitation Blocks, and ConvMixer architecture. Its effectiveness was evaluated on two datasets: a private dataset with 2316 images and the public Noor dataset. Key performance metrics, including accuracy, precision, recall, and F1 score, were calculated to assess the method’s diagnostic performance. Results: On the private dataset, the PM achieved outstanding performance: 97.98% accuracy, 97.95% precision, 97.77% recall, and 97.86% F1 score. When tested on the public Noor dataset, the method reached 100% across all evaluation metrics, outperforming existing DL approaches. Conclusions: These results highlight the promising role of AI-based systems in AMD diagnosis, of-fering advanced feature extraction capabilities that can potentially enable early detection and in-tervention, ultimately improving patient care and outcomes. While the proposed model demon-strates promising performance on the datasets tested, the study is limited by the size and diversity of the datasets. Future work will focus on external clinical validation to address these limita-tions.
- 2 May 2024
- Published by Springer Nature in International Ophthalmology
- Vol. 44 (1) , 1-7
The effect of sodium hyaluronate on dry eye and corneal epithelial thickness following cataract surgery
Purpose To evaluate the effects of sodium hyaluronate drops on dry eye parameters and corneal epithelial thickness following cataract surgery. Methods The study included 84 patients who underwent uncomplicated phacoemulsification. In Group A, 0.15% sodium hyaluronate drops were added to the postoperative antibiotic/anti-inflammatory treatment. In Group B, only antibiotic/anti-inflammatory treatment was applied. Preoperatively and at 1 week and 1 month postoperatively, all the patients were evaluated in respect of tear break-up time (TBUT), the Schirmer test under anesthesia, the corneal fluorescein staining (CFS) score, mean central corneal thickness (CCT) and mean central corneal epithelial thickness (CCET), and the two groups were compared. Results A statistically significant difference was determined between the two groups at postoperative 1 month in respect of TBUT, Schirmer test, CFS score, and CCET (p < 0.01). In Group A, a statistically significant increase was determined in the TBUT and Schirmer values at 1 month postoperatively (p < 0.01, p = 0.01, respectively) and in Group B, these values were decreased compared to preoperatively (p < 0.01). The CCET was determined to be significantly thinner in Group B 1 month postoperatively (p < 0.01). A significant increase in CCT was observed in both groups at postoperative 1 week (p < 0.01) and preoperative values were reached at 1 month postoperatively. Conclusion In the patient group using sodium hyaluronate, significant differences were determined in all dry eye parameters and CCET. The use of hyaluronate sodium drops after cataract surgery was seen to improve dry eye parameters and contribute to a healthy ocular surface by ensuring continuity of the corneal epithelium.
- 2 May 2024
- Published by Springer Nature in International Ophthalmology
- Vol. 44 (1) , 1-7
The effect of sodium hyaluronate on dry eye and corneal epithelial thickness following cataract surgery
Purpose To evaluate the effects of sodium hyaluronate drops on dry eye parameters and corneal epithelial thickness following cataract surgery. Methods The study included 84 patients who underwent uncomplicated phacoemulsification. In Group A, 0.15% sodium hyaluronate drops were added to the postoperative antibiotic/anti-inflammatory treatment. In Group B, only antibiotic/anti-inflammatory treatment was applied. Preoperatively and at 1 week and 1 month postoperatively, all the patients were evaluated in respect of tear break-up time (TBUT), the Schirmer test under anesthesia, the corneal fluorescein staining (CFS) score, mean central corneal thickness (CCT) and mean central corneal epithelial thickness (CCET), and the two groups were compared. Results A statistically significant difference was determined between the two groups at postoperative 1 month in respect of TBUT, Schirmer test, CFS score, and CCET (p < 0.01). In Group A, a statistically significant increase was determined in the TBUT and Schirmer values at 1 month postoperatively (p < 0.01, p = 0.01, respectively) and in Group B, these values were decreased compared to preoperatively (p < 0.01). The CCET was determined to be significantly thinner in Group B 1 month postoperatively (p < 0.01). A significant increase in CCT was observed in both groups at postoperative 1 week (p < 0.01) and preoperative values were reached at 1 month postoperatively. Conclusion In the patient group using sodium hyaluronate, significant differences were determined in all dry eye parameters and CCET. The use of hyaluronate sodium drops after cataract surgery was seen to improve dry eye parameters and contribute to a healthy ocular surface by ensuring continuity of the corneal epithelium.
- Open Access
- 29 January 2024
- Published by MediHealth Academy in Archives of Ophthalmological Research
- Vol. 1 (1) , 16-17
Miller-Fisher syndrome after COVID-19 pneumonia: case report
Miller-Fisher syndrome is a rare variant of Guillain-Barré syndrome and is a disease with ataxia, areflexia, and ophthalmoplegia. Recent viral diseases are generally blamed for their etiologies. Our aim is to present a case of Miller-Fisher syndrome that developed after COVID-19 pneumonia.
- 29 January 2024
- Published by MediHealth Academy in Archives of Ophthalmological Research
- Vol. 1 (1) , 16-17
Open AccessMiller-Fisher syndrome after COVID-19 pneumonia: case report
Miller-Fisher syndrome is a rare variant of Guillain-Barré syndrome and is a disease with ataxia, areflexia, and ophthalmoplegia. Recent viral diseases are generally blamed for their etiologies. Our aim is to present a case of Miller-Fisher syndrome that developed after COVID-19 pneumonia.
- Open Access
- 1 January 2024
- Published by Turkiye Klinikleri in Turkiye Klinikleri Journal of Ophthalmology
- Vol. 33 (1) , 56-61
Evaluation of Dry Eye, Central Corneal and Epithelial Thickness in Psoriasis Patients: Case Control Study
Objective: The evaluation of diagnostic tests for dry eye, mean central corneal and epithelial thickness in psoriasis patients. Material and Methods: The study was conducted with 54 psoriasis patients and 51 healthy volunteers. Psoriasis was diagnosed based on clinical findings and histopathological examination of the lesions. Those who were under systemic psoriasis treatment drugs were not included in the study. In all cases, in addition to routine eye examination, tear break-up time (TBUT), anesthetized Schirmer test, and corneal fluorescein staining (CFS) score were evaluated. In addition, mean central corneal thickness (CCT), and mean central corneal epithelial thickness (CCET) were measured by anterior segment optical coherence tomography. The study parameters were compared between the two groups. Results: The mean age of the psoriasis patients was 37.85 ± 12.44 and the mean age of the control group was 39.96 ± 10.11. (p=0.331) There were 36 (66.7%) women and 18 (33.3%) men in the psoriasis group, and 36 (70.6%) women and 15 (29.4%) men in the control group. (p=0.665) TBUT was significantly lower (p=0.011) and CFS score was significantly higher (p=0.039) in the psoriasis group when compared to the control group. No significant difference was determined in the Schirmer test, CCT, and CCET between the two groups (p>0.05). Conclusion: Dry eye is common in psoriasis, especially due to impaired tear stability, and could negatively affect the ocular surface. In our study, no statistically significant changes were detected in CCT and CCET. However, we think that new studies are needed on this subject.
- 1 January 2024
- Published by Turkiye Klinikleri in Turkiye Klinikleri Journal of Ophthalmology
- Vol. 33 (1) , 56-61
Open AccessEvaluation of Dry Eye, Central Corneal and Epithelial Thickness in Psoriasis Patients: Case Control Study
Objective: The evaluation of diagnostic tests for dry eye, mean central corneal and epithelial thickness in psoriasis patients. Material and Methods: The study was conducted with 54 psoriasis patients and 51 healthy volunteers. Psoriasis was diagnosed based on clinical findings and histopathological examination of the lesions. Those who were under systemic psoriasis treatment drugs were not included in the study. In all cases, in addition to routine eye examination, tear break-up time (TBUT), anesthetized Schirmer test, and corneal fluorescein staining (CFS) score were evaluated. In addition, mean central corneal thickness (CCT), and mean central corneal epithelial thickness (CCET) were measured by anterior segment optical coherence tomography. The study parameters were compared between the two groups. Results: The mean age of the psoriasis patients was 37.85 ± 12.44 and the mean age of the control group was 39.96 ± 10.11. (p=0.331) There were 36 (66.7%) women and 18 (33.3%) men in the psoriasis group, and 36 (70.6%) women and 15 (29.4%) men in the control group. (p=0.665) TBUT was significantly lower (p=0.011) and CFS score was significantly higher (p=0.039) in the psoriasis group when compared to the control group. No significant difference was determined in the Schirmer test, CCT, and CCET between the two groups (p>0.05). Conclusion: Dry eye is common in psoriasis, especially due to impaired tear stability, and could negatively affect the ocular surface. In our study, no statistically significant changes were detected in CCT and CCET. However, we think that new studies are needed on this subject.
- 1 November 2023
- Published by Elsevier in Photodiagnosis and Photodynamic Therapy
- Vol. 44, 103873
Evaluation of the retinal nerve fiber layer and choroidal thickness with optical coherence tomography in patients with psoriasis
- 1 November 2023
- Published by Elsevier in Photodiagnosis and Photodynamic Therapy
- Vol. 44, 103873
Evaluation of the retinal nerve fiber layer and choroidal thickness with optical coherence tomography in patients with psoriasis
- 4 July 2022
- Published by Springer Nature in International Ophthalmology
- Vol. 43 (1) , 27-33
Evaluation of central corneal epithelial thickness with anterior segment OCT in patients with type 2 diabetes mellitus
Background This study aimed to evaluate the central corneal thickness (CCT) and central corneal epithelial thickness (CCET) in patients with Type 2 diabetes mellitus (DM), and the effect of the duration of diabetes, the degree of diabetic retinopathy (DR), and HbA1c level. Methods CCT and CCET values of 72 patients diagnosed with type 2 DM and 72 healthy individuals were measured by anterior segment optical coherence tomography. The eye tear function was evaluated with the Tear Break-up Time test (TBUT) and the Schirmer test. From the results of fundus examination, the diabetic patients were grouped as those without DR, non-proliferative DR, and proliferative DR. The disease duration and the HbA1c levels were recorded. Results In the diabetic patients, the mean CCT was determined to be thicker (p = 0.025), the CCET was thinner (p = 0.003), and the TBUT and Schirmer values were lower (p p 0.05). The presence of retinopathy had no significant effect on CCT, TBUT, and Schirmer values. The CCET was determined to be thinner in patients with retinopathy (p <0.001). Conclusions As the corneal epithelial thickness is reduced in patients with advanced diabetic retinopathy, corneal epithelial pathologies can be seen more often. Therefore, early and effective treatment can be started by taking into consideration the complications which may develop associated with the corneal epithelium following surgical procedures, especially those applied to the cornea.
- 4 July 2022
- Published by Springer Nature in International Ophthalmology
- Vol. 43 (1) , 27-33
Evaluation of central corneal epithelial thickness with anterior segment OCT in patients with type 2 diabetes mellitus
Background This study aimed to evaluate the central corneal thickness (CCT) and central corneal epithelial thickness (CCET) in patients with Type 2 diabetes mellitus (DM), and the effect of the duration of diabetes, the degree of diabetic retinopathy (DR), and HbA1c level. Methods CCT and CCET values of 72 patients diagnosed with type 2 DM and 72 healthy individuals were measured by anterior segment optical coherence tomography. The eye tear function was evaluated with the Tear Break-up Time test (TBUT) and the Schirmer test. From the results of fundus examination, the diabetic patients were grouped as those without DR, non-proliferative DR, and proliferative DR. The disease duration and the HbA1c levels were recorded. Results In the diabetic patients, the mean CCT was determined to be thicker (p = 0.025), the CCET was thinner (p = 0.003), and the TBUT and Schirmer values were lower (p p 0.05). The presence of retinopathy had no significant effect on CCT, TBUT, and Schirmer values. The CCET was determined to be thinner in patients with retinopathy (p <0.001). Conclusions As the corneal epithelial thickness is reduced in patients with advanced diabetic retinopathy, corneal epithelial pathologies can be seen more often. Therefore, early and effective treatment can be started by taking into consideration the complications which may develop associated with the corneal epithelium following surgical procedures, especially those applied to the cornea.