Counteracting Data Bias and Class Imbalance—Towards a Useful and Reliable Retinal Disease Recognition System
Multiple studies presented satisfactory performances for the treatment of various ocular diseases. To date, there has been no study that describes a multiclass model, medically accurate, and trained on large diverse dataset. No study has addressed a class imbalance problem in one giant dataset origi...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/13/11/1904 |
| _version_ | 1797597728583188480 |
|---|---|
| author | Adam R. Chłopowiec Konrad Karanowski Tomasz Skrzypczak Mateusz Grzesiuk Adrian B. Chłopowiec Martin Tabakov |
| author_facet | Adam R. Chłopowiec Konrad Karanowski Tomasz Skrzypczak Mateusz Grzesiuk Adrian B. Chłopowiec Martin Tabakov |
| author_sort | Adam R. Chłopowiec |
| collection | DOAJ |
| description | Multiple studies presented satisfactory performances for the treatment of various ocular diseases. To date, there has been no study that describes a multiclass model, medically accurate, and trained on large diverse dataset. No study has addressed a class imbalance problem in one giant dataset originating from multiple large diverse eye fundus image collections. To ensure a real-life clinical environment and mitigate the problem of biased medical image data, 22 publicly available datasets were merged. To secure medical validity only Diabetic Retinopathy (DR), Age-Related Macular Degeneration (AMD) and Glaucoma (GL) were included. The state-of-the-art models ConvNext, RegNet and ResNet were utilized. In the resulting dataset, there were 86,415 normal, 3787 GL, 632 AMD and 34,379 DR fundus images. ConvNextTiny achieved the best results in terms of recognizing most of the examined eye diseases with the most metrics. The overall accuracy was 80.46 ± 1.48. Specific accuracy values were: 80.01 ± 1.10 for normal eye fundus, 97.20 ± 0.66 for GL, 98.14 ± 0.31 for AMD, 80.66 ± 1.27 for DR. A suitable screening model for the most prevalent retinal diseases in ageing societies was designed. The model was developed on a diverse, combined large dataset which made the obtained results less biased and more generalizable. |
| first_indexed | 2024-03-11T03:09:35Z |
| format | Article |
| id | doaj.art-709b76efe60f471db832ff68b50137ab |
| institution | Directory Open Access Journal |
| issn | 2075-4418 |
| language | English |
| last_indexed | 2024-03-11T03:09:35Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj.art-709b76efe60f471db832ff68b50137ab2023-11-18T07:42:41ZengMDPI AGDiagnostics2075-44182023-05-011311190410.3390/diagnostics13111904Counteracting Data Bias and Class Imbalance—Towards a Useful and Reliable Retinal Disease Recognition SystemAdam R. Chłopowiec0Konrad Karanowski1Tomasz Skrzypczak2Mateusz Grzesiuk3Adrian B. Chłopowiec4Martin Tabakov5Department of Artificial Intelligence, Wroclaw University of Science and Technology, Wybrzeże Wyspianskiego 27, 50-370 Wroclaw, PolandDepartment of Artificial Intelligence, Wroclaw University of Science and Technology, Wybrzeże Wyspianskiego 27, 50-370 Wroclaw, PolandFaculty of Medicine, Wroclaw Medical University, Wybrzeże Ludwika Pasteura 1, 50-367 Wroclaw, PolandDepartment of Artificial Intelligence, Wroclaw University of Science and Technology, Wybrzeże Wyspianskiego 27, 50-370 Wroclaw, PolandDepartment of Artificial Intelligence, Wroclaw University of Science and Technology, Wybrzeże Wyspianskiego 27, 50-370 Wroclaw, PolandDepartment of Artificial Intelligence, Wroclaw University of Science and Technology, Wybrzeże Wyspianskiego 27, 50-370 Wroclaw, PolandMultiple studies presented satisfactory performances for the treatment of various ocular diseases. To date, there has been no study that describes a multiclass model, medically accurate, and trained on large diverse dataset. No study has addressed a class imbalance problem in one giant dataset originating from multiple large diverse eye fundus image collections. To ensure a real-life clinical environment and mitigate the problem of biased medical image data, 22 publicly available datasets were merged. To secure medical validity only Diabetic Retinopathy (DR), Age-Related Macular Degeneration (AMD) and Glaucoma (GL) were included. The state-of-the-art models ConvNext, RegNet and ResNet were utilized. In the resulting dataset, there were 86,415 normal, 3787 GL, 632 AMD and 34,379 DR fundus images. ConvNextTiny achieved the best results in terms of recognizing most of the examined eye diseases with the most metrics. The overall accuracy was 80.46 ± 1.48. Specific accuracy values were: 80.01 ± 1.10 for normal eye fundus, 97.20 ± 0.66 for GL, 98.14 ± 0.31 for AMD, 80.66 ± 1.27 for DR. A suitable screening model for the most prevalent retinal diseases in ageing societies was designed. The model was developed on a diverse, combined large dataset which made the obtained results less biased and more generalizable.https://www.mdpi.com/2075-4418/13/11/1904deep learningmedical image classificationconvolutional neural networks |
| spellingShingle | Adam R. Chłopowiec Konrad Karanowski Tomasz Skrzypczak Mateusz Grzesiuk Adrian B. Chłopowiec Martin Tabakov Counteracting Data Bias and Class Imbalance—Towards a Useful and Reliable Retinal Disease Recognition System Diagnostics deep learning medical image classification convolutional neural networks |
| title | Counteracting Data Bias and Class Imbalance—Towards a Useful and Reliable Retinal Disease Recognition System |
| title_full | Counteracting Data Bias and Class Imbalance—Towards a Useful and Reliable Retinal Disease Recognition System |
| title_fullStr | Counteracting Data Bias and Class Imbalance—Towards a Useful and Reliable Retinal Disease Recognition System |
| title_full_unstemmed | Counteracting Data Bias and Class Imbalance—Towards a Useful and Reliable Retinal Disease Recognition System |
| title_short | Counteracting Data Bias and Class Imbalance—Towards a Useful and Reliable Retinal Disease Recognition System |
| title_sort | counteracting data bias and class imbalance towards a useful and reliable retinal disease recognition system |
| topic | deep learning medical image classification convolutional neural networks |
| url | https://www.mdpi.com/2075-4418/13/11/1904 |
| work_keys_str_mv | AT adamrchłopowiec counteractingdatabiasandclassimbalancetowardsausefulandreliableretinaldiseaserecognitionsystem AT konradkaranowski counteractingdatabiasandclassimbalancetowardsausefulandreliableretinaldiseaserecognitionsystem AT tomaszskrzypczak counteractingdatabiasandclassimbalancetowardsausefulandreliableretinaldiseaserecognitionsystem AT mateuszgrzesiuk counteractingdatabiasandclassimbalancetowardsausefulandreliableretinaldiseaserecognitionsystem AT adrianbchłopowiec counteractingdatabiasandclassimbalancetowardsausefulandreliableretinaldiseaserecognitionsystem AT martintabakov counteractingdatabiasandclassimbalancetowardsausefulandreliableretinaldiseaserecognitionsystem |