Deep Learning and QR Code based Automated Diagnosis of Premature Retinopathy Prematüre Retinopatisinin Derin Öǧrenme ve Kare Kod Tabanli Otomatik Teşhisi
2023 Medical Technologies Congress, TIPTEKNO 2023, Famagusta, Kıbrıs (Gkry), 10 - 12 Kasım 2023, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/tiptekno59875.2023.10359202
- Basıldığı Şehir: Famagusta
- Basıldığı Ülke: Kıbrıs (Gkry)
- Anahtar Kelimeler: Deep learning, QR code, Retinopathy of prematurity
- Sağlık Bilimleri Üniversitesi Adresli: Evet
Özet
Retinopathy of prematurity (ROP) is a condition that affects premature neonates and can cause blindness. Deep learning can assist ophthalmologists in diagnosing retinopathy of prematurity. Although medical images are needed for deep learning-based decision support systems, these images are difficult to obtain. As an alternative to medical images, patient examination information can also be used in the diagnosis of retinopathy of prematurity. This study focuses on converting patient examination information into QR codes and making deep learning models usable in detecting retinopathy of prematurity, its zone, type and stage. This approach may offer an effective alternative for detecting and grading retinopathy of prematurity in the absence of medical images. Experiments were carried out with the 10-fold cross-validation principle of Xception, EfficientNet B7 and DenseNet-201 deep learning models. The results show that the Xception model is the model with the highest performance for the detection of PR disease, detection of its zone, detection of its stage and type.