D�z el radyografilerinden el k�r�klar�n�n tespiti i�in derin ��renme y�ntemlerinin kullan�lmas�

�reten, Kemal; Sevin�, H�seyin Fatih; ��deli, Ufuk; Onay, Asl�han; Mara�, Y�ksel

Cilt : 30 Say� : 12 Y�l : 2024

30/12Gelecek Say� Ar�iv Kapaklar Pop�ler Makaleler

ICMJE COI Form

H�zl� Arama

D�z el radyografilerinden el k�r�klar�n�n tespiti i�in derin ��renme y�ntemlerinin kullan�lmas� [Ulus Travma Acil Cerrahi Derg]

Ulus Travma Acil Cerrahi Derg. 2022; 28(2): 196-201 | DOI: 10.14744/tjtes.2020.06944

D�z el radyografilerinden el k�r�klar�n�n tespiti i�in derin ��renme y�ntemlerinin kullan�lmas�

Kemal �reten¹, H�seyin Fatih Sevin�², Ufuk ��deli³, Asl�han Onay⁴, Y�ksel Mara�⁵
¹K�r�kkale �niversitesi T�p Fak�ltesi, Romatoloji Bilim Dal�, K�r�kkale
²Nev�ehir Devlet Hastanesi, Ortopedi ve Travmatoloji Klini�i, Nev�ehir
³K�r�kkale �niversitesi T�p Fak�ltesi, �� Hastal�klar� Klini�i, K�r�kkale
⁴Ba�kent �niversitesi T�p Fak�ltesi, Radyoloji Anabilim Dal�, Ankara
⁵Ankara �ehir Hastanesi, Romatoloji B�l�m�, Ankara

AMA�: El travmas� olan hastalar genellikle hastanelerin acil servislerinde muayene edilir. El travmalar�nda el kemik k�r�klar� s�kl�kla g�r�l�r. Bu �al��mada, derin ��renme y�ntemlerini kullanarak el k�r�klar�n�n tan�s�nda hekimlere yard�mc� olmak i�in bilgisayar destekli bir y�ntem geli�tirmeyi hedefledik.
GERE� VE Y�NTEM: Bu �al��mada, konvol�syonel sinir a�lar� kullan�lm�� ve ��renme transferi y�ntemi uygulanm��t�r. Veri k�mesinde 275 el bilek k�r��, 257 falanks k�r�� ve 270 normal el radyografisi vard�. Bu �al��mada derin ��renme y�ntemi olan konvol�syonel sinir a�lar� kullan�lm��t�r. Modelin performans�n� art�rmak i�in �nceden e�itilmi� VGG-16, GoogLeNet ve ResNet-50 a�lar� ile ��renme transferi uygulanm��t�r.
BULGULAR: Grup 1 (el bilek k�r�� ve normal el) veri setindeki do�ruluk, duyarl�l�k, �zg�ll�k ve kesinlik sonu�lar� VGG-16 ile s�ras�yla %93.3, %96.8, %90.3 ve %89.7, ResNet-50 ile s�ras�yla %88.9, %94.9, %84.2 ve %82.4 ve GoogLeNet ile s�ras�yla %88.1, %90.6, %85.9 ve %85.3 idi. Grup 2 (falanks k�r�� & normal el) veri setindeki do�ruluk, duyarl�l�k, �zg�ll�k ve kesinlik sonu�lar�, VGG-16 ile s�ras�yla %84.0, %84.1, %83.8 ve %82.8, Resnet-50 ile s�ras�yla %79.4, %78.5, %80.3 ve %79.7 ve GoogLeNet ile s�ras�yla %81.7, %81.3, %82.1 ve %81.3 idi.
TARTI�MA: Derin ��renme uygulamalar�nda son teknoloji uygulamalar olan transfer ��renme, veri art�rma gibi y�ntemler uygulayarak geli�tirdi�imiz bu bilgisayar destekli tan� y�nteminde umut verici sonu�lar elde ettik. Bu bilgisayar destekli tan� y�ntemi, el radyografilerini yorumlarken, �zellikle gece vardiyalar� ve hafta sonlar� gibi uzman meslekta�lara ula�mak zor oldu�unda, k��k hastanelerin acil servislerinde �al��an hekimlere yard�mc� olabilir.

Anahtar Kelimeler: Bilgisayar destekli tan�, derin ��renme; el k�r�klar�; konvol�syonel sinir a�lar�; ��renme transferi; veri art�rma.

Use of deep learning methods for hand fracture detection from plain hand radiographs

Kemal �reten¹, H�seyin Fatih Sevin�², Ufuk ��deli³, Asl�han Onay⁴, Y�ksel Mara�⁵
¹Department of Rheumatology, K�r�kkale University Faculty of Medicine, K�r�kkale-Turkey
²Department of Orthopedics and Traumatology, Nev�ehir State Hospital, Nev�ehir-Turkey
³Department of Internal Medicine, K�r�kkale University Faculty of Medicine, K�r�kkale-Turkey
⁴Department of Radiology, Ba�kent University Faculty of Medicine, Ankara-Turkey
⁵Department of Rheumatology, Ankara City Hospital, Ankara-Turkey

BACKGROUND: Patients with hand trauma are usually examined in emergency departments of hospitals. Hand fractures are frequently observed in patients with hand trauma. Here, we aim to develop a computer-aided diagnosis (CAD) method to assist physicians in the diagnosis of hand fractures using deep learning methods.
METHODS: In this study, Convolutional Neural Networks (CNN) were used and the transfer learning method was applied. There were 275 fractured wrists, 257 fractured phalanx, and 270 normal hand radiographs in the raw dataset. CNN, a deep learning method, were used in this study. In order to increase the performance of the model, transfer learning was applied with the pre-trained VGG-16, GoogLeNet, and ResNet-50 networks.
RESULTS: The accuracy, sensitivity, specificity, and precision results in Group 1 (wrist fracture and normal hand) dataset were 93.3%, 96.8%, 90.3%, and 89.7%, respectively, with VGG-16, were 88.9%, 94.9%, 84.2%, and 82.4%, respectively, with Resnet-50, and were 88.1%, 90.6%, 85.9%, and 85.3%, respectively, with GoogLeNet. The accuracy, sensitivity, specificity, and precision results in Group 2 (phalanx fracture and normal hand) dataset were 84.0%, 84.1%, 83.8%, and 82.8%, respectively, with VGG-16, were 79.4%, 78.5%, 80.3%, and 79.7%, respectively, with Resnet-50, and were 81.7%, 81.3%, 82.1%, and 81.3%, respectively, with GoogLeNet.
CONCLUSION: We achieved promising results in this CAD method, which we developed by applying methods such as transfer learning, data augmentation, which are state-of-the-art practices in deep learning applications. This CAD method can assist physicians working in the emergency departments of small hospitals when interpreting hand radiographs, especially when it is difficult to reach qualified colleagues, such as night shifts and weekends.

Keywords: Computer-aided diagnosis, convolutional neural networks; data augmentation; deep learning; hand fractures; transfer learning.

Sorumlu Yazar: Kemal �reten, T�rkiye
Makale Dili: �ngilizce

ATIF KOPYALA

Tam Metin PDF At�f dosyas� indir RIS EndNote BibTex Medlars Procite Reference Manager Yazara e-posta g�nder Benzer makaleler PubMed Google Scholar