Multi-Input CNN Models for Breast Cancer Detection Using BreakHis Database

Adnan Köşker; Ümit Budak; Musa Çıbuk; Abdülkadir Şengür

Makale detayları

Multi-Input CNN Models for Breast Cancer Detection Using BreakHis Database

Yazarlar:

Adnan Köşker | Ümit Budak | Musa Çıbuk | Abdülkadir Şengür

Yayıncı:

Fırat University

Ana Başlıklar:

Derin Öğrenme | Yapay Görme

Dergi:

Fırat Üniversitesi Mühendislik Bilimleri Dergisi Vol.37 Issue 2, pp.711-721

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001357257

pISSN:

1308-9072

Anahtar kelimeler:

Özet:

 Breast cancer BC is one of the diseases that women suffer most, especially in the world. Routine breast checks are vital for both early diagnosis and early treatment of the person concerned. Computer aided diagnosis systems have also come a long way in being a helpful tool for pathologists during diagnosis. In this work, a novel convolutional neural network CNN is proposed for the effective diagnosis of BC from histopathological images. Since classical CNNs have only one input, the network is to use only the raw images from the dataset in the training process. This limits the network from using an extra feature as an input. However, the proposed model has two inputs, unlike classical CNN structures. One input of the network uses histopathological raw images and the other input uses deep features of related images. All of the experimental studies were performed on the widely used BreaKHis dataset. For the test of performance, the accuracy criterion was preferred and the 5-fold cross-validation technique was taken into account. Accuracy scores of 99.94, 98.94, 99.05, and 97.30 were obtained in 40, 100, 200 and 400 sub-datasets, respectively. While the results obtained were highly effective values for the diagnosis of BC, they were also far superior to other results reported in the literature.

 Breast cancer BC is one of the diseases that women suffer most, especially in the world. Routine breast checks are vital for both early diagnosis and early treatment of the person concerned. Computer aided diagnosis systems have also come a long way in being a helpful tool for pathologists during diagnosis. In this work, a novel convolutional neural network CNN is proposed for the effective diagnosis of BC from histopathological images. Since classical CNNs have only one input, the network is to use only the raw images from the dataset in the training process. This limits the network from using an extra feature as an input. However, the proposed model has two inputs, unlike classical CNN structures. One input of the network uses histopathological raw images and the other input uses deep features of related images. All of the experimental studies were performed on the widely used BreaKHis dataset. For the test of performance, the accuracy criterion was preferred and the 5-fold cross-validation technique was taken into account. Accuracy scores of 99.94, 98.94, 99.05, and 97.30 were obtained in 40, 100, 200 and 400 sub-datasets, respectively. While the results obtained were highly effective values for the diagnosis of BC, they were also far superior to other results reported in the literature.

 Breast cancer BC is one of the diseases that women suffer most, especially in the world. Routine breast checks are vital for both early diagnosis and early treatment of the person concerned. Computer aided diagnosis systems have also come a long way in being a helpful tool for pathologists during diagnosis. In this work, a novel convolutional neural network CNN is proposed for the effective diagnosis of BC from histopathological images. Since classical CNNs have only one input, the network is to use only the raw images from the dataset in the training process. This limits the network from using an extra feature as an input. However, the proposed model has two inputs, unlike classical CNN structures. One input of the network uses histopathological raw images and the other input uses deep features of related images. All of the experimental studies were performed on the widely used BreaKHis dataset. For the test of performance, the accuracy criterion was preferred and the 5-fold cross-validation technique was taken into account. Accuracy scores of 99.94, 98.94, 99.05, and 97.30 were obtained in 40, 100, 200 and 400 sub-datasets, respectively. While the results obtained were highly effective values for the diagnosis of BC, they were also far superior to other results reported in the literature.

 Meme kanseri MK, özellikle dünya genelinde kadınların en çok muzdarip olduğu hastalıklardan biridir. Rutin meme kontrolleri, hem erken teşhis hem de ilgili kişinin erken tedavisi açısından hayati öneme sahiptir. Bilgisayar destekli teşhis sistemleri, patoloji uzmanlarına teşhis sırasında yardımcı olan bir araç olma konusunda önemli bir yol kat etmiştir. Bu çalışmada, histopatolojik görüntülerden BCnin etkili bir şekilde teşhisi için yeni bir evrişimli sinir ağı CNN önerilmiştir. Geleneksel CNNlerin yalnızca bir girişi olduğu için, eğitim sürecinde yalnızca veri kümesindeki ham görüntüler kullanılmaktadır. Bu durum, ağın ek bir özelliği giriş olarak kullanmasını sınırlandırmaktadır. Ancak, önerilen model klasik CNN yapılarından farklı olarak iki girişe sahiptir. Ağın bir girişi histopatolojik ham görüntüleri, diğer girişi ise ilgili görüntülerin derin özelliklerini kullanmaktadır. Tüm deneysel çalışmalar, yaygın olarak kullanılan BreaKHis veri kümesi üzerinde gerçekleştirilmiştir. Performans testi için doğruluk ölçütü tercih edilmiş ve 5 katlı çapraz doğrulama tekniği kullanılmıştır. 40, 100, 200 ve 400 alt veri kümelerinde sırasıyla 99,94, 98,94, 99,05 ve 97,30 doğruluk skorları elde edilmiştir. Elde edilen sonuçlar, MK teşhisi için son derece etkili değerler olmasının yanı sıra, literatürde raporlanan diğer sonuçlara kıyasla çok daha üstün niteliktedir.

Makale detayları

Multi-Input CNN Models for Breast Cancer Detection Using BreakHis Database

Yazarlar:

Adnan Köşker | Ümit Budak | Musa Çıbuk | Abdülkadir Şengür

Yayıncı:

Fırat University

Ana Başlıklar:

Derin Öğrenme | Yapay Görme

Dergi:

Fırat Üniversitesi Mühendislik Bilimleri Dergisi Vol.37 Issue 2, pp.711-721

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001357257

pISSN:

1308-9072

Multi-input CNN

breast cancer diagnosis

histopathological image

BreaKHis database.Çok girişli CNN

meme kanseri teşhisi

histopatolojik görüntü

BreaKHis veritabanı.

Sayı

Makale detayları

Yazarlar: Adnan Köşker | Ümit Budak | Musa Çıbuk | Abdülkadir Şengür

Yayıncı: Fırat University

Ana Başlıklar: Derin Öğrenme | Yapay Görme

Dergi: Fırat Üniversitesi Mühendislik Bilimleri Dergisi Vol.37 Issue 2, pp.711-721

Belge Tipi: Makale

Yayın Tarihi: 9/2025

Belge Dili: İngilizce

Document ID (Org No.): 2001357257

pISSN: 1308-9072

Sayı

Yazarlar:

Adnan Köşker | Ümit Budak | Musa Çıbuk | Abdülkadir Şengür

Yayıncı:

Fırat University

Ana Başlıklar:

Derin Öğrenme | Yapay Görme

Dergi:

Fırat Üniversitesi Mühendislik Bilimleri Dergisi Vol.37 Issue 2, pp.711-721

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001357257

pISSN:

1308-9072