A Novel Multi-Head Attention Framework for COVID-19 Detection: Hybrid Integration of MobileNet and VGG19 with Enhanced Feature Learning

Şafak Kılıç

Makale detayları

A Novel Multi-Head Attention Framework for COVID-19 Detection: Hybrid Integration of MobileNet and VGG19 with Enhanced Feature Learning

Yazarlar:

Şafak Kılıç

Yayıncı:

Cukurova University

Ana Başlıklar:

Bilgisayar Görüşü | Görüntü İşleme

Dergi:

Çukurova Üniversitesi Mühendislik Fakültesi Dergisi Vol.40 Issue 3, pp.655-670

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001358969

pISSN:

2757-9255

Anahtar kelimeler:

Özet:

 The COVID-19 pandemic has underscored the urgent need for rapid, accurate, and affordable diagnostic tools to complement RT-PCR testing. This study proposes a novel multi-head attention framework that integrates VGG19 and MobileNet for automated COVID-19 detection from chest X-rays. The model employs a hybrid mechanism combining spatial, channel, and self-attention components, enhancing feature representation while preserving efficiency. Evaluations on 7,132 chest X-ray images across four categories COVID-19, Normal, Pneumonia, Tuberculosis demonstrated outstanding performance: 99.0 accuracy, 99.0 macro and weighted F1-scores, with near-perfect class-specific results 100 Tuberculosis, 99.7 COVID-19, 99.5 Normal, 96.0 Pneumonia. Inference time was only 63 ms per image, with a compact 14.8 MB model size. These results surpass baseline MobileNet and DenseNet121 by 2.63 and 4.32, respectively. The proposed framework offers reliable rapid screening and differential diagnosis, supported by interpretable attention maps, making it highly suitable for deployment in resource-limited healthcare and point-of-care settings.

 The COVID-19 pandemic has underscored the urgent need for rapid, accurate, and affordable diagnostic tools to complement RT-PCR testing. This study proposes a novel multi-head attention framework that integrates VGG19 and MobileNet for automated COVID-19 detection from chest X-rays. The model employs a hybrid mechanism combining spatial, channel, and self-attention components, enhancing feature representation while preserving efficiency. Evaluations on 7,132 chest X-ray images across four categories COVID-19, Normal, Pneumonia, Tuberculosis demonstrated outstanding performance: 99.0 accuracy, 99.0 macro and weighted F1-scores, with near-perfect class-specific results 100 Tuberculosis, 99.7 COVID-19, 99.5 Normal, 96.0 Pneumonia. Inference time was only 63 ms per image, with a compact 14.8 MB model size. These results surpass baseline MobileNet and DenseNet121 by 2.63 and 4.32, respectively. The proposed framework offers reliable rapid screening and differential diagnosis, supported by interpretable attention maps, making it highly suitable for deployment in resource-limited healthcare and point-of-care settings.

 The COVID-19 pandemic has underscored the urgent need for rapid, accurate, and affordable diagnostic tools to complement RT-PCR testing. This study proposes a novel multi-head attention framework that integrates VGG19 and MobileNet for automated COVID-19 detection from chest X-rays. The model employs a hybrid mechanism combining spatial, channel, and self-attention components, enhancing feature representation while preserving efficiency. Evaluations on 7,132 chest X-ray images across four categories COVID-19, Normal, Pneumonia, Tuberculosis demonstrated outstanding performance: 99.0 accuracy, 99.0 macro and weighted F1-scores, with near-perfect class-specific results 100 Tuberculosis, 99.7 COVID-19, 99.5 Normal, 96.0 Pneumonia. Inference time was only 63 ms per image, with a compact 14.8 MB model size. These results surpass baseline MobileNet and DenseNet121 by 2.63 and 4.32, respectively. The proposed framework offers reliable rapid screening and differential diagnosis, supported by interpretable attention maps, making it highly suitable for deployment in resource-limited healthcare and point-of-care settings.

 COVID-19 pandemisi, RT-PCR testlerini destekleyecek hızlı, doğru ve maliyet etkin tanı araçlarına duyulan ihtiyacı ortaya koymuştur. Bu çalışmada, göğüs röntgeni görüntülerinden otomatik COVID-19 tespiti için VGG19 ve MobileNet mimarilerini entegre eden yeni bir çok başlı dikkat çerçevesi önerilmektedir. Model, uzamsal, kanal ve çok başlı öz-dikkat mekanizmalarını birleştirerek özellik çıkarımını güçlendirirken hesaplama verimliliğini korumaktadır. Yaklaşımımız 7.132 görüntüden oluşan dört sınıflı veri kümesinde test edilmiştir COVID-19, Normal, Pnömoni, Tüberküloz. Dikkat mekanizmasıyla geliştirilmiş MobileNet 99,0 doğruluk, makro ve ağırlıklı F1 skorları elde etmiştir. Sınıf bazında 100 Tüberküloz, 99,7 COVID-19, 99,5 Normal ve 96,0 Pnömoni doğruluğu kaydedilmiştir. Ayrıca model, 14,8 MB boyutu ve 63 ms çıkarım süresi ile klinik uygulanabilirliğe sahiptir. Sonuçlar, mevcut yöntemlere göre 2,634,32 iyileşme göstermekte olup, modelin güvenilir hızlı tarama ve ayırıcı tanıda etkili olduğunu göstermektedir.

Makale detayları

A Novel Multi-Head Attention Framework for COVID-19 Detection: Hybrid Integration of MobileNet and VGG19 with Enhanced Feature Learning

Yazarlar:

Şafak Kılıç

Yayıncı:

Cukurova University

Ana Başlıklar:

Bilgisayar Görüşü | Görüntü İşleme

Dergi:

Çukurova Üniversitesi Mühendislik Fakültesi Dergisi Vol.40 Issue 3, pp.655-670

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001358969

pISSN:

2757-9255

COVID-19 Detection

Deep Learning

Multi-Head Attention Mechanism

Medical Image Analysis

VGG19COVID-19 Tespiti

Derin Öğrenme

Çok Başlı Dikkat Mekanizması

Tıbbi Görüntü İşleme

VGG19

Sayı

Makale detayları

Yazarlar: Şafak Kılıç

Yayıncı: Cukurova University

Ana Başlıklar: Bilgisayar Görüşü | Görüntü İşleme

Dergi: Çukurova Üniversitesi Mühendislik Fakültesi Dergisi Vol.40 Issue 3, pp.655-670

Belge Tipi: Makale

Yayın Tarihi: 9/2025

Belge Dili: İngilizce

Document ID (Org No.): 2001358969

pISSN: 2757-9255

Sayı

Yazarlar:

Şafak Kılıç

Yayıncı:

Cukurova University

Ana Başlıklar:

Bilgisayar Görüşü | Görüntü İşleme

Dergi:

Çukurova Üniversitesi Mühendislik Fakültesi Dergisi Vol.40 Issue 3, pp.655-670

Belge Tipi:

Makale

Yayın Tarihi:

9/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001358969

pISSN:

2757-9255