Segmentation of Meningeal Contrast Enhancement in Post-Contrast T1-Weighted Images Using the Deep Learning Method

Nevin Aydın; Suzan Şaylısoy; Uğur Toprak; Burcu Mert; Özer Çelik

Makale detayları

Segmentation of Meningeal Contrast Enhancement in Post-Contrast T1-Weighted Images Using the Deep Learning Method

Yazarlar:

Nevin Aydın | Suzan Şaylısoy | Uğur Toprak | Burcu Mert | Özer Çelik

Yayıncı:

Eskişehir Osmangazi University

Ana Başlıklar:

Radyoloji ve Organ Görüntüleme

Dergi:

Osmangazi Tıp Dergisi Vol.47 Issue 4, pp.600-605

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001343014

pISSN:

1305-4953

Anahtar kelimeler:

Özet:

 To evaluate the success of segmentation of meningeal contrast enhancement on post-contrast T1-weighted images using the deep learning method. The study retrospectively included 313 sections obtained from post-contrast T1-weighted sequences of 83 patients with meningeal enhancement. The dataset was divided into three groups. A total of 300 epochs of training were performed using PyTorch U-Net, and the best model was identified. The results were calculated by selecting 50 as the threshold for the intersection over union statistics. In total, images of 83 patients were evaluated, of whom 36 43.4 were female and 47 56.6 were male. The mean  standard deviation of the patients age was 57.06  16.73 years. Of the 313 sections obtained, 251 were allocated in the training group, 31 to the validation group, and 31 to the test group. The results of the test group were as follows: 35 true positives, 12 false positives, and 12 false negatives. The precision, sensitivity, and F1 score values were all calculated to be 74. This is one of the pioneering studies in the literature on the segmentation of meningeal contrast-enhanced areas using the deep learning-based U-net architecture. Further studies are needed in this area.

 To evaluate the success of segmentation of meningeal contrast enhancement on post-contrast T1-weighted images using the deep learning method. The study retrospectively included 313 sections obtained from post-contrast T1-weighted sequences of 83 patients with meningeal enhancement. The dataset was divided into three groups. A total of 300 epochs of training were performed using PyTorch U-Net, and the best model was identified. The results were calculated by selecting 50 as the threshold for the intersection over union statistics. In total, images of 83 patients were evaluated, of whom 36 43.4 were female and 47 56.6 were male. The mean  standard deviation of the patients age was 57.06  16.73 years. Of the 313 sections obtained, 251 were allocated in the training group, 31 to the validation group, and 31 to the test group. The results of the test group were as follows: 35 true positives, 12 false positives, and 12 false negatives. The precision, sensitivity, and F1 score values were all calculated to be 74. This is one of the pioneering studies in the literature on the segmentation of meningeal contrast-enhanced areas using the deep learning-based U-net architecture. Further studies are needed in this area.

 To evaluate the success of segmentation of meningeal contrast enhancement on post-contrast T1-weighted images using the deep learning method. The study retrospectively included 313 sections obtained from post-contrast T1-weighted sequences of 83 patients with meningeal enhancement. The dataset was divided into three groups. A total of 300 epochs of training were performed using PyTorch U-Net, and the best model was identified. The results were calculated by selecting 50 as the threshold for the intersection over union statistics. In total, images of 83 patients were evaluated, of whom 36 43.4 were female and 47 56.6 were male. The mean  standard deviation of the patients age was 57.06  16.73 years. Of the 313 sections obtained, 251 were allocated in the training group, 31 to the validation group, and 31 to the test group. The results of the test group were as follows: 35 true positives, 12 false positives, and 12 false negatives. The precision, sensitivity, and F1 score values were all calculated to be 74. This is one of the pioneering studies in the literature on the segmentation of meningeal contrast-enhanced areas using the deep learning-based U-net architecture. Further studies are needed in this area.

 Çalışmamızın amacı; postkontrast T1 ağırlıklı görüntülerde meningeal kontrastlanmanın derin öğrenme yöntemi ile segmentasyonunun başarısını değerlendirmektir. Retrospektif olarak 2013-2020 yılları arasında meningeal kontrastlanması olan 83 hastanın postkontrast T1 ağırlıklı sekanslarından elde edilen 313 kesit çalışmaya dahil edildi. Veri seti train  validation- test grubu olarak ayrıldı. Pytorch Unet ile 300 epoch eğitim yapıldı, en iyi model kaydedildi. Birleşim Üzerinde Kesişim The Intersection over Union, IoU, Jaccard Endeksi istatistiğinin eşik değeri olarak 50 seçilerek sonuçlar hesaplandı. Toplamda 83 hastanın görüntüleri değerlendirilmiş olup bu hastalardan 36 43.4sı kadın, 47 56.6si erkek hasta idi. Hastaların yaş ortalaması  standart sapması 57.06 16.73 idi. 83 hastanın görüntüsünden elde edilen 313 kesitte; 251 kesit eğitime, 31 kesit validasyona, 31 kesitteki etiketler test aşamasına ayrıldı. Test grubunda Doğru Bulunan: 35, Yanlış Bulunan: 12, Bulunamayan: 12 olarak tespit edildi. Çalışmamızda Precision, Sensitivity, F1 Score değerleri sırasıyla 74, 74, 74 olarak hesaplandı. Çalışmamız derin öğrenme temelli U-net mimarisi kullanarak meningeal kontrastlanma alanlarının segmentasyonunda literatürde öncü çalışmalardan biri olup bu alanda yapılacak yeni çalışmalara ihtiyaç vardır.

Makale detayları

Segmentation of Meningeal Contrast Enhancement in Post-Contrast T1-Weighted Images Using the Deep Learning Method

Yazarlar:

Nevin Aydın | Suzan Şaylısoy | Uğur Toprak | Burcu Mert | Özer Çelik

Yayıncı:

Eskişehir Osmangazi University

Ana Başlıklar:

Radyoloji ve Organ Görüntüleme

Dergi:

Osmangazi Tıp Dergisi Vol.47 Issue 4, pp.600-605

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001343014

pISSN:

1305-4953

Deep learning method

Meningeal enhancement

Pachymeningeal enhancement

Leptomeningeal enhancement

Magnetic resonance imagingDerin Öğrenme

Meningeal Kontrastlanma,

Pakimeningeal Kontrastlanma

Leptomeningeal Kontrastlanma

Manyetik Rezonans Görüntüleme

Sayı

Makale detayları

Yazarlar: Nevin Aydın | Suzan Şaylısoy | Uğur Toprak | Burcu Mert | Özer Çelik

Yayıncı: Eskişehir Osmangazi University

Ana Başlıklar: Radyoloji ve Organ Görüntüleme

Dergi: Osmangazi Tıp Dergisi Vol.47 Issue 4, pp.600-605

Belge Tipi: Makale

Yayın Tarihi: 6/2025

Belge Dili: İngilizce

Document ID (Org No.): 2001343014

pISSN: 1305-4953

Sayı

Yazarlar:

Nevin Aydın | Suzan Şaylısoy | Uğur Toprak | Burcu Mert | Özer Çelik

Yayıncı:

Eskişehir Osmangazi University

Ana Başlıklar:

Radyoloji ve Organ Görüntüleme

Dergi:

Osmangazi Tıp Dergisi Vol.47 Issue 4, pp.600-605

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001343014

pISSN:

1305-4953