The Metrisor device has been developed using gas sensors for rapid, highly accurate and effective diagnosis of metritis. 513 cattle uteri were collected from abattoirs and swabs were taken for microbiological testing. The Metrisor device was used to measure intrauterine gases. The results showed a bacterial growth rate of 75.75 in uteri with clinical metritis. In uteri positive for clinical metritis, the most commonly isolated and identified bacteria were Trueperella pyogenes, Fusobacterium necrophorum and Escherichia coli. Measurements taken with Metrisor to determine the presence of metritis in the uterus yielded the most successful results in evaluations of relevant machine learning algorithms. The ICO (Iterative Classifier Optimizer) algorithm achieved 71.22 accuracy, 64.40 precision and 71.20 recall. Experiments were conducted to examine bacterial growth in the uterus and the random forest algorithm produced the most successful results with accuracy, precision and recall values of 78.16 , 75.30 and 78.20 respectively. ICO also showed high performance in experiments to determine bacterial growth in metritis-positive uteri, with accuracy, precision and recall values of 78.97 , 77.20 and 79.00 , respectively. In conclusion, the Metrisor device demonstrated high accuracy in detecting metritis and bacterial growth in uteri and could identify bacteria such as E. coli, S. aureus, coagulase-negative staphylococci, T. pyogenes, Bacillus spp., Clostridium spp. and F. necrophorum with rates up to 80 . It provides a reliable, rapid and effective means of detecting metritis in animals in the field without the need for laboratory facilities.
Keyword: cow; machine learning; metrisor; metritis; sensor
Yayın Adı (dc.title) | Metrisor: A novel diagnostic method for metritis detection in cattle based on machine learning and sensors |
Yazar/lar (dc.contributor.yazarlar) | Ali Risvanli, Burak Tanyeri, Güngör Yildirim, Yetkin Tatar, Mehmet Gedikpinar, Hakan Kalender, Tarik Safak, Burak Yuksel, Burcu Karagulle, Oznur Yilmaz, Mehmet Akif Kilinc |
Yayın Türü (dc.type) | Makale |
Dil (dc.language) | İngilizce |
Yayımlanma Yılı (dc.date.issued) | 2024 |
Ulusal/Uluslararası (dc.identifier.ulusaluluslararasi) | Uluslararası |
Kaynak (dc.relation.journal) | Theriogenology |
Cilt/Sayı (dc.identifier.volume) | 223 |
Sayfa (dc.identifier.startpage) | 115-121 |
ISSN/ISBN (dc.identifier.issn) | ISSN: 0093-691X; Online ISSN: 1879-3231 |
Yayıncı (dc.publisher) | Elsevier |
Veri Tabanları (dc.contributor.veritaban) | Web of Science Core Collection |
Veri Tabanları (dc.contributor.veritaban) | Sciencedirect |
Veri Tabanları (dc.contributor.veritaban) | Scopus |
İndex Türü (dc.identifier.index) | SCI Expanded |
İndex Türü (dc.identifier.index) | Scopus |
Etki Faktörü (dc.identifier.etkifaktoru) | 2,4 / 2023-WOS / Son 5 yıl: 2,5 |
Özet (dc.description.abstract) | The Metrisor device has been developed using gas sensors for rapid, highly accurate and effective diagnosis of metritis. 513 cattle uteri were collected from abattoirs and swabs were taken for microbiological testing. The Metrisor device was used to measure intrauterine gases. The results showed a bacterial growth rate of 75.75 in uteri with clinical metritis. In uteri positive for clinical metritis, the most commonly isolated and identified bacteria were Trueperella pyogenes, Fusobacterium necrophorum and Escherichia coli. Measurements taken with Metrisor to determine the presence of metritis in the uterus yielded the most successful results in evaluations of relevant machine learning algorithms. The ICO (Iterative Classifier Optimizer) algorithm achieved 71.22 accuracy, 64.40 precision and 71.20 recall. Experiments were conducted to examine bacterial growth in the uterus and the random forest algorithm produced the most successful results with accuracy, precision and recall values of 78.16 , 75.30 and 78.20 respectively. ICO also showed high performance in experiments to determine bacterial growth in metritis-positive uteri, with accuracy, precision and recall values of 78.97 , 77.20 and 79.00 , respectively. In conclusion, the Metrisor device demonstrated high accuracy in detecting metritis and bacterial growth in uteri and could identify bacteria such as E. coli, S. aureus, coagulase-negative staphylococci, T. pyogenes, Bacillus spp., Clostridium spp. and F. necrophorum with rates up to 80 . It provides a reliable, rapid and effective means of detecting metritis in animals in the field without the need for laboratory facilities. |
Özet (dc.description.abstract) | Keyword: cow; machine learning; metrisor; metritis; sensor |
URL (dc.rights) | https://www.sciencedirect.com/science/article/abs/pii/S0093691X24001705 |
DOI (dc.identifier.doi) | 10.1016/j.theriogenology.2024.05.002 |
Fakültesi / Enstitütü (dc.identifier.fakulte) | Veteriner Fakültesi |
Bölümü (dc.identifier.bolum) | Veteriner |
Kurumdaki Yazar/lar (dc.contributor.author) | Ali RIŞVANLI |
Kayıt No (dc.identifier.kayitno) | BLD5F4796F |
Kayıt Giriş Tarihi (dc.date.available) | 2024-05-10 |
Not (Yayımlanma Yılı) (dc.identifier.notyayinyili) | July 2024 |
Wos No (dc.identifier.wos) | WOS:001239590100001 |
Konu Başlıkları (dc.subject) | cow |
Konu Başlıkları (dc.subject) | machine learning |
Konu Başlıkları (dc.subject) | metrisor |
Konu Başlıkları (dc.subject) | metritis |
Konu Başlıkları (dc.subject) | sensor |