Kocaeli University Research Information System
Not a member yet
80615 research outputs found
Sort by
Pediatrik Hastalarda Pnömotoraks ve Plevral Efüzyon Yönetimi İçin Yapay Zekâ Destekli Karar Sistemi
No full textAmaç: Bu çalışmanın amacı, pediatrik hastalarda pnömotoraks ve plevral efüzyonun tespiti ve tüp torakostomi gerekliliğinin belirlenmesi için derin öğrenme temelli bir yapay zekâ destekli karar sisteminin geliştirilmesi ve etkinliğinin değerlendirilmesidir.Yöntem: Retrospektif olarak, 2005-2022 yılları arasında tedavi gören 170 pediatrik hastaya ait toplam 510 akciğer grafisi analiz edildi. Hastalar tüp torakostomi gerektiren ve konservatif tedaviyle yönetilen olmak üzere iki gruba ayrıldı. Derin öğrenme modeli, uzman cerrahlar tarafından işaretlenmiş görüntülerle eğitildi. Modelin performansı ortalama hassasiyet (mean Average Precision - mAP), duyarlılık, özgüllük, doğruluk, pozitif ve negatif öngörü değerleri ve F1 skoru gibi çeşitli metrikler kullanılarak değerlendirildi.Bulgular: Derin öğrenme modeli, müdahale gerektiren pnömotoraks ve plevral efüzyon alanlarının tespitinde 0.918 gibi yüksek bir mAP değeri elde etti. Bağımsız test veri setinde ise modelin duyarlılığı %64, özgüllüğü ise %96,15 olarak belirlendi. Pozitif prediktif değeri %94,12, negatif prediktif değeri %73,53 ve genel doğruluk oranı %80,39 olarak hesaplandı.Sonuç: Bu çalışma, pediatrik pnömotoraks ve plevral efüzyon tanısında ve tedavi kararlarının desteklenmesinde yapay zekâ tabanlı derin öğrenme sistemlerinin etkili olduğunu ortaya koymaktadır. Geliştirilen modelin klinik uygulamalara entegrasyonu öncesinde daha fazla doğrulama ve iyileştirmeye ihtiyaç duyulmakla birlikte, mevcut sonuçlar bu tür sistemlerin sağlık hizmetlerinde kullanımını güçlü biçimde desteklemektedir. Gelecekteki çalışmalar hibrit görüntüleme modelleri üzerine yoğunlaşmalı, yorumlanabilirliği artırmalı, veri kalitesi sorunlarını çözmeli ve düzenleyici standartlara uyumu sağlamalıdır.Anahtar Kelimeler: karar destek sistemleri, derin öğrenme, tanısal görüntüleme, plevral efüzyon, pnömotoraksAim: The aim of this study is to develop and evaluate a deep learning-based artificial intelligence decision support system capable of accurately detecting pneumothorax and pleural effusion in pediatric patients and determining the necessity of tube thoracostomy.Methods: In our retrospective diagnostic accuracy study, we analyzed a total of 510 chest X-ray images from 170 pediatric patients treated between 2005 and 2022. Patients were categorized into two groups: those requiring tube thoracostomy (Tube group) and those managed conservatively (Conservative group). The deep learning model was trained using images annotated by expert surgeons. Performance metrics such as mean Average Precision (mAP), sensitivity, specificity, accuracy, positive and negative predictive values, and F1 score were used to evaluate the model's effectiveness.Results: The deep learning model achieved a high mAP score of 0.918 in detecting regions requiring intervention for pneumothorax and pleural effusion. In the independent test data set, the model demonstrated a sensitivity of 64.00% and a specificity of 96.15%. Positive predictive value was calculated as 94.12%, negative predictive value as 73.53%, and overall accuracy as 80.39%.Conclusions: This study demonstrates that artificial intelligence-based deep learning systems effectively support diagnostic and treatment decisions in pediatric pneumothorax and pleural effusion. Although further validation and refinement are necessary before clinical integration, current results strongly support the adoption of such systems in healthcare. Future research should focus on developing hybrid imaging models, enhancing interpretability, resolving data quality issues, and ensuring compliance with regulatory standards.Keywords: decision support systems, deep learning, diagnostic imaging, pleural effusion, pneumothorax</p
Real-Time Deep-Learning-Based Recognition of Helmet-Wearing Personnel on Construction Sites from a Distance
No full textOn construction sites, it is crucial and and in most cases mandatory to wear safety equipment such as helmets, safety shoes, vests, and belts. The most important of these is the helmet, as it protects against head injuries and can also serve as a marker for detecting and tracking workers, since a helmet is typically visible to cameras on construction sites. Checking helmet usage, however, is a labor-intensive and time-consuming process. A lot of work has been conducted on detecting and tracking people. Some studies have involved hardware-based systems that require batteries and are often perceived as intrusive by workers, while others have focused on vision-based methods. The aim of this work is not only to detect workers and helmets, but also to identify workers through labeled helmets using symbol detection methods. Person and helmet detection tasks were handled by training existing datasets and gained accurate results. For symbol detection, 14 different shapes were selected and put on helmets in a triple format side by side. A total of 11,243 images have been annotated. YOLOv5 and YOLOv8 were used to train the dataset and obtain models. The results show that both methods achieved high precision and recall. However, YOLOv5 slightly outperformed YOLOv8 in real-time identification tests, correctly detecting the helmet symbols. A testing dataset containing different distances was generated in order to measure accuracy by distance. According to the results, accurate identification was achieved at distances of up to 10 meters. Also, a location-based symbol-ordering algorithm is proposed. Since symbol detection does not follow any order and works with confidence values in the inference mode, a left to right approach is followed
Improvement of Urban Life and Environmental Quality in the Reuse of Historical Industrial Area: Izmit Çuha Factory
No full textAristotelian Equity and Discretion Thesis in Modern Legal Philosophy
No full textThis essay explores the Aristotelian concept of equity and the discretionary power of judges within the context of modern legal philosophy. Modern jurisprudence has changed profoundly as a result of the debates between Hart and Dworkin on judicial discretion and the principles of equity. The discretion thesis asserts that judges possess both the authority to fill gaps in law and to incorporate equitable principles when interpreting the law. By contrast, Aristotle's account of equity comprises two main dimensions: first, as a moral virtue; second, as a form of legal justice. Aristotle emphasises that applying equity requires careful consideration of case-specific irregularities. This essay argues that Hart and Dworkin's discretion thesis does not fully account for Aristotle's concept of equity due to insufficient emphasis on particularity in each case. Conversely, I propose that discretionary powers within contemporary codified law systems impose a continuing legal obligation upon judges, enabling them to more effectively balance discretion with equitable considerations
