Due to the fast-paced development of technology, the software has become a crucial aspect of modern life, facilitating the operation and management of hardware devices. Nevertheless, using substandard software can result in severe complications for users, putting human lives at risk. This underscores the significance of error-free and premium-quality software. Verification and validation are essential in ensuring high-quality software development; software testing is integral to this process. Although code coverage is a prevalent method for assessing the efficacy of test suites, it has some limitations. Therefore, mutation testing is proposed as a remedy to tackle these limitations. Furthermore, mutation testing is recognized as a method for directing test case creation and evaluating the effectiveness of test suites. Our proposed method involves autonomously learning mutations from faults in real-world software applications. Firstly, our approach involves extracting bug fixes at the method-level, classifying them according to mutation types, and performing code abstraction. Subsequently, the approach utilizes a deep learning technique based on neural machine translation to develop mutation models. Our method has been trained and assessed using approximately ~588k bug fix commits extracted from GitHub. The results of our experimental assessment indicate that our models can forecast mutations resembling resolved bugs in 6% to 35% of instances. The models effectively revert fixed code to its original buggy version, reproducing the original bug and generating various other buggy codes with up to 94% accuracy. More than 96% of the generated mutants also demonstrate lexical and syntactic accuracy.
Background/Objectives: Diagnosing Rotator Cuff Tears (RCTs) via Magnetic Resonance Imaging (MRI) is clinically challenging due to complex 3D anatomy and significant interobserver variability. Traditional slice-centric Convolutional Neural Networks (CNNs) often fail to capture the necessary volumetric context for accurate grading. This study aims to develop and validate the Patient-Aware Vision Transformer (Pa-ViT), an explainable deep-learning framework designed for the automated, patient-level classification of RCTs (Normal, Partial-Thickness, and Full-Thickness). Methods: A large-scale retrospective dataset comprising 2447 T2-weighted coronal shoulder MRI examinations was utilized. The proposed Pa-ViT framework employs a Vision Transformer (ViT-Base) backbone within a Weakly-Supervised Multiple Instance Learning (MIL) paradigm to aggregate slice-level semantic features into a unified patient diagnosis. The model was trained using a weighted cross-entropy loss to address class imbalance and was benchmarked against widely used CNN architectures and traditional machine-learning classifiers. Results: The Pa-ViT model achieved a high overall accuracy of 91% and a macro-averaged F1-score of 0.91, significantly outperforming the standard VGG-16 baseline (87%). Notably, the model demonstrated superior discriminative power for the challenging Partial-Thickness Tear class (ROC AUC: 0.903). Furthermore, Attention Rollout visualizations confirmed the model’s reliance on genuine anatomical features, such as the supraspinatus footprint, rather than artifacts. Conclusions: By effectively modeling long-range dependencies, the Pa-ViT framework provides a robust alternative to traditional CNNs. It offers a clinically viable, explainable decision support tool that enhances diagnostic sensitivity, particularly for subtle partial-thickness tears.
Background/Objectives: Rotator cuff tears are a leading cause of shoulder disability. While multi-sequence MRI is standard, the optimal deep learning integration of heterogeneous image series and clinical metadata remains unresolved. This study evaluated a hierarchical, sequence-aware multimodal framework for patient-level binary rotator cuff tear classification. Methods: A single-center cohort of 199 patients (100 tears, 99 controls) was analyzed across four MRI sequences (T1 coronal, T2 fat-suppressed sagittal, and proton density [PD] fat-suppressed coronal and transverse/axial) and nine demographic features. Under a patient-level stratified three-fold cross-validation scheme preventing data leakage, we evaluated ResNet50 and Vision Transformer baselines (Study 0), full-protocol fusion topologies (Study 1), and systematically mapped sequence-subset combinations with or without metadata (Study 2). Results: In Study 0, the PD coronal ResNet50 model was the top baseline (AUC = 0.9834, F1 = 0.9515). In Study 1, late decision fusion yielded the highest AUC (0.9909), while feature concatenation optimized threshold balance (F1 = 0.9502). In Study 2, a streamlined three-sequence subset with metadata (C14M: T2 + PDc + PDt) achieved peak performance (AUC = 0.9961, 95% CI: 0.9823–0.9987, F1 = 0.9618, MCC = 0.9238), outperforming the full protocol (AUC = 0.9909, F1 = 0.9355). Metadata utility was configuration-dependent, assisting only fluid-sensitive combinations. Conclusions: Rather than indiscriminately aggregating entire clinical protocols, multimodal fusion is optimized by selecting complementary imaging series. For binary classification, excluding non-fat-suppressed T1 images in favor of a streamlined T2 and PD set stabilized by clinical demographics maximized classification performance in this internally validated, single-center cohort.
Background/Objectives: Diagnosing Rotator Cuff Tears (RCTs) via Magnetic Resonance Imaging (MRI) is clinically challenging due to complex 3D anatomy and significant in-terobserver variability. Traditional slice-centric Convolutional Neural Networks (CNNs) often fail to capture the necessary volumetric context for accurate grading. This study aims to develop and validate the Patient-Aware Vision Transformer (Pa-ViT), an explainable deep learning framework designed for the automated, patient-level classification of RCTs (Normal, Partial-Thickness, and Full-Thickness). Methods: A large-scale retrospective dataset comprising 2,447 T2-weighted coronal shoulder MRI examinations was utilized. The proposed Pa-ViT framework employs a Vision Transformer (ViT-Base) backbone within a Weakly-Supervised Multiple Instance Learning (MIL) paradigm to aggregate slice-level semantic features into a unified patient diagnosis. The model was trained using a weighted cross-entropy loss to address class imbalance and was benchmarked against widely used CNN architectures and traditional machine learning classifiers. Results: The Pa-ViT model achieved a high overall accuracy of 91% and a macro-averaged F1-score of 0.91, significantly outperforming the standard VGG-16 baseline (87%). Notably, the model demonstrated superior discriminative power for the challenging Partial-Thickness Tear class (ROC AUC: 0.903). Furthermore, Attention Rollout visualizations confirmed the model’s reliance on genuine anatomical features, such as the supraspinatus footprint, rather than artifacts. Conclusions: By effectively modeling long-range dependencies, the Pa-ViT framework provides a robust alternative to traditional CNNs. It offers a clinically viable, explainable decision support tool that enhances diagnostic sensitivity, particularly for subtle partial-thickness tears.
20 March 2022 in Fırat Üniversitesi Mühendislik Bilimleri Dergisi
Bu çalışma, çok sensörlü veri füzyonu ve darboğaz katmanı ile optimize edilmiş evrişimli sinir ağı tabanlı otonom taşıyıcı araçlar için yeni bir hata tespit yöntemi sunmaktadır. Daha zengin öznitelikler elde etmek için tek sensörden gelen sinyal verileri yerine çoklu heterojen sensörlerden gelen sinyal verilerini resimlere dönüştüren bir dönüştürme yönteminden yararlanılmıştır. Çoklu heterojen sensör verilerinin dönüşümü ile elde edilen resimler kullanılarak yeni bir evrişimli sinir ağı oluşturulmuştur. Önerilen yaklaşımın etkinliğini doğrulamak için otonom taşıyıcı araç kullanılarak sensörlerden veriler toplanmış ve gerçek bir fiziksel ortamda çeşitli hata senaryoları oluşturulmuştur. Tek sensör kullanarak ortalama %85 ve çoklu heterojen sensör kullanılarak ortalama %93 hata tahmin doğruluğu elde edilmiştir. Sonuçlarda gösterildiği gibi önerilen evrişimli sinir ağı, daha yüksek tanıma doğruluğu elde etmektedir.
Background/Objectives: Rotator cuff tears are a leading cause of shoulder disability. While multi-sequence MRI is standard, the optimal deep learning integration of het-erogeneous image series and clinical metadata remains unresolved. This study evalu-ated a hierarchical, sequence-aware multimodal framework for patient-level binary rotator cuff tear classification. Methods: A single-center cohort of 199 patients (100 tears, 99 controls) was analyzed across four MRI sequences (T1 coronal, T2 fat-suppressed sagittal, and proton density [PD] fat-suppressed coronal and trans-verse/axial) and nine demographic features. Under a patient-level stratified three-fold cross-validation scheme preventing data leakage, we evaluated ResNet50 and Vision Transformer baselines (Study 0), full-protocol fusion topologies (Study 1), and system-atically mapped sequence-subset combinations with or without metadata (Study 2). Results: In Study 0, the PD coronal ResNet50 model was the top baseline (AUC = 0.9834, F1 = 0.9515). In Study 1, late decision fusion yielded the highest AUC (0.9909), while feature concatenation optimized threshold balance (F1 = 0.9502). In Study 2, a streamlined three-sequence subset with metadata (C14M: T2 + PDc + PDt) achieved peak performance (AUC = 0.9961, 95% CI: 0.9806–0.9986, F1 = 0.9618, MCC = 0.9238), outperforming the full protocol (AUC = 0.9909, F1 = 0.9355). Metadata utility was con-figuration-dependent, assisting only fluid-sensitive combinations. Conclusions: Rather than indiscriminately aggregating entire clinical protocols, multimodal fusion is opti-mized by selecting complementary imaging series. For binary classification, excluding non-fat-suppressed T1 images in favor of a streamlined T2 and PD set stabilized by clinical demographics maximizes diagnostic capacity.