بعد أكاديمية كاوست
As of October 2026, Yahya works at Anan Alsama. Yahya served as a teaching assistant for two KAUST Academy bioinformatics courses, finished the Tamheer program at SABIC as an AI Specialist in August 2025, and passed the Microsoft Azure AI Fundamentals exam in October 2026.
المشروع الصيفي
تخصص المعلوماتية الحيوية · دفعة 2024
From CNA Profiles to Prognostic Predictions: Utilizing Hilbert Curve Images for Cancer Subtype Classification and Survival Analysis
Cancer results from DNA mutations, including Copy Number Alterations (CNAs) where large DNA segments are duplicated or deleted. This study examines the potential of CNAs as biomarkers for cancer subtype classification, tumor stage prediction, and survival analysis, using nearly 10,000 samples across 25 cancer types. We employed Hilbert Curve visualizations of CNA profiles and applied computer vision models, including YOLOv8 and VGG16, for classification tasks. A logistic regression model using patient cytoband data served as a control. YOLOv8 performed best in cancer subtype classification, surpassing VGG16 and the control model, although tumor stage classification yielded poor results across all models. Survival analysis with DeepSurv showed fluctuating C-index results, indicating issues with overfitting. Despite these challenges, the study successfully explored the utility of CNAs in cancer research. Some subtypes were classified effectively, suggesting CNAs provide valuable insights into tumor characteristics. Key limitations included dataset size and the mismatch between CNA image simplicity and model complexity. Future work will expand the dataset and develop models tailored to CNA image analysis, improving interpretability and performance.
مع أكاديمية كاوست
- دفعة 2024
جامعة تورنتو · البرنامج الصيفي للمعلوماتية الحيوية
هل تريد تحديث بياناتك أو حذفها؟ تواصل معنا.