سامانه همایش های دانشگاه آزاد اسلامی واحد نجف آباد

صفحه اصلی / اولین همایش بین المللی تحول خدمات سلامت: تلفیق مراقبت، درمان، بازتوانی و امیدبخشی

Artificial Intelligence from Diagnosis to Treatment of Disease: Spinal Imaging, Biomaterials, and Drug Delivery in Biomedical Engineering

نویسندگان :

Javid Mirzaei¹ Mitra Vahed Pirkoohi²

1- دانشگاه آزاد اسلامی واحد لاهیجان 2- دانشگاه آزاد اسلامی واحد لاهیجان

کلمات کلیدی :

Artificial Intelligence،Machine Learning،Tissue Scaffolds،Drug Delivery Systems،Magnetic Resonance Imaging،Intervertebral Disc Degeneration

چکیده :

Abstract Background: Intervertebral disc degeneration is a leading cause of chronic low back pain, and AI is increasingly applied from imaging-based diagnosis to regenerative treatment design and drug delivery. Although typically studied separately, both domains increasingly rely on similar AI and machine learning (ML) methods. This study aimed to provide an integrated review of these applications across this diagnosis-to-treatment pathway. Method: Twenty open-access primary research articles were selected through a narrative (nonsystematic) PubMed search conducted during the first six months of 2026, using AI-, ML-, and biomedical-engineering-related keywords, then screened from an initial pool of about forty candidates for relevance to diagnostic imaging, regenerative treatment design, or drug delivery. Finding: U-Net deep-learning models trained on lumbar MRI achieved a 0.935 Dice similarity coefficient for automatic disc segmentation [1], and convolutional neural network (CNN) models graded disc degeneration with 92%-97% accuracy, depending on whether the training, testing, or external-validation set was used [2]. In treatment: a genetic-algorithm approach for directional bone-scaffold design [3], an online tool predicting electrospun fibre-diameter distribution with 84% overlap versus experimental data [4], and a physics-informed neural network predicting nanocarrier drug toxicity (mean R²=0.87, 95% CI 0.85-0.89) [5] were reported. Conclusion: AI plays a growing, complementary role across this pathway; diagnostic applications rely mainly on MRI-based deep learning, while treatment applications draw on a broader mix of classical and deep-learning methods, though clinical validation in human patients remains limited.