Diffusion generative model-empowered modeling and analytics for smart engineering systems
Abstract
Smart engineering systems, including advanced manufacturing and healthcare systems, operate under complex stochastic dynamics characterized by nonlinear variability, noisy observations, and limited or imbalanced data. Although large volumes of different types of data are increasingly available, extracting reliable predictive and generative insights remains challenging due to uncertainty in both process evolution and data distributions. This dissertation investigates diffusion generative models as a unified probabilistic framework for modeling and analytics in smart engineering systems, specifically in advanced manufacturing and healthcare systems. By leveraging the iterative diffusion mechanism, the proposed methodologies transform diffusion models from image-centric generators into adaptable engines capable of addressing variability, uncertainty, and data scarcity across heterogeneous domains: 1. Diffusion-based Variation Modeling (DDIM/KIDIS): A diffusion architecture incorporating advanced distance metrics is proposed to predict nonlinear process variations in additive manufacturing (AM) and tumor trajectory evolution in ultrasound-based image-guided drug delivery (IGDD). 2. Diffusion-based Distribution Learning (IC-DREAM): A probabilistic boundary and intensity distribution learning framework is introduced for ultrasound tumor segmentation, enabling improved gradient estimation and robust region-of-interest identification under severe imaging noise. 3. Diffusion-based…
Read the paper
Related studies
- Learning from First-Generation Qualitative Approaches in the IS Discipline: An Evolutionary View and Some Implications for Authors and Evaluators (PART 1/2)
- Commentary on the clinical and preclinical dosage limits of interstitially administered magnetic fluids for therapeutic hyperthermia based on current practice and efficacy models
- Prosthesis: New Technological Opportunities and Innovative Biomedical Devices
- The history and future of neural modeling for cochlear implants
- Ryanodine receptors: waking up from refractoriness
- Novel Transgenic Rabbit Model Sheds Light on the Puzzling Role of Matrix Metalloproteinase-12 in Atherosclerosis
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.