Deep Generative Models : 4th MICCAI Workshop, DGM4MICCAI 2024, Held in Conjunction with MICCAI 2024, Marrakesh, Morocco, October 10, 2024, Proceedings / / edited by Anirban Mukhopadhyay, Ilkay Oksuz, Sandy Engelhardt, Dorit Mehrof, Yixuan Yuan

• Autore: Mukhopadhyay Anirban
• Pubblicazione: Cham : , : Springer Nature Switzerland : , : Imprint : Springer, , 2025
• Edizione: 1st ed. 2025.
• Descrizione fisica: 1 online resource (235 pages)
• Disciplina: 006.37
• Soggetto topico: Computer vision; Machine learning; Education - Data processing; Application software; Computer Vision; Machine Learning; Computers and Education; Computer and Information Systems Applications
• Altri autori: OksuzIlkay EngelhardtSandy MehrofDorit YuanYixuan
• Nota di contenuto: Intro -- Preface -- Organization -- Contents -- DeReStainer: H& -- E to IHC Pathological Image Translation via Decoupled Staining Channels -- 1 Introduction -- 2 Methods -- 2.1 DeStainer -- 2.2 Feature Fusion Module -- 2.3 ReStainer -- 2.4 Loss Functions -- 3 Experiment -- 3.1 Dataset and Implementation Details -- 3.2 Results -- 3.3 Ablation Study -- 4 Conclusion -- References -- WDM: 3D Wavelet Diffusion Models for High-Resolution Medical Image Synthesis -- 1 Introduction -- 2 Background -- 3 Method -- 4 Experiments -- 4.1 Experimental Settings -- 4.2 Experimental Results -- 5 Conclusion -- References -- Energy-Based Prior Latent Space Diffusion Model for Reconstruction of Lumbar Vertebrae from Thick Slice MRI -- 1 Introduction -- 2 Previous Work -- 3 Method -- 4 Results -- 4.1 Datasets and Metrics -- 4.2 Lumbar Vertebrae Reconstruction -- 4.3 Convergence in the Latent Space -- 5 Conclusions -- References -- Anatomically-Guided Inpainting for Local Synthesis of Normal Chest Radiographs -- 1 Introduction -- 2 Methods -- 2.1 Datasets -- 2.2 Chest Radiography Inpainting -- 2.3 Anatomically-Guided Chest Radiography Inpainting -- 3 Experiments -- 4 Results and Discussion -- 5 Conclusion -- References -- Enhancing Cross-Modal Medical Image Segmentation Through Compositionality -- 1 Introduction -- 2 Methodology -- 3 Experiments -- 4 Conclusions -- References -- Unpaired Modality Translation for Pseudo Labeling of Histology Images -- 1 Introduction -- 2 Methods -- 2.1 Data -- 2.2 Models -- 2.3 Experiments -- 3 Results and Discussion -- 4 Conclusion -- References -- SNAFusion: Distilling 2D Axial Plane Diffusion Priors for Sparse-View 3D Cone-Beam CT Imaging -- 1 Introduction -- 2 Background -- 3 Method -- 3.1 Main Idea -- 3.2 Density Initialization -- 3.3 Density Refinement -- 4 Experiments -- 4.1 Experimential Settings -- 4.2 Results and Discussion. ; 5 Conclusion -- References -- SynthBrainGrow: Synthetic Diffusion Brain Aging for Longitudinal MRI Data Generation in Young People -- 1 Introduction -- 2 Method -- 3 Experiments and Results -- 3.1 Experimental Setup and Dataset -- 3.2 Quantitative Image Quality -- 3.3 Uncertainty Maps as an Explainability Surrogate -- 3.4 Limitations and Future Directions -- 4 Conclusion -- Appendix -- References -- Denoising Diffusion Models for 3D Healthy Brain Tissue Inpainting -- 1 Introduction -- 1.1 Related Work -- 1.2 Contribution -- 2 Methods -- 2.1 Denoising Diffusion Probabilistic Models -- 2.2 Modifying Diffusion Models for Inpainting -- 3 Experiments -- 4 Results -- 5 Discussion -- 6 Conclusion -- References -- Panoptic Segmentation of Mammograms with Text-to-Image Diffusion Model -- 1 Introduction -- 2 Materials and Methods -- 2.1 Segmentation Framework -- 2.2 Datasets -- 2.3 Implementation Details -- 2.4 Evaluation Metrics -- 3 Results -- 4 Discussion -- References -- Interactive Generation of Laparoscopic Videos with Diffusion Models -- 1 Introduction -- 2 Related Work -- 3 Video Generation Pipeline -- 3.1 StableDiffusion -- 3.2 ControlNet -- 3.3 ControlVideo -- 4 Experiments -- 4.1 Data -- 4.2 Training and Inference Details -- 4.3 Evaluation -- 5 Discussion and Conclusion -- References -- Multi-parametric MRI to FMISO PET Synthesis for Hypoxia Prediction in Brain Tumors -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 FMISO Dataset -- 3.2 FDG Dataset -- 3.3 Data Preprocessing -- 3.4 Model -- 4 Results -- 5 Conclusions -- References -- qMRI Diffuser: Quantitative T1 Mapping of the Brain Using a Denoising Diffusion Probabilistic Model -- 1 Introduction -- 2 Method -- 3 Experiments -- 3.1 Scanning Protocol -- 3.2 Datasets -- 3.3 Network Training -- 3.4 Evaluation -- 4 Results -- 5 Discussion -- 6 Conclusion -- References. ; On Differentially Private 3D Medical Image Synthesis with Controllable Latent Diffusion Models -- 1 Introduction -- 2 Methods -- 2.1 Latent Diffusion Model -- 2.2 Differential Privacy -- 2.3 Data and Pre-processing -- 2.4 Evaluation Metrics -- 3 Experiments and Results -- 3.1 Implementation Details -- 3.2 Results -- 4 Discussion and Conclusion -- References -- Five Pitfalls When Assessing Synthetic Medical Images with Reference Metrics -- 1 Introduction -- 2 Data and Methods -- 2.1 Normalization and Binning -- 2.2 Reference Metrics -- 3 Experiments and Results -- 3.1 Pitfall 1: Inappropriate Normalization -- 3.2 Pitfall 2: Similarity of Misaligned Images -- 3.3 Pitfall 3: Background, Foreground and Region of Interest Similarity -- 3.4 Pitfall 4: Error Metrics Prefer Blurred Images -- 3.5 Pitfall 5: Perceptual and Task-Specific Similarity -- 4 Discussion and Conclusion -- References -- Augmenting Prostate MRI Dataset with Synthetic Volumetric Images from Zone-Conditioned Diffusion Generative Model -- 1 Introduction -- 2 Methods -- 2.1 Generative Model -- 3 Experiments -- 3.1 Dataset -- 3.2 New Sample Generation -- 3.3 Downstream Task -- 4 Experiments -- 5 Conclusion -- References -- TiBiX: Leveraging Temporal Information for Bidirectional X-Ray and Report Generation -- 1 Introduction -- 1.1 Multimodal Tokenization -- 1.2 Multimodal Generation Framework -- 2 Datasets and Experiments -- 3 Results and Discussion -- 3.1 SOTA Performance Comparison -- 3.2 Self-comparison -- 3.3 Qualitative Analysis -- 4 Conclusion -- References -- Segmentation-Guided MRI Reconstruction for Meaningfully Diverse Reconstructions -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Background Diffusion Models and MRI-Reconstruction -- 3.2 Segmentation Guidance for Diverse Sampling -- 3.3 Data and Models -- 4 Experiments and Results -- 5 Conclusion and Discussion. ; References -- Non-reference Quality Assessment for Medical Imaging: Application to Synthetic Brain MRIs -- 1 Introduction -- 2 Methods -- 2.1 Generative Network -- 2.2 Quality Metrics -- 2.3 Quality Network -- 2.4 Augmentation and Inference -- 3 Experiments and Results -- 3.1 Data -- 3.2 Setup -- 3.3 Results -- 4 Conclusion -- References -- LatentArtiFusion: An Effective and Efficient Histological Artifacts Restoration Framework -- 1 Introduction -- 2 Method -- 2.1 Preliminary in Diffusion Models -- 2.2 LatentArtiFusion -- 3 Experiments -- 3.1 Implementation Details -- 3.2 Datasets -- 3.3 Comparison and Results -- 3.4 Downstream Classification Evaluation -- 4 Conclusion -- References -- How to Segment in 3D Using 2D Models: Automated 3D Segmentation of Prostate Cancer Metastatic Lesions on PET Volumes Using Multi-angle Maximum Intensity Projections and Diffusion Models -- 1 Introduction -- 2 Methods and Materials -- 2.1 Dataset -- 2.2 Data Preprocessing -- 2.3 Segmentation Network Architecture -- 2.4 3D Reconstruction of 2D Masks Using OSEM Algorithm -- 3 Experiment Details -- 4 Results and Discussion -- 5 Conclusion -- References -- Author Index.
• Sommario/riassunto: This book constitutes the proceedings of the 4th workshop on Deep Generative Models for Medical Image Computing and Computer Assisted Intervention, DGM4MICCAI 2024, held in conjunction with the 27th International conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2024, in Marrakesh, Morocco in October 2024. The 21 papers presented here were carefully reviewed and selected from 40 submissions. These papers deal with a broad range of topics, ranging from methodology (such as Causal inference, Latent interpretation, Generative factor analysis) to Applications (such as Mammography, Vessel imaging, Surgical videos and more).
• Titolo autorizzato: Deep Generative Models
• ISBN: 3-031-72744-4
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 9910983044503321
• Lo trovi qui: Univ. Federico II
• Serie: Lecture Notes in Computer Science, . 1611-3349 ; ; 15224