Communication-Efficient Federated Learning -- 3 Experiment -- 3.1 Datasets and Evaluation Metrics -- 3.2 Implementation Details -- 3.3 Comparison of State-of-the-Arts -- 3.4 Detailed Analysis -- 4 Conclusion -- References -- AViT: Adapting Vision Transformers for Small Skin Lesion Segmentation Datasets -- 1 Introduction -- 2 Methodology -- 2.1 Basic ViT -- 2.2 AViT -- 3 Experiments -- 4 Conclusion -- References -- Test-Time Selection for Robust Skin Lesion Analysis -- 1 Introduction -- 2 Related Work -- 3 Methodology -- 4 Results -- 5 Conclusion -- References -- Global and Local Explanations for Skin Cancer Diagnosis Using Prototypes -- 1 Introduction -- 2 Proposed Approach -- 2.1 Clustering -- 2.2 Global Prototypes -- 2.3 Local Prototypes -- 2.4 Pruning and Final Classifier -- 2.5 Visual Explanations -- 3 Experimental Setup -- 4 Results -- 5 Conclusions -- References -- Evidence-Driven Differential Diagnosis of Malignant Melanoma -- 1 Introduction -- 2 Method -- 3 Experiments -- 3.1 Data -- 3.2 Experimental Settings -- 4 Results and Discussion -- 5 Conclusion -- References.
Proceedings of the First Clinically-Oriented and Responsible AI for Medical Data Analysis (Care-AI 2023) Workshop -- An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer's Disease -- 1 Introduction -- 2 Methods -- 2.1 Study Population -- 2.2 Data Preprocessing -- 2.3 Explainable Boosting Machine (EBM) -- 2.4 Proposed Extension -- 2.5 Extraction of the DL-Biomarkers -- 3 Experiments -- 3.1 Validation Study -- 3.2 EBM Using DL-Biomarkers -- 3.3 Baseline Methods -- 4 Results -- 4.1 Glo-CNN and ROIs -- 4.2 Comparison Study -- 5 Discussion and Conclusions -- References -- Generating Chinese Radiology Reports from X-Ray Images: A Public Dataset and an X-ray-to-Reports Generation Method -- 1 Introduction -- 2 Related Work -- 3 Datasets -- 3.1 CN-CXR Dataset -- 3.2 CN-RadGraph Dataset -- 4 Method -- 5 Results and Analysis -- 5.1 Pre-processing and Evaluation -- 5.2 Performance and Comparisons -- References -- Gradient Self-alignment in Private Deep Learning -- 1 Introduction -- 2 Background and Related Work -- 3 Methodology -- 3.1 DP-SGD with a Cosine Similarity Filter -- 3.2 DP-SGD with Dimension-Filtered Cosine Similarity Filter -- 4 Experiments and Discussion -- 4.1 Experimental Setup -- 4.2 Results and Discussion -- 5 Conclusion -- References -- Cellular Features Based Interpretable Network for Classifying Cell-Of-Origin from Whole Slide Images for Diffuse Large B-cell Lymphoma Patients -- 1 Introduction -- 2 Methodology -- 2.1 Nuclei Segmentation and Classification -- 2.2 Cellular Feature Extraction -- 2.3 AMIL Model Training -- 2.4 CellFiNet Interpretability -- 2.5 Benchmark Methods -- 3 Experiments and Results -- 3.1 Data -- 3.2 Results -- 4 Interpretability -- 5 Conclusion and Discussion -- References.
Multimodal Learning for Improving Performance and Explainability of Chest X-Ray Classification -- 1 Introduction -- 2 Method -- 2.1 Classification Performance Experiments -- 2.2 Explainability Experiments -- 3 Results -- 3.1 Classification Performance Results -- 3.2 Explainability Results -- 4 Conclusion -- References -- Proceedings of the First International Workshop on Foundation Models for Medical Artificial General Intelligence (MedAGI 2023) -- Cross-Task Attention Network: Improving Multi-task Learning for Medical Imaging Applications -- 1 Introduction -- 2 Methods and Materials -- 2.1 Cross-Task Attention Network (CTAN) -- 2.2 Training Details -- 2.3 Evaluation -- 2.4 Datasets -- 3 Experiments and Results -- 4 Discussion -- References -- Input Augmentation with SAM: Boosting Medical Image Segmentation with Segmentation Foundation Model -- 1
Introduction -- 2 Related Work -- 3 Methodology -- 3.1 Segmentation and Boundary Prior Maps -- 3.2 Augmenting Input Images -- 3.3 Model Training with SAM-Augmented Images -- 3.4 Model Deployment with SAM-Augmented Images -- 4 Experiments and Results -- 4.1 Datasets and Setups -- 4.2 Polyp Segmentation on Five Datasets -- 4.3 Cell Segmentation on the MoNuSeg Dataset -- 4.4 Gland Segmentation on the GlaS Dataset -- 5 Conclusions -- References -- Empirical Analysis of a Segmentation Foundation Model in Prostate Imaging -- 1 Introduction -- 2 Related Works -- 2.1 UniverSeg -- 2.2 Prostate MR Segmentation -- 3 Experiments -- 3.1 Datasets -- 3.2 UniverSeg Inference -- 3.3 nnUNet -- 3.4 Empirical Evaluation -- 4 Results -- 4.1 Computational Resource -- 4.2 Segmentation Performance -- 4.3 Conclusion -- References -- GPT4MIA: Utilizing Generative Pre-trained Transformer (GPT-3) as a Plug-and-Play Transductive Model for Medical Image Analysis -- 1 Introduction -- 2 Approach -- 2.1 Theoretical Analyses.
2.2 Prompt Construction -- 2.3 Workflow and Use Cases -- 3 Experiments -- 3.1 On Detecting Prediction Errors -- 3.2 On Improving Classification Accuracy -- 3.3 Ablation Studies -- 4 Discussion and Conclusions -- References -- SAM-Path: A Segment Anything Model for Semantic Segmentation in Digital Pathology -- 1 Introduction -- 2 Method -- 2.1 Pathology Encoder -- 2.2 Class Prompts -- 2.3 Optimization -- 3 Experiments -- 3.1 Dataset -- 3.2 Results -- 4 Conclusion -- References -- Multi-task Cooperative Learning via Searching for Flat Minima -- 1 Introduction -- 2 Method -- 2.1 Bi-Level Optimization for Cooperative Two-Task Learning -- 2.2 Finding Flat Minima via Injecting Noise -- 3 Experiments -- 3.1 Dataset -- 3.2 Results on MNIST Dataset -- 3.3 Comparison on REFUGE2018 Dataset -- 3.4 Comparison on HRF-AV Dataset -- 4 Conclusion -- References -- MAP: Domain Generalization via Meta-Learning on Anatomy-Consistent Pseudo-Modalities -- 1 Introduction -- 2 Methods -- 2.1 Problem Definition -- 2.2 Pseudo-modality Synthesis -- 2.3 Meta-learning on Anatomy Consistent Image Space -- 2.4 Structural Correlation Constraints -- 2.5 Experimental Settings -- 3 Results -- 4 Conclusion -- References -- A General Computationally-Efficient 3D Reconstruction Pipeline for Multiple Images with Point Clouds -- 1 Introduction -- 2 Related Works -- 3 Method -- 4 Implementation -- 5 Quantitative Results -- 6 Conclusion and Future Work -- References -- GPC: Generative and General Pathology Image Classifier -- 1 Introduction -- 2 Methodology -- 2.1 Problem Formulation -- 2.2 Network Architecture -- 3 Experiments -- 3.1 Datasets -- 3.2 Comparative Models -- 3.3 Experimental Design -- 3.4 Training Details -- 3.5 Metrics -- 4 Results and Discussion -- 5 Conclusions -- References.
Towards Foundation Models and Few-Shot Parameter-Efficient Fine-Tuning for Volumetric Organ Segmentation -- 1 Introduction -- 2 Related Work -- 3 Methodology -- 4 Experiments -- 5 Conclusions -- References -- Concept Bottleneck with Visual Concept Filtering for Explainable Medical Image Classification -- 1 Introduction -- 2 Related Works -- 2.1 Concept Bottleneck Models -- 2.2 Large Language Models -- 3 Method -- 3.1 Preliminary -- 3.2 Concept Selection with Visual Activation Score -- 4 Experiments -- 4.1 Experimental Results -- 5 Analysis -- 5.1 Analysis on a Visual Activation Score V(c) -- 5.2 Analysis on Target Image Set X -- 5.3 Qualitative Examples -- 6 Conclusion -- References -- SAM Meets Robotic Surgery: An Empirical Study on Generalization, Robustness and Adaptation -- 1 Introduction -- 2 Experimental Settings -- 3 Surgical Instruments Segmentation with Prompts -- 4 Robustness Under Data Corruption -- 5 Automatic
Surgical Scene Segmentation -- 6 Parameter-Efficient Finetuning with Low-Rank Adaptation -- 7 Conclusion -- References -- Evaluation and Improvement of Segment Anything Model for Interactive Histopathology Image Segmentation -- 1 Introduction -- 2 Method -- 2.1 Overview of Segment Anything Model (SAM) -- 2.2 SAM Fine-Tuning Scenarios -- 2.3 Decoder Architecture Modification -- 3 Experiments -- 3.1 Data Description -- 3.2 Implementation Details -- 4 Results -- 4.1 Zero-Shot Performance -- 4.2 Fine-Tuned SAM Performance -- 4.3 Comparison Between SAM and SOTA Interactive Methods -- 4.4 Modified SAM Decoder Performance -- 5 Conclusion -- References -- Task-Driven Prompt Evolution for Foundation Models -- 1 Introduction -- 1.1 Our Approach -- 1.2 Contributions -- 2 Methodology -- 2.1 Prompt Optimization by Oracle Scoring -- 2.2 Learning to Score -- 3 Experiments and Results -- 3.1 Dataset Description -- 3.2 Segmentation Regressor.