arXivSub Start free trial

MICCAI 2024 Papers with Code

International Conference on Medical Image Computing and Computer-Assisted Intervention Β· 487 papers with a public code repository

3D Vessel Graph Generation Using Denoising Diffusion

Prabhakar, Chinmay (University of Zurich), Menze, Bjoern (University of Zurich)

CodeGenerationData SynthesisGraph Neural NetworkTransformerDiffusion modelGraphBiomedical Data

🎯 What it does: A two-stage generative model based on denoising diffusion is proposed, which first generates the coordinates of vascular nodes and then generates the connectivity between nodes, resulting in a three-dimensional vascular network diagram.

3D-SAutoMed: Automatic Segment Anything Model for 3D Medical Image Segmentation from Local-Global Perspective

Liang, Junjie (Northeastern University), Zaiane, Osmar R. (Amii)

CodeSegmentationTransformerSupervised Fine-TuningBiomedical DataMagnetic Resonance Imaging

🎯 What it does: The 3D-SAutoMed framework is proposed to achieve automation of SAM in 3D medical image segmentation and 3D awareness.

3DDX: Bone Surface Reconstruction from a Single Standard-Geometry Radiograph via Dual-Face Depth Estimation

Gu, Yi (OMRON SINIC X Corporation), Sato, Yoshinobu (Nara Institute of Science and Technology)

CodeSegmentationDepth EstimationConvolutional Neural NetworkAuto EncoderImagePoint CloudComputed Tomography

🎯 What it does: The 3DDX method is proposed, which completes the full 3D bone surface reconstruction by simultaneously predicting the depth maps of the anterior and posterior surfaces of the bone from a single standard geometric X-ray image.

3DGR-CAR: Coronary artery reconstruction from ultra-sparse 2D X-ray views with a 3D Gaussians representation

Fu, Xueming (University of Science and Technology of China), Zhou, S. Kevin (Hohai University)

CodeSegmentationGenerationOptimizationGaussian SplattingImageBiomedical DataComputed Tomography

🎯 What it does: This paper proposes the use of 3D Gaussian representation to reconstruct the three-dimensional structure of coronary arteries from extremely sparse 2D X-ray views.

7T MRI Synthesization from 3T Acquisitions

Cui, Qiming (University of California San Francisco), Abbasi-Asl, Reza (University of California San Francisco)

CodeImage TranslationSegmentationData SynthesisConvolutional Neural NetworkGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: Developed and evaluated various V-Net-based deep learning models to synthesize 7T MRI from 3T MRI to improve image quality and lesion visualization.

A Bayesian Approach to Weakly-supervised Laparoscopic Image Segmentation

Zheng, Zhou (Nagoya University), Mori, Kensaku (Aichi Institute of Technology)

CodeSegmentationConvolutional Neural NetworkAuto EncoderImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: A weakly supervised laparoscopic image segmentation method based on a Bayesian framework is proposed, which generates high-quality pseudo-labels using sparse annotations and trains the segmentation model, while also quantifying prediction uncertainty.

A Domain Adaption Approach for EEG-based Automated Seizure Classification with Temporal-Spatial-Spectral Attention

Fan, Xiaoya (Dalian University of Technology), Wang, Zhong (Dalian University of Technology)

CodeClassificationDomain AdaptationConvolutional Neural NetworkTime SeriesBiomedical Data

🎯 What it does: This study proposes a classification model for EEG seizure types based on domain adaptation and a spatiotemporal frequency spectrum attention mechanism.

A Foundation Model for Brain Lesion Segmentation with Mixture of Modality Experts

Zhang, Xinru (Beijing Institute of Technology), Bai, Wenjia (Imperial College London)

CodeSegmentationConvolutional Neural NetworkMixture of ExpertsImageMultimodalityMagnetic Resonance Imaging

🎯 What it does: The MoME foundational model is proposed to achieve multi-modal brain lesion segmentation.

A Graph-Embedded Latent Space Learning and Clustering Framework for Incomplete Multimodal Multiclass Alzheimer’s Disease Diagnosis

Ou, Zaixin (ShanghaiTech University), Shen, Dinggang (Shanghai United Imaging Intelligence Co., Ltd.)

CodeClassificationGraph Neural NetworkMultimodalityBiomedical DataMagnetic Resonance ImagingPositron Emission TomographyAlzheimer's Disease

🎯 What it does: This paper proposes a graph embedding latent space learning and clustering framework (Graph-SLC) for multi-class diagnosis of Alzheimer's disease (AD) in the presence of missing multimodal PET data.

A Hybrid CNN-Transformer Feature Pyramid Network for Granular Abdominal Aortic Calcification Detection from DXA Images

Ilyas, Zaid (Edith Cowan University), Gilani, Syed Zulqarnain (University of Western Australia)

CodeClassificationObject DetectionConvolutional Neural NetworkTransformerImage

🎯 What it does: A CNN-Transformer model called Hybrid-FPN-AACNet based on Hybrid FPN is proposed, which achieves fine-grained scoring regression of abdominal aortic calcification in DXA VFA images (AAC-24 scores for segments L1-L4) through Dual Resolution Self-Attention (DRSA) and Efficient Feature Fusion Module (EFFM);

A New Benchmark In Vivo Paired Dataset for Laparoscopic Image De-smoking

Xia, Wenyao (Western University), Chen, Elvis C. S. (Western University)

CodeImage TranslationRestorationConvolutional Neural NetworkTransformerGenerative Adversarial NetworkOptical FlowImageVideoBenchmark

🎯 What it does: A real paired dataset of surgical smoke and smoke-free images was constructed for the first time, and existing smoke removal algorithms were evaluated using this dataset.

A New Cine-MRI Segmentation Method of Tongue Dorsum for Postoperative Swallowing Function Analysis

Sun, Minghao (Nanjing University of Posts and Telecommunications), Yu, Han (Nanjing University of Posts and Telecommunications)

CodeSegmentationConvolutional Neural NetworkTransformerImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: A tongue dorsum segmentation network based on convolution + temporal Transformer is proposed, and quantitative analysis of swallowing function is achieved by automatically extracting tongue dorsum feature points.

A New Perspective to Boost Performance Fairness For Medical Federated Learning

Yan, Yunlu (Hong Kong University of Science and Technology (Guangzhou)), Feng, Chun-Mei (Agency for Science, Technology and Research)

CodeSegmentationFederated LearningContrastive LearningImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: The Fed-LWR method is proposed to address the unfairness caused by feature shift in medical federated learning through hierarchical weight re-aggregation.

A Novel Adaptive Hypergraph Neural Network for Enhancing Medical Image Segmentation

Chai, Shurong (Ritsumeikan University), Chen, Yen-Wei (Ritsumeikan University)

CodeSegmentationGraph Neural NetworkImageBiomedical DataMagnetic Resonance ImagingComputed Tomography

🎯 What it does: A medical image segmentation framework based on adaptive hypergraphs is proposed, integrating hypergraph construction and convolutional hypergraph techniques to enhance segmentation accuracy.

A Refer-and-Ground Multimodal Large Language Model for Biomedicine

Huang, Xiaoshuang (Baidu Inc), Liu, Jia (Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences)

CodeObject DetectionSegmentationTransformerLarge Language ModelSupervised Fine-TuningVision Language ModelImageTextMultimodalityBiomedical DataMagnetic Resonance ImagingComputed TomographyUltrasound

🎯 What it does: A medical imaging reference-localization instruction tuning dataset, Med-GRIT-270k, was constructed, and the BiRD (Biomedical Refer-and-Ground Multimodal Large Language Model) model was fine-tuned based on this dataset to address the lack of reference and localization capabilities in biomedical multimodal models.

A Region-Based Approach to Diabetic Retinopathy Classification with Superpixel Tokenization

Playout, Clément, Cheriet, Farida (Polytechnique Montreal)

CodeClassificationSegmentationTransformerImage

🎯 What it does: Perform SLIC superpixel segmentation on retinal images, extract feature vectors for each superpixel, and use them as tokens for DR grading in the Vision Transformer.

A Unified Model for Longitudinal Multi-Modal Multi-View Prediction with Missingness

Chen, Boqi (University of North Carolina at Chapel Hill), Niethammer, Marc (UCSD)

CodeTransformerMultimodalityTabularBiomedical Data

🎯 What it does: A unified longitudinal multimodal multi-view prediction model is proposed, which can flexibly handle missing data and inputs at any time point.

A Universal and Flexible Framework for Unsupervised Statistical Shape Model Learning

El Amrani, Nafie (University of Bonn), Bernard, Florian (University of Bonn)

CodeSegmentationGenerationBiomedical DataMagnetic Resonance Imaging

🎯 What it does: A completely unsupervised deep learning framework is proposed for constructing statistical shape models (SSM) of medical anatomical structures, achieving high-quality shape matching and deformation through both correspondence and deformation methods.

A Weakly-supervised Multi-lesion Segmentation Framework Based on Target-level Incomplete Annotations

Ju, Jianguo (Northwest University), Guan, Ziyu (Northwest University)

CodeSegmentationConvolutional Neural NetworkImageBiomedical DataComputed Tomography

🎯 What it does: A weakly supervised multi-lesion segmentation framework based on Target-level Incomplete Annotations (TIA) is proposed, which can segment Crohn's disease CT images with very few annotations.

Achieving Fairness Through Channel Pruning for Dermatological Disease Diagnosis

Kong, Qingpeng (University of Notre Dame), Shi, Yiyu (University of Notre Dame)

CodeClassificationConvolutional Neural NetworkSupervised Fine-TuningImage

🎯 What it does: A channel pruning framework based on Soft Nearest Neighbor Loss (SNNL) is proposed to enhance the fairness of skin disease diagnosis models by identifying and removing channels that significantly affect sensitive attributes.

AcneAI: A new acne severity assessment method using digital images and deep learning

Gazeau, LΓ©a (Torus AI), Wolfe, Jonathan (BelleTorus Corporation)

CodeClassificationObject DetectionSegmentationConvolutional Neural NetworkImage

🎯 What it does: The AcneAI system has been developed to achieve full-process automatic processing of facial acne images: first segmenting all acne and acne-like lesions, then scoring each lesion, and finally synthesizing an overall severity score from 0 to 100.

Across-subject ensemble-learning alleviates the need for large samples for fMRI decoding

Aggarwal, Himanshu (Inria), Thirion, Bertrand (UniversitΓ© Paris-Saclay)

CodeClassificationRecognitionBiomedical DataMagnetic Resonance Imaging

🎯 What it does: This paper improves the accuracy of fMRI brain state decoding through stacked ensemble learning, utilizing classifiers pre-trained on other subjects, and compares it with traditional single-subject training methods.

Active Label Refinement for Robust Training of Imbalanced Medical Image Classification Tasks in the Presence of High Label Noise

Khanal, Bidur (Rochester Institute of Technology), Linte, Cristian (Rochester Institute of Technology)

CodeClassificationConvolutional Neural NetworkImageBiomedical Data

🎯 What it does: A two-stage process is proposed: first, clean samples are identified using Co-Teaching and Gradient Variance (VOG), and then residual noisy labels are gradually cleaned through expert polling-based active learning to enhance the robustness of imbalanced medical image classification.

AdaCBM: An Adaptive Concept Bottleneck Model for Explainable and Accurate Diagnosis

Chowdhury, Townim F. (University of Adelaide), Liao, Zhibin (University of Adelaide)

CodeDomain AdaptationExplainability and InterpretabilityLarge Language ModelPrompt EngineeringBiomedical Data

🎯 What it does: This paper proposes AdaCBM, an interpretable diagnostic model that combines CLIP with the Concept Bottleneck Model (CBM) and inserts learnable adapters between the two, utilizing GPT-4 for prompt-based concept generation and filtering.

Adaptive Curriculum Query Strategy for Active Learning in Medical Image Classification

Ma, Siteng (University College Dublin), Dong, Ruihai (University College Dublin)

CodeClassificationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: An adaptive query strategy based on curriculum learning, ACAL, is proposed for active learning in medical image classification.

Adaptive Smooth Activation Function for Improved Organ Segmentation and Disease Diagnosis

Biswas, Koushik (Northwestern University), Bagci, Ulas (Northwestern University)

CodeClassificationSegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingComputed TomographyBenchmark

🎯 What it does: The Adaptive Smooth Activation Unit (ASAU) is proposed as a new smooth activation function, applied to multi-class disease classification and liver segmentation tasks in medical imaging.

Advancing Brain Imaging Analysis Step-by-step via Progressive Self-paced Learning

Yang, Yanwu (University of TΓΌbingen), Ma, Ting (Harbin Institute of Technology at Shenzhen)

CodeClassificationKnowledge DistillationConvolutional Neural NetworkBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease

🎯 What it does: An advanced self-adaptive self-supervised knowledge distillation framework (PSPD) is proposed for brain imaging classification tasks.

Advancing H&E-to-IHC Virtual Staining with Task-Specific Domain Knowledge for HER2 Scoring

Peng, Qiong (Xiamen University), Wang, Liansheng (Shanghai Changhai Hospital)

CodeImage TranslationGenerationGenerative Adversarial NetworkImage

🎯 What it does: This paper proposes a GAN-based virtual staining method that converts H&E stained images into HER2 IHC stained images, achieving precise retention of nuclear distribution and membrane staining intensity through a kernel density estimator and a membrane staining enhancement branch.

Advancing UWF-SLO Vessel Segmentation with Source-Free Active Domain Adaptation and a Novel Multi-Center Dataset

Wang, Hongqiu (Hong Kong University of Science and Technology), Zhu, Lei (Hong Kong University of Science and Technology)

CodeSegmentationDomain AdaptationConvolutional Neural NetworkImage

🎯 What it does: To address the problem of vessel segmentation in ultra-wide field scanning laser ophthalmoscope (UWF-SLO) images, the authors propose a patch-based source-free active domain adaptation (SFADA) framework and design a cascading uncertainty-dominant (CUP) selection strategy, which can significantly enhance cross-center segmentation performance with a minimal amount of annotated patches.

Affinity Learning Based Brain Function Representation for Disease Diagnosis

Liu, Mengjun (Shanghai Jiao Tong University), Wang, Qian (United Imaging Intelligence)

CodeClassificationRepresentation LearningGraph Neural NetworkLarge Language ModelSupervised Fine-TuningBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease

🎯 What it does: A brain function representation framework based on affinity learning is proposed, using randomly sampled 3D patches to obtain an information-rich functional connectivity matrix through self-supervised learning, which is then used for disease diagnosis.

Airway segmentation based on topological structure enhancement using multi-task learning

Yang, Xuan (Shenzhen University), Liao, Hongen (Tsinghua University)

CodeSegmentationConvolutional Neural NetworkImageBiomedical DataComputed Tomography

🎯 What it does: This paper proposes a 3D U-Net framework based on multi-task learning, using centerline detection as an auxiliary task to enhance the topological integrity of airway segmentation. It further improves branch connectivity and the identification of small bronchi through a newly established Topological Embedding Interaction Module (TEIM) and a Topological Enhanced Attention Module (TEAM).

Algebraic Sphere Surface Fitting for Accurate and Efficient Mesh Reconstruction from Cine CMR Images

He, Jin (Beijing Normal University), Wang, Shuo (Fudan University)

CodeSegmentationGenerationComputational EfficiencyImagePoint CloudMeshMagnetic Resonance Imaging

🎯 What it does: A framework for reconstructing the ventricular surface based on algebraic spherical fitting is proposed, which can directly generate a complete three-dimensional mesh from a sparse layered wireframe point cloud of a single frame cine CMR image.

Algorithmic Fairness in Lesion Classification by Mitigating Class Imbalance and Skin Tone Bias

Ansari, Faizanuddin (Indian Statistical Institute), Das, Swagatam (Indian Statistical Institute)

CodeClassificationConvolutional Neural NetworkSupervised Fine-TuningImage

🎯 What it does: A data augmentation framework that combines adaptive sampling and Mixup is proposed to simultaneously alleviate skin tone bias and class imbalance, improving the accuracy and fairness of skin lesion classification.

Aligning Human Knowledge with Visual Concepts Towards Explainable Medical Image Classification

Gao, Yunhe (Rutgers University), Metaxas, Dimitris (Rutgers University)

CodeClassificationExplainability and InterpretabilityTransformerLarge Language ModelVision Language ModelContrastive LearningImageBiomedical Data

🎯 What it does: Proposes the Explicd framework, which transforms LLM or expert knowledge into diagnostic criteria and learns visual concepts in visual language models to achieve interpretable medical image classification.

All-In-One Medical Image Restoration via Task-Adaptive Routing

Yang, Zhiwen (Beihang University), Xu, Yan (Beihang University)

CodeRestorationSuper ResolutionConvolutional Neural NetworkMixture of ExpertsImageBiomedical DataMagnetic Resonance ImagingComputed TomographyPositron Emission Tomography

🎯 What it does: A full-task medical image restoration network called AMIR is proposed, which can simultaneously accomplish MRI super-resolution, CT denoising, and PET synthesis within a single model.

AMONuSeg: A Histological Dataset for African Multi-Organ Nuclei Semantic Segmentation

Zerouaoui, Hasnae (Mohammed VI Polytechnic University), Navab, Nassir (Technische UniversitΓ€t MΓΌnchen)

CodeSegmentationTransformerContrastive LearningImageBiomedical Data

🎯 What it does: The first publicly available African multi-organ H&E stained tissue nucleus segmentation dataset, AMONuSeg, collected using low-resource microscopy, is proposed and evaluated with various existing segmentation models and stain normalization methods.

An Evaluation of State-of-the-Art Projectors in the Presence of Noise and Nonlinearity in the Beer-Lambert Law

Xie, Shiyu (University of Florida), Entezari, Alireza (University of Florida)

CodeImageComputed Tomography

🎯 What it does: In the context of low-dose/limited-angle CT reconstruction considering the Beer-Lambert nonlinearity and Poisson noise, the reconstruction errors of three mainstream fast projectors (SF, LTRI, CNSF) are evaluated against an accurate projector.

An Uncertainty-guided Tiered Self-training Framework for Active Source-free Domain Adaptation in Prostate Segmentation

Luo, Zihao (University of Electronic Science and Technology of China), Wang, Guotai (University Of Electronic Science And Technology Of China)

CodeSegmentationDomain AdaptationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: An Active Source Free Domain Adaptation (ASFDA) framework UGTST is proposed for prostate MRI segmentation.

Anatomic-constrained Medical Image Synthesis via Physiological Density Sampling

Chu, Yuetan, Gao, Xin (Chinese Academy of Sciences)

CodeSegmentationGenerationData SynthesisReinforcement LearningGenerative Adversarial NetworkImageBiomedical DataComputed Tomography

🎯 What it does: This paper proposes a medical image synthesis framework based on anatomical constraints, ACIS, which uses pseudo-masks as anatomical priors to generate diverse and realistic CT images along with corresponding annotations. It significantly enhances segmentation and reconstruction performance in few-shot learning through post-processing refinement using registration uncertainty.

Anatomical Positional Embeddings

Goncharov, Mikhail (Skolkovo Institute of Science and Technology), Oseledets, Ivan (Skolkovo Institute of Science and Technology)

CodeSegmentationRetrievalConvolutional Neural NetworkContrastive LearningImageBiomedical DataComputed Tomography

🎯 What it does: A self-supervised 3D Anatomical Position Embedding (APE) model is proposed, capable of generating three-dimensional position embeddings for each voxel in medical CT images at once, and applying it to anatomical label retrieval and few-shot organ localization.

Anatomically-Guided Segmentation of Cerebral Microbleeds in T1-weighted and T2*-weighted MRI

Kwon, Junmo (Sungkyunkwan University), Park, Hyunjin (VUNO Inc.)

CodeObject DetectionSegmentationConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: This paper achieves automatic segmentation and detection of cerebral microbleeds by incorporating an anatomical proxy task (segmenting the brainstem, deep structures, and the lower ventricular regions) into the CMB segmentation network.

Anatomy-Aware Gating Network for Explainable Alzheimer’s Disease Diagnosis

Jiang, Hongchao (Nanyang Technological University), Miao, Chunyan (Nanyang Technological University)

CodeClassificationExplainability and InterpretabilityConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease

🎯 What it does: A gated network based on anatomical information (AAGN) has been designed and implemented, achieving interpretable Alzheimer's disease diagnosis by dynamically selecting features from different brain regions.

Are We Ready for Out-of-Distribution Detection in Digital Pathology?

Oh, Ji-Hun (University of Illinois Urbana Champaign), Bhargava, Rohit (University of Illinois Urbana Champaign)

CodeAnomaly DetectionConvolutional Neural NetworkTransformerSupervised Fine-TuningImageBiomedical DataBenchmark

🎯 What it does: Conducted benchmark experiments on out-of-distribution detection in digital pathology (S-OODD and MC-OODD), systematically evaluating various detectors, transfer learning strategies, and model architectures.

ASPS: Augmented Segment Anything Model for Polyp Segmentation

Li, Huiqian (University of Science and Technology of China), Han, Junwei (Xi'an Jiaotong University)

CodeSegmentationDomain AdaptationConvolutional Neural NetworkTransformerImageComputed Tomography

🎯 What it does: This paper proposes the ASPS (Augmented Segment Anything Model for Polyp Segmentation) model, which enhances multi-scale feature extraction and domain adaptation capabilities for colonoscopic images by adding a trainable CNN encoder branch and uncertainty-guided prediction regularization based on SAM.

Automated Spinal MRI Labelling from Reports Using a Large Language Model

Park, Robin Y. (University of Oxford), Zisserman, Andrew (University of Oxford)

CodeClassificationConvolutional Neural NetworkLarge Language ModelPrompt EngineeringTextBiomedical DataMagnetic Resonance Imaging

🎯 What it does: This paper proposes a general pipeline based on large language models that can automatically extract structured labels from radiology reports.

Automatic Mandibular Semantic Segmentation of Teeth Pulp Cavity and Root Canals, and Inferior Alveolar Nerve on Pulpy3D Dataset

Gamal, Mahmoud (Alexandria University), Torki, Marwan (Alexandria University)

CodeSegmentationConvolutional Neural NetworkImageBiomedical DataComputed Tomography

🎯 What it does: A complete 3D CBCT semantic segmentation system for the mandibular pulp chamber, root canal, and inferior alveolar nerve (IAN) has been developed, and the Pulpy3D dataset has been proposed.

Average Calibration Error: A Differentiable Loss for Improved Reliability in Image Segmentation

Barfoot, Theodore (King's College London), Vercauteren, Tom (King's College London)

CodeSegmentationConvolutional Neural NetworkBiomedical DataMagnetic Resonance Imaging

🎯 What it does: A differentiable auxiliary loss function mL1-ACE is proposed to enhance the confidence calibration of medical image segmentation models, and it is jointly trained with traditional losses (cross-entropy, Dice) on the BraTS 2021 dataset.

BackMix: Mitigating Shortcut Learning in Echocardiography with Minimal Supervision

Bransby, Kit M. (Ultromics Ltd), Gomez, Alberto (Ultromics Ltd)

CodeClassificationSegmentationConvolutional Neural NetworkSupervised Fine-TuningContrastive LearningImageBiomedical DataUltrasound

🎯 What it does: The study proposes two random background mixing enhancement methods, BackMix and wBackMix, to address the shortcut learning problem caused by background information in echocardiogram view classification, achieving performance improvement under semi-supervised conditions with only 5% segmentation annotations.

Baikal: Unpaired Denoising of Fluorescence Microscopy Images using Diffusion Models

Chaudhary, Shivesh (Calico Life Sciences LLC), Xu, Jun (Heilongjiang University)

CodeRestorationDiffusion modelImage

🎯 What it does: A denoising framework for fluorescence microscopy images called Baikal is constructed based on unaligned data, utilizing a Denoising Diffusion Probabilistic Model (DDPM) to learn the generative prior of clean images, and then recovering clean images from noisy images during inference through various conditional sampling strategies (Forward-Backward, Mixing, Repaint).

BAPLe: Backdoor Attacks on Medical Foundational Models using Prompt Learning

Hanif, Asif (Mohamed Bin Zayed University of Artificial Intelligence), Anwer, Rao Muhammad (Mohamed Bin Zayed University of Artificial Intelligence)

CodeAdversarial AttackData-Centric LearningTransformerPrompt EngineeringContrastive LearningBiomedical DataMagnetic Resonance Imaging

🎯 What it does: This paper proposes a backdoor attack method called BAPLe that utilizes prompt learning to embed backdoors in medical foundation models (Med-FM), achieving efficient backdoor injection in data-scarce scenarios while keeping the model frozen.

Beyond Adapting SAM: Towards End-to-End Ultrasound Image Segmentation via Auto Prompting

Lin, Xian (Huazhong University of Science and Technology), Yan, Zengqiang (Huazhong University of Science and Technology)

CodeSegmentationConvolutional Neural NetworkTransformerPrompt EngineeringImageBiomedical DataUltrasound

🎯 What it does: This paper presents SAMUS and its end-to-end automatic segmentation version AutoSAMUS, which improves the performance of SAM on ultrasound images by using parallel CNN branches, position/feature adapters, and cross-branch attention to supplement local information, achieving fully automatic segmentation through an automatic prompt generator.

BGDiffSeg: a Fast Diffusion Model for Skin Lesion Segmentation via Boundary Enhancement and Global Recognition Guidance

Guo, Yilin (Sun Yat-sen University), Cai, Qingling (Sun Yat-sen University)

CodeSegmentationDiffusion modelGenerative Adversarial NetworkImageBiomedical Data

🎯 What it does: This paper proposes a fast skin lesion segmentation framework called BGDiffSeg based on diffusion models, aiming to achieve more accurate and clearer lesion boundary segmentation under limited computational resources and inference time.

BGF-YOLO: Enhanced YOLOv8 with Multiscale Attentional Feature Fusion for Brain Tumor Detection

Kang, Ming (Monash University), Phan, RaphaΓ«l C.-W. (Monash University)

CodeObject DetectionConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: Proposed and implemented BGF-YOLO, an improved YOLOv8 network for brain tumor MRI detection.

Biomechanics-informed Non-rigid Medical Image Registration and its Inverse Material Property Estimation with Linear and Nonlinear Elasticity

Min, Zhe (Shandong University), Hu, Yipeng (King's College London)

CodePoint CloudBiomedical DataMagnetic Resonance Imaging

🎯 What it does: Using Physics-Informed Neural Networks (PINN) to achieve non-rigid point set registration of soft tissues and material property estimation in both forward and inverse problems, and comparing linear and nonlinear elastic models.

Black-Box Adaptation for Medical Image Segmentation

Paranjape, Jay N. (Johns Hopkins University), Patel, Vishal M. (Johns Hopkins University)

CodeSegmentationTransformerPrompt EngineeringSimultaneous Localization and MappingImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: The BAPS (Black-Box Adapter for Prompted Segmentation) method is proposed, which enhances the performance of medical image segmentation through visual prompts by utilizing a pre-trained image encoder and a trainable image-prompt decoder without accessing the underlying model parameters.

Boosting FFPE-to-HE Virtual Staining with Cell Semantics from Pretrained Segmentation Model

Hu, Yihuang (Xiamen University), Wang, Liansheng (Shanghai Changhai Hospital)

CodeImage TranslationGenerationGenerative Adversarial NetworkImage

🎯 What it does: This paper proposes a cGAN framework for enhancing FFPE-to-HE virtual staining using cell semantics extracted from a pre-trained cell segmentation model.

BPaCo: Balanced Parametric Contrastive Learning for Long-tailed Medical Image Classification

Cai, Zhiyuan (Southern University of Science and Technology), Tang, Xiaoying (Chinese University of Hong Kong)

CodeClassificationContrastive LearningImage

🎯 What it does: This paper proposes the BPaCo framework, which combines contrastive learning to achieve end-to-end training for the long-tail classification problem in medical images.

BrainSCK: Brain Structure and Cognition Alignment via Knowledge Injection and Reactivation for Diagnosing Brain Disorders

Wang, Lilong (Shanghai AI Laboratory), Wang, Xiaosong (Shanghai Artificial Intelligence Laboratory)

CodeTransformerPrompt EngineeringVision Language ModelImageTextBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease

🎯 What it does: The BrainSCK framework is proposed, which aligns brain structural images with cognitive knowledge using a two-stage visual-language model to achieve diagnosis of brain diseases across the lifespan;

BrainWaveNet: Wavelet-based Transformer for Autism Spectrum Disorder Diagnosis

Jeong, Ah-Yeong (Korea University), Suk, Heung-Il (Korea University)

CodeClassificationTransformerTime SeriesBiomedical DataMagnetic Resonance Imaging

🎯 What it does: This paper proposes BrainWaveNet, a waveform network based on Continuous Wavelet Transform (CWT) and Transformer for the diagnosis of Autism Spectrum Disorder (ASD).

Cache-Driven Spatial Test-Time Adaptation for Cross-Modality Medical Image Segmentation

Li, Xiang (Nanyang Technological University), Xu, Yanwu (South China University of Technology)

CodeSegmentationDomain AdaptationBiomedical DataMagnetic Resonance ImagingComputed Tomography

🎯 What it does: This paper proposes a cache-driven Spatial Test-Time Adaptation (STTA) method to address the domain shift problem in cross-modal medical image segmentation.

Can LLMs’ Tuning Methods Work in Medical Multimodal Domain?

Chen, Jiawei (Fudan University), Zhang, Lihua (Fudan University)

CodeTransformerLarge Language ModelVision Language ModelMultimodalityBiomedical Data

🎯 What it does: This paper transfers the parameter-efficient fine-tuning (PEFT) methods of LLM to medical multimodal vision-language models, constructing the MILE framework and systematically evaluating the effects of techniques such as LoRA, Prefix, IA3, and P-Tuning v2 on medical VLMs.

CAR-MFL: Cross-Modal Augmentation by Retrieval for Multimodal Federated Learning with Missing Modalities

Poudel, Pranav (NepAl Applied Mathematics and Informatics Institute for research), Bhattarai, Binod (University of Aberdeen)

CodeFederated LearningSafty and PrivacyConvolutional Neural NetworkTransformerImageTextMultimodalityElectronic Health RecordsRetrieval-Augmented Generation

🎯 What it does: A cross-modal retrieval-enhanced framework CAR-MFL is proposed, which utilizes a small amount of public multimodal data to retrieve and complete missing modalities in multimodal federated learning while maintaining data privacy.

Cardiovascular Disease Detection from Multi-View Chest X-rays with BI-Mamba

Yang, Zefan (Rensselaer Polytechnic Institute), Yan, Pingkun (Rensselaer Polytechnic Institute)

CodeClassificationAnomaly DetectionComputational EfficiencyConvolutional Neural NetworkTransformerImageMultimodalityComputed Tomography

🎯 What it does: This study investigates the use of multi-view chest X-rays to predict cardiovascular disease risk.

CausalCLIPSeg: Unlocking CLIP’s Potential in Referring Medical Image Segmentation with Causal Intervention

Chen, Yaxiong (Wuhan University of Technology), Mou, Lichao (Xidian University)

CodeSegmentationConvolutional Neural NetworkContrastive LearningImageTextMultimodalityComputed Tomography

🎯 What it does: CausalCLIPSeg is proposed, an end-to-end multimodal medical image segmentation framework that utilizes CLIP's visual and text encoders and achieves pixel-level text-image alignment through cross-modal decoding, while introducing a causal intervention module to eliminate confounding bias.

CAVM: Conditional Autoregressive Vision Model for Contrast-Enhanced Brain Tumor MRI Synthesis

Gui, Lujun (Beijing Institute of Technology), Yan, Tianyi (Beijing Institute of Technology)

CodeSegmentationGenerationData SynthesisTransformerGenerative Adversarial NetworkImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: A framework based on a Conditional Autoregressive Visual Model (CAVM) is proposed to generate contrast-enhanced T1Gd images from non-contrast brain tumor MRI (T1w, T2w, FLAIR), achieving high-quality synthesis through a stepwise increasing dosage.

Centerline Boundary Dice Loss for Vascular Segmentation

Shi, Pengcheng (Harbin Institute of Technology (Shenzhen)), Ma, Ting (Harbin Institute of Technology at Shenzhen)

CodeSegmentationConvolutional Neural NetworkMultimodalityBiomedical DataMagnetic Resonance ImagingComputed Tomography

🎯 What it does: A centerline boundary-based Dice loss function (cbDice) is proposed to improve the topological preservation and geometric detail capture in vascular segmentation.

CheXtriev: Anatomy-Centered Representation for Case-Based Retrieval of Chest Radiographs

Akash R. J., Naren (International Institute of Information Technology Hyderabad), Sivaswamy, Jayanthi (International Institute of Information Technology Hyderabad)

CodeRetrievalGraph Neural NetworkTransformerImageBiomedical Data

🎯 What it does: The CheXtriev framework is proposed, utilizing a graph Transformer to extract features from 18 anatomical regions and achieve case retrieval of chest X-ray images.

Class and Region-Adaptive Constraints for Network Calibration

Murugesan, Balamurali (ETS Montreal), Dolz, Jose (CHU Sainte-Justine, University of Montreal)

CodeSegmentationOptimizationConvolutional Neural NetworkSupervised Fine-TuningImageBiomedical DataMagnetic Resonance ImagingComputed TomographyBenchmark

🎯 What it does: This paper proposes a Class and Region Adaptive Constraint (CRaC), which introduces class and region-specific penalty terms into the semantic segmentation network and adaptively learns these penalty weights during training using the Augmented Lagrangian Method, thereby achieving network calibration.

CLEFT: Language-Image Contrastive Learning with Efficient Large Language Model and Prompt Fine-Tuning

Du, Yuexi (Yale University), Dvornek, Nicha C. (Yale University)

CodeClassificationRepresentation LearningTransformerLarge Language ModelPrompt EngineeringContrastive LearningImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: Proposes the CLEFT framework, which combines pre-trained LLMs with parameter-efficient fine-tuning (PEFT) for contrastive learning in medical images, and enhances model generalization through context prompt learning after pre-training.

Clinical-grade Multi-Organ Pathology Report Generation for Multi-scale Whole Slide Images via a Semantically Guided Medical Text Foundation Model

Tan, Jing Wei (Korea University), Jeong, Won-Ki (Korea University)

CodeGenerationTransformerLarge Language ModelSupervised Fine-TuningImageText

🎯 What it does: A pathology report generation model PMPRG based on a multi-scale regional visual Transformer (MR-ViT) is proposed, which can automatically generate clinical-grade pathology reports for multiple organs and multiple slices.

CLIP-DR: Textual Knowledge-Guided Diabetic Retinopathy Grading with Ranking-aware Prompting

Yu, Qinkai (University of Exeter), Meng, Yanda (University Of Exeter)

CodeClassificationDomain AdaptationTransformerPrompt EngineeringVision Language ModelImageBenchmark

🎯 What it does: A CLIP-based framework for diabetic retinopathy (DR) grading, named CLIP-DR, is proposed, which utilizes text prompts to learn the natural order of DR grading and addresses the long-tail data problem.

Coarse-Grained Mask Regularization for Microvascular Obstruction Identification from non-contrast Cardiac Magnetic Resonance

Yan, Yige (Nanyang Technological University), Rajapakse, Jagath C. (Nanyang Technological University)

CodeRecognitionConvolutional Neural NetworkTransformerImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: A framework for identifying microvascular obstruction (MVO) based on contrast-free cardiac CINE MRI is proposed, enhancing model learning through coarse-grained mask regularization.

Coarse-to-Fine Latent Diffusion Model for Glaucoma Forecast on Sequential Fundus Images

Zhang, Yuhan (Chinese University of Hong Kong), Heng, Pheng-Ann (Chinese University of Hong Kong)

CodeGenerationData SynthesisTransformerDiffusion modelImageSequentialBiomedical Data

🎯 What it does: Based on continuous retinal fluorescence imaging, a coarse-to-fine latent diffusion model (C2F-LDM) is used to generate latent features for future time points, thereby predicting glaucoma risk and reconstructing future fundus images.

Comprehensive Generative Replay for Task-Incremental Segmentation with Concurrent Appearance and Semantic Forgetting

Li, Wei (Shanghai Jiao Tong University), Gu, Lixu (Shanghai Jiao Tong University)

CodeSegmentationGenerationDiffusion modelImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: A Task-Incremental Learning (TIL) framework called Comprehensive Generative Replay (CGR) is proposed, which generates image-mask pairs of past tasks at each step to recover lost appearance and semantic knowledge, while simultaneously updating the model to support future replay during new task learning.

Conditional Diffusion Model for Versatile Temporal Inpainting in 4D Cerebral CT Perfusion Imaging

Bae, Juyoung (Hong Kong University of Science and Technology), Chen, Hao (Harvard Medical School)

CodeRestorationData SynthesisConvolutional Neural NetworkDiffusion modelBiomedical DataComputed Tomography

🎯 What it does: Using conditional diffusion models for temporal interpolation of brain CT perfusion sequences to fill in missing scans at high temporal resolution, outputting synthetic scans at one-second intervals.

Conditional diffusion model with spatial attention and latent embedding for medical image segmentation

Hejrati, Behzad (Wayne State University), Dong, Ming (Wayne State University)

CodeSegmentationConvolutional Neural NetworkDiffusion modelImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: A conditional diffusion model cDAL is proposed, utilizing spatial attention and random latent embeddings for medical image segmentation.

Confidence intervals uncovered: Are we ready for real-world medical imaging AI?

Christodoulou, Evangelia (German Cancer Research Center), Maier-Hein, Lena (Dresden University of Technology)

CodeSegmentationBiomedical DataReview/Survey PaperBenchmark

🎯 What it does: The system evaluates and summarizes the reporting of performance variability in medical image segmentation models in MICCAI 2023 papers, and estimates the unreported standard deviations and confidence intervals using polynomial models.

Confidence Matters: Enhancing Medical Image Classification Through Uncertainty-Driven Contrastive Self-Distillation

Sharma, Saurabh (Indian Institute of Technology Patna), Chandra, Joydeep (Indian Institute of Technology Patna)

CodeClassificationKnowledge DistillationContrastive LearningImageBiomedical Data

🎯 What it does: A framework for uncertainty-driven contrastive self-distillation (UDCD) is proposed to address issues of data scarcity, class imbalance, and high intra-class variance in medical image classification.

Confidence-guided Semi-supervised Learning for Generalized Lesion Localization in X-ray Images

Das, Abhijit (Jio Institute), Roy, Sudipta (Inception Institute of AI)

CodeObject DetectionSegmentationKnowledge DistillationImageBiomedical Data

🎯 What it does: A confidence-guided semi-supervised framework called AnoMed is proposed for multi-scale lesion localization in chest X-rays, addressing the issues of pseudo-label inconsistency and small class bias by combining Scale-Invariant Bottleneck (SIB) and Confidence-Guided Pseudo-Label Optimizer (PLO).

Context-guided Continual Reinforcement Learning for Landmark Detection with Incomplete Data

Wan, Kaiwen (Fudan University), Zhuang, Xiahai (Fudan University)

CodeObject DetectionReinforcement LearningPrompt EngineeringImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: This paper studies a context-guided continuous reinforcement learning framework (CgCRL) for multi-point label detection in sequentially incomplete medical imaging data.

Contrast Representation Learning from Imaging Parameters for Magnetic Resonance Image Synthesis

Xiong, Honglin (ShanghaiTech University), Wang, Qian (United Imaging Intelligence)

CodeGenerationData SynthesisRepresentation LearningAuto EncoderContrastive LearningImageBiomedical DataMagnetic Resonance Imaging

🎯 What it does: A contrast representation learning (CRL) method based on imaging parameters is proposed for synthesizing images with different contrasts in magnetic resonance imaging.

Controllable Counterfactual Generation for Interpretable Medical Image Classification

Liu, Shiyu (Xi'an Jiaotong University), Ma, Jianhua (Pazhou Lab)

CodeClassificationGenerationData SynthesisExplainability and InterpretabilityDiffusion modelImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease

🎯 What it does: This paper proposes a controllable counterfactual generation method based on diffusion models to generate image differences between different pathological categories (such as Alzheimer's disease and healthy normals), thereby providing interpretable diagnostic results and performing data augmentation.

CoReEcho: Continuous Representation Learning for 2D+time Echocardiography Analysis

Maani, Fadillah Adamsyah (Mohamed bin Zayed University of Artificial Intelligence), Yaqub, Mohammad (Mohamed bin Zayed University of Artificial Intelligence)

CodeRepresentation LearningTransformerContrastive LearningVideoBiomedical DataUltrasound

🎯 What it does: Proposes the CoReEcho framework, achieving direct ejection fraction regression from 2D+time cardiac ultrasound videos through continuous representation learning.

Correlation-adaptive Multi-view CEUS Fusion for Liver Cancer Diagnosis

Wan, Peng (Nanjing University of Aeronautics and Astronautics), Zhang, Daoqiang (Nanjing University of Aeronautics and Astronautics)

CodeClassificationBiomedical DataUltrasound

🎯 What it does: A correlation adaptive method CAMVF based on multi-view CEUS fusion is proposed for the diagnosis of hepatocellular carcinoma and intrahepatic cholangiocarcinoma.

Cortical Surface Reconstruction from 2D MRI with Segmentation-Constrained Super-Resolution and Representation Learning

Wu, Wenxuan (South China University of Technology), Zhang, Xin (South China University of Technology)

CodeSegmentationSuper ResolutionRepresentation LearningConvolutional Neural NetworkTransformerImageBiomedical DataMagnetic Resonance ImagingOrdinary Differential Equation

🎯 What it does: A two-stage process for reconstructing the complete cortical surface from 2D low-resolution MRI is proposed, including super-resolution with segmentation constraints and surface reconstruction with feature alignment.

COVID19 to Pneumonia: Multi Region Lung Severity Classification using CNN Transformer Position-Aware Feature Encoding Network

Lee, Jong Bub (Inha University), Lee, Hyun Gyu (Inha University)

CodeClassificationConvolutional Neural NetworkTransformerImage

🎯 What it does: This paper proposes a multi-region lung severity classification method based on a CNN-Transformer hybrid network, utilizing position-aware feature encoding and region-shared MLP to achieve automatic scoring of COVID-19 and heterogeneous pneumonia images.

CriDiff: Criss-cross Injection Diffusion Framework via Generative Pre-train for Prostate Segmentation

Liu, Tingwei (Dalian University of Technology), Lu, Huchuan (Dalian University of Technology)

CodeSegmentationTransformerDiffusion modelImageBiomedical DataMagnetic Resonance ImagingUltrasound

🎯 What it does: A two-stage Criss-cross Injection Diffusion (CriDiff) framework is proposed, combining Boundary Enhancement (BEC) and Core Enhancement (CEC) conditioners with Generative Pre-training (GP) methods for prostate segmentation.

Cross Prompting Consistency with Segment Anything Model for Semi-supervised Medical Image Segmentation

Miao, Juzheng (Chinese University of Hong Kong), Heng, Pheng-Ann (Chinese University of Hong Kong)

CodeSegmentationSupervised Fine-TuningImageBiomedical DataMagnetic Resonance ImagingUltrasound

🎯 What it does: A cross-prompt consistency semi-supervised medical image segmentation method based on the Segment Anything model, CPC-SAM, has been developed.

Cross-Modality Cardiac Insight Transfer: A Contrastive Learning Approach to Enrich ECG with CMR Features

Ding, Zhengyao (Zhejiang University), Huang, Zhengxing (First Affiliated Hospital of Zhejiang University School of Medicine)

CodeClassificationRepresentation LearningTransformerAuto EncoderContrastive LearningMultimodalityTime SeriesBiomedical DataMagnetic Resonance ImagingElectrocardiogram

🎯 What it does: By using contrastive learning to transfer cardiac magnetic resonance (CMR) features to electrocardiogram (ECG) embeddings, ECG can more accurately predict cardiac structural indicators and disease states.

CryoSAM: Training-free CryoET Tomogram Segmentation with Foundation Models

Zhao, Yizhou (Carnegie Mellon University), Xu, Min (University of Alabama at Birmingham)

CodeSegmentationPrompt EngineeringContrastive LearningImage

🎯 What it does: A training-free CryoSAM framework is proposed to complete the segmentation of all particles in CryoET using single-point prompts.

CS3: Cascade SAM for Sperm Segmentation

Shi, Yi (Nanjing University), Zhan, De-Chuan (Nanjing University)

CodeSegmentationImage

🎯 What it does: This paper proposes an unsupervised CS3 algorithm that utilizes Cascade SAM to accurately segment overlapping sperm images through a three-stage progression: head separation, simple tail, and complex tail.

CT-based brain ventricle segmentation via diffusion Schrödinger Bridge without target domain ground truths

Teimouri, Reihaneh (Concordia University), Xiao, Yiming (Concordia University)

CodeSegmentationDiffusion modelImageBiomedical DataComputed Tomography

🎯 What it does: This study proposes a CT brain ventricle segmentation method based on the diffusion Schrâdinger bridge, which does not require target domain labeling.

CT2Rep: Automated Radiology Report Generation for 3D Medical Imaging

Hamamci, Ibrahim Ethem (University Of Zurich), Menze, Bjoern (University of Zurich)

CodeGenerationTransformerImageTextMultimodalityComputed Tomography

🎯 What it does: This paper presents CT2Rep, which achieves the automatic generation of radiology reports from 3D chest CT images for the first time.

Curriculum Prompting Foundation Models for Medical Image Segmentation

Zheng, Xiuqi (Beijing University of Posts and Telecommunications), Lao, Qicheng (Ningbo Fregty Optoelectronics Technology Co., Ltd)

CodeSegmentationPrompt EngineeringImageBiomedical DataUltrasound

🎯 What it does: This paper proposes a medical image segmentation framework based on curriculum-based prompts, utilizing self-generated multi-granularity prompts (boxes, points, masks) to achieve automated segmentation on the SAM model.

Customized Relationship Graph Neural Network for Brain Disorder Identification

Xia, Zhengwang (Nanjing University of Science and Technology), Lu, Jianfeng (Nanjing University of Science and Technology)

CodeClassificationRecognitionGraph Neural NetworkBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease

🎯 What it does: A CRGNN model is proposed that can jointly learn brain network structure and classification tasks for brain disease recognition.

Cut to the Mix: Simple Data Augmentation Outperforms Elaborate Ones in Limited Organ Segmentation Datasets

Liu, Chang (Friedrich-Alexander-Universitt Erlangen-Nrnberg), Maier, Andreas (Friedrich-Alexander-UniversitΓ€t Erlangen)

CodeSegmentationConvolutional Neural NetworkImageBiomedical DataComputed Tomography

🎯 What it does: A systematic evaluation and comparison of four algorithms based on image fusion and object-level augmentation (CutMix, CarveMix, ObjectAug, AnatoMix) was conducted on a limited multi-organ segmentation dataset, demonstrating that the simple CutMix is the most effective in improving segmentation accuracy.

CUTS: A Deep Learning and Topological Framework for Multigranular Unsupervised Medical Image Segmentation

Liu, Chen (Yale University), Krishnaswamy, Smita (Yale University)

CodeSegmentationConvolutional Neural NetworkContrastive LearningImageBiomedical DataMagnetic Resonance ImagingAlzheimer's Disease

🎯 What it does: An unsupervised multi-granularity medical image segmentation framework called CUTS is proposed, which generates embeddings using contrastive learning and reconstruction loss based on pixel-centered patches, and achieves multi-scale segmentation through diffusion condensation.

D-CoRP: Differentiable Connectivity Refinement for Functional Brain Networks

Hu, Haoyu, Li, Chao (University of Dundee)

CodeGraph Neural NetworkBiomedical DataMagnetic Resonance Imaging

🎯 What it does: D-CoRP is proposed, a differentiable functional brain network connectivity refinement plugin that can automatically remove noise or redundant edges while maintaining the overall structure.

Data Augmentation with Multi-armed Bandit on Image Deformations Improves Fluorescence Glioma Boundary Recognition

Xiao, Anqi (Institute of Automation, Chinese Academy of Sciences), Hu, Zhenhua (Institute of Automation, Chinese Academy of Sciences)

CodeRecognitionSegmentationData-Centric LearningConvolutional Neural NetworkReinforcement LearningImageMultimodalityMagnetic Resonance Imaging

🎯 What it does: This paper proposes an adaptive data augmentation method based on multi-armed bandits (MAB) called EEA, aimed at enhancing the generalization ability of deep learning models for brain glioma boundary recognition under low sample size NIR-II fluorescence imaging, and it is validated on three modalities of data: white light, NIR-I, and NIR-II.

Data-Driven Tissue- and Subject-Specific Elastic Regularization for Medical Image Registration

Reithmeir, Anna (Technical University of Munich), Zimmer, Veronika A. (Technical University of Munich)

CodeConvolutional Neural NetworkImageBiomedical DataMagnetic Resonance ImagingComputed Tomography

🎯 What it does: A medical image registration method based on hypernetworks is proposed, which can learn patient- and tissue-specific linear elastic regularization parameters without retraining.