Segment Anything in Medical Images

By Jun Ma, Bo Wang. arXiv.org, 2023

1,770 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/4dd869a2c17cacd03b3e8a9b7efcd48f40651925

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

ASAM: Anatomy-Encoded Segment Anything Model for Medical Images (2025, positive): “Currently, most medical foundation models [2, 6–10, To address the domain gap between natural and medical images, we propose an Anatomy-encoded SAM (ASAM).”
Personalize Segment Anything Model with One Shot (2023, neutral): “…SAM for high-quality segmentation (Ke et al., 2023), faster inference speed (Zhao et al., 2023; Zhang et al., 2023a), all-purpose matching (Liu et al., 2023), 3D reconstruction (Cen et al., 2023), object tracking (Yang et al., 2023), medical (Ma & Wang, 2023; Huang et al., 2023) image processing.”
Medical SAM Adapter: Adapting Segment Anything Model for Medical Image Segmentation (2023, methodology): “Considering SAM’s performance in 1, we observe that SAM’s zero-shot performance is generally inferior to that of fully-trained models (e.g., MedSAM (Ma and Wang 2023)) in the target medical image segmentation tasks, regardless of the prompt used.”
The Platonic Representation Hypothesis (2024, neutral): “…sometimes called foundation models (Bommasani et al., 2021), that support a large range of tasks, including robotics (Driess et al., 2023; Brohan et al., 2023), bioinformatics (Ma et al., 2024), and health-* Proceedings of the 41 st International Conference on Machine Learning , Vienna, Austria.”
Customized Segment Anything Model for Medical Image Segmentation (2023, background): “First, the large-scale CV models make the decisions of boundaries between different segmentation regions based on the variance of intensity [15,24], which is reasonable in natural images but not medical images because the analysis of anatomical or pathological structure plays a critical role in medi”
Segment Anything Model for medical image analysis: an experimental study (2023, background): “Some preprints already showed extensions of SAM to medical imaging [12,23] and one paper showed integration of SAM into 3D Slicer [22].”
Foundation Models Defining a New Era in Vision: A Survey and Outlook (2025, neutral): “Further, such a generic visual prompt-based segmentation model can be adapted for speciﬁc downstream tasks such as medical image segmentation [12], [13], video object segmentation [14], robotics [15], and remote sensing [16].”
On the Challenges and Perspectives of Foundation Models for Medical Image Analysis (2023, background): “SAM can be further tuned to achieve state-of-the-art performance by leveraging high-quality downstream data and performing proper fine-tuning strategies Ma and Wang [2023], Paranjape et al. [2023], Cui et al. [2023], adding adapters with specially designed architectures Wu et al. [2023a], Gong et…”
LangSplat: 3D Language Gaussian Splatting (2023, background): “[24] proposed Seal to explore the potential of VFMs including SAM for point cloud segmentation.”
RSPrompter: Learning to Prompt for Remote Sensing Instance Segmentation Based on Visual Foundation Model (2023, methodology): “This advancement opens new horizons for intel-1558-0644 ligent image analysis and understanding [35], [36], [37], [38], [39], [40], [41].”
EfficientSAM: Leveraged Masked Image Pretraining for Efficient Segment Anything (2023, background): “There are many other works attempting to generalize SAM to real-world scenarios, including medical image segmentation[41], camouflaged object detection[51], transparent object detection[24], concept-based explaination[50], semantic communitation[52], and helping people with visual impairments[37].”
SAM 2: Segment Anything in Images and Videos (2024, background): “The success of SAM led to its adoption in a wide range of applications, such as medical imaging (Ma et al., 2024; Deng et al., 2023; Mazurowski et al., 2023; Wu et al., 2023a), remote sensing (Chen et al., 2024; Ren et al., 2024), motion segmentation (Xie et al., 2024), and camouflaged object…”
Segment Anything Model for Medical Image Segmentation: Current Applications and Future Directions (2024, background): “MedSAM [13] is introduced for universal medical image segmentation, which adapts from SAM on an unprecedented scale by curating a diverse and comprehensive dataset containing more than one million medical image-mask pairs of 11 modalities.”