The DOI matches its Crossref record; 105 of the 105 references it cites were found in a registry or an index; 2 of the 31 sources its claims cite were checked for support.
Segment anything model for medical images?
Open accessAbstract
The Segment Anything Model (SAM) is the first foundation model for general image segmentation. It has achieved impressive results on various natural image segmentation tasks. However, medical image segmentation (MIS) is more challenging because of the complex modalities, fine anatomical structures, uncertain and complex object boundaries, and wide-range object scales. To fully validate SAM's performance on medical data, we collected and sorted 53 open-source datasets and built a large medical segmentation dataset with 18 modalities, 84 objects, 125 object-modality paired targets, 1050K 2D images, and 6033K masks. We comprehensively analyzed different models and strategies on the so-called COSMOS 1050K dataset. Our findings mainly include the following: (1) SAM showed remarkable performance in some specific objects but was unstable, imperfect, or even totally failed in other situations. (2) SAM with the large ViT-H showed better overall performance than that with the small ViT-B. (3) SAM performed better with manual hints, especially box, than the Everything mode. (4) SAM could help human annotation with high labeling quality and less time. (5) SAM was sensitive to the randomness in the center point and tight box prompts, and may suffer from a serious performance drop. (6) SAM performed better than interactive methods with one or a few points, but will be outpaced as the number of points increases. (7) SAM's performance correlated to different factors, including boundary complexity, intensity differences, etc. (8) Finetuning the SAM on specific medical tasks could improve its average DICE performance by 4.39% and 6.68% for ViT-B and ViT-H, respectively. Codes and models are available at: https://github.com/yuhoo0302/Segment-Anything-Model-for-Medical-Images. We hope that this comprehensive report can help researchers explore the potential of SAM applications in MIS, and guide how to appropriately use and develop SAM.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
The paper builds COSMOS 1050K, a large medical segmentation benchmark from 53 public datasets, and comprehensively evaluates the Segment Anything Model (SAM) across modalities, objects, prompt strategies, and model sizes. SAM is highly variable in medical image segmentation—strong on some targets but unstable or failing on others—while box prompts, larger ViT-H, and task-specific finetuning improve performance.
Key findings
- SAM’s performance was unstable across medical objects: remarkable for some, but imperfect or totally failed in others. Manual prompts outperformed Everything mode, and box prompts were best; adding a point to a box did not obviously help (ViT-B DICE 0.49% ↓, ViT-H DICE 0.06% ↓). Increasing point prompts improved DICE from 56.02% to 63.81% for ViT-B and from 56.78% to 71.61% for ViT-H.
- ViT-H generally outperformed ViT-B. Under Everything mode, ViT-H exceeded ViT-B by 7.47% DICE and had 10.61 pixels lower HD. Modality-wise, SAM achieved the highest average DICE on X-ray for both models; histopathology, colonoscopy, and DW MRI also reached >90% DICE, while 6 (ViT-B) and 7 (ViT-H) modalities had DICE between 80 and 90 but with large standard deviations.
- SAM was sensitive to prompt randomness: for one positive point, DICE dropped by 2.67%, 7.38%, and 14.62% as shifting increased; for one box, shifts of 20–30 pixels caused a 24.11% decrease, and one box plus one point caused a 29.93% decrease. In interactive comparison, SAM beat FocalClick and SimpleClick with one or a few points but was outpaced as points increased, performing worse at 10 points.
- Finetuning SAM on specific medical tasks improved average DICE by 4.39% for ViT-B and 6.68% for ViT-H; 32/45 objects improved for ViT-B and 37/45 for ViT-H. DICE showed moderate correlation with Fourier order and intensity difference, weak correlation with size, and no correlation with modality or aspect ratio.
Methods
- The authors collected and standardized 53 public datasets into COSMOS 1050K: 18 modalities, 84 objects, 125 object-modality paired targets, 1,050,311 2D images/slices (1,003,809 slices from 8,653 3D volumes and 46,502 standalone 2D images), and 6,033,198 masks.
- They evaluated SAM with ViT-B (91M parameters, 12 transformer layers) and ViT-H (636M parameters, 32 transformer layers) under six strategies: S1 Everything; S2 one positive point; S3 five positive points; S4 five positive and five negative points; S5 one box; S6 one box and one positive point. Metrics were DICE, JAC, and HD, with a mask-matching mechanism; Spearman partial correlation analyzed 191,779 anatomical structures for size, aspect ratio, intensity difference, modality, and boundary complexity via Elliptical Fourier Descriptors/Fourier order.
- Finetuning used 45 common objects with box prompts, frozen image and prompt encoders, and only the mask decoder adjusted for 20 epochs at learning rate 1e-4 and batch size 2. A separate annotation study involved three doctors, 100 images, 55 objects, 620 masks, and 9 modalities, and SAM was compared with FocalClick and SimpleClick.
Limitations
- The paper states that SAM is unstable in medical image segmentation, has weak point-based multi-round iteration capability, is sensitive to prompt randomness, cannot output object categories/semantics, and was trained on natural images without medical data. It also notes that finetuning decreased performance for objects with small numbers, RGB color coding, etc., requiring more careful task-specific finetuning.
- Evident from the text: the evaluation is 2D-only, with 3D/4D data converted to 2D slices; the dataset excludes extremely small objects, objects that become significantly separated across slices, and discrete structures such as some histopathology and lung trachea trees.
- Evident from the text: the main paper presents detailed performance for only 40 objects, with full results in supplementary materials, and the text notes that SAM’s text prompt had not been released in the official repository at the time.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
58 Own result · 24 Own interpretation · 19 No basis given · 18 Not checked yet · 5 Own method · 1 Partly supported · 1 Not supported · accuracy not yet measured
At least one cited source doesn’t support this claim or points to a different paper, and none confirmed it — see the verdicts below.
In the paper
“Most previous studies (Ji et al., 2023a; Mazurowski et al., 2023; Zhou et al., 2023) evaluated SAM with limited or even only one type of testing mode / strategy.”
What the sources say
- Not checked yetcites Ji et al., 2023a
- Disputedcites Mazurowski et al., 2023 sourceExperimental — accuracy not yet measured
Section 2.2 (Experiments) describes a thorough evaluation using non-iterative prompts, iterative prompts, and segment-everything mode, including five non-iterative prompting strategies. This paper therefore contradicts its use as an example of a study that evaluated SAM with limited or only one testing strategy. (LLM, conf=0.90)
Checked by a reasoning model — may be wrong.
- Not checked yetcites Zhou et al., 2023
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Thus, the correlation between SAM's segmentation performance and medical objects' attributes has not been conducted carefully (He et al., 2023; Mazurowski et al., 2023; Ma and Wang, 2023).”
What the sources say
- Not checked yetcites He et al., 2023
- Partly supportedcites Mazurowski et al., 2023 sourceExperimental — accuracy not yet measured
Section 3.5 examines the relationship between SAM performance and object size, so the paper does not support a blanket claim that correlations with object attributes have gone unexamined. It does acknowledge a limitation: its correlation analysis uses dataset-average object sizes and does not consider variation within datasets. The broader judgment that these correlations have not been studied carefully is not established by this paper. (LLM, conf=0.78)
- Not checked yetcites Ma and Wang, 2023
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“we collected and sorted 53 open-source datasets and built a large medical segmentation dataset with 18 modalities, 84 objects, 125 object-modality paired targets, 1050K 2D images, and 6033K masks.”
- Also stated in: § 1.2. SAM in Medical Image Analysis · p. 5 · plain
“we build a large medical image dataset named COSMOS 1050K , including 1050 K images with 18 different modalities (see Fig. 1) and 84 objects ( e.g., anatomical structures, lesions, cells, tools, etc.), to cover the entire body (see Fig. 2).”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 1.2. SAM in Medical Image Analysis · p. 5 · plain
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“we collected and sorted 53 open-source datasets and built a large medical segmentation dataset with 18 modalities, 84 objects, 125 object-modality paired targets, 1050K 2D images, and 6033K masks.”
- Also stated in: § 2.2. Dataset Preprocessing Specification · p. 7 · plain
“In total, COSMOS 1050K consists of 1,050,311 2D images or slices, with 1,003,809 slices originating from 8,653 3D volumes and 46,502 being standalone 2D images.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 2.2. Dataset Preprocessing Specification · p. 7 · plain
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“SAM was sensitive to the randomness in the center point and tight box prompts, and may suffer from a serious performance drop.”
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“SAM performed better than interactive methods with one or a few points, but will be outpaced as the number of points increases.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“SAM performed better with manual hints, especially box, than the Everything mode.”
- Also stated in: § 5. Conclusion · p. 19 · plain
“3) SAM performed better with manual hints, especially box, than the Everything mode.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 5. Conclusion · p. 19 · plain
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“SAM could help human annotation with high labeling quality and less time.”
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“SAM performed better than interactive methods with one or a few points, but will be outpaced as the number of points increases.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Finetuning the SAM on specific medical tasks could improve its average DICE performance by 4.39% and 6.68% for ViT-B and ViT-H , respectively.”
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- It may state two separate claims as one. 25
- Probably not a substantial claim. 17
- The quote only partly states this claim. 1
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“SAM showed remarkable performance in some specific objects but was unstable, imperfect, or even totally failed in other situations.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In total, COSMOS 1050K consists of 1,050,311 2D images or slices, with 1,003,809 slices originating from 8,653 3D volumes and 46,502 being standalone 2D images.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In total, COSMOS 1050K consists of 1,050,311 2D images or slices, with 1,003,809 slices originating from 8,653 3D volumes and 46,502 being standalone 2D images.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“While for the manual prompt encoding and mask decoding ( S 2 -S 6), the reference times of different models and strategies are similar and less than 0.01s.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“There are 6 ( ViTB ) and 7 ( ViT-H ) modalities with DICE performance between 80 and 90, but most of them have large standard deviations.”
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The ground-truth masks used for evaluation accurately represent the true segmentation targets.
- Rests on an assumption: The selected datasets and anatomical structures are representative of medical images, so the evaluation results can provide general guidance.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“See BrainTumor in Table 2 and 3, its performance ranges from 71.76% to 77.74% in DICE and 13.02 pixels to 15.35 pixels in HD.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The ground-truth masks used for evaluation accurately represent the true segmentation targets.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“SAM performs worse in the remaining modalities, and the results are also more unstable.”
plaincorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The ground-truth masks used for evaluation accurately represent the true segmentation targets.
- Rests on an assumption: The selected datasets and anatomical structures are representative of medical images, so the evaluation results can provide general guidance.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Accordingly, the image embeddings could be reused to relieve the computational burden of the inference pipeline. Thereby, the overall efficiency of the SAM testing could be improved by nearly n times.”
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The image embeddings computed once can be reused across different testing strategies for the same image without affecting segmentation results.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“We observed that the process of image embedding extraction was time-consuming.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“SAM with a box prompt yields the best performance, while adding one point to the box will not bring obvious changes (ViT-B: DICE 0.49% ↓ , ViT-H: DICE 0.06% ↓ ).”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The ground-truth masks used for evaluation accurately represent the true segmentation targets.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“SAM generated multiple binary masks for each input image, but not all of them included the corresponding object.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“When different testing strategies were applied, only the corresponding npy files needed to be loaded, which significantly improved the testing efficiency (approximately n × ).”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
- Rests on an assumption: The official GitHub implementation of SAM is correct and faithful.
- Rests on an assumption: The image embeddings computed once can be reused across different testing strategies for the same image without affecting segmentation results.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For the point prompts ( S 2 -S 4), adding more points will bring stable performance improvement (ViTB: DICE from 56.02% to 63.81%, ViT-H: DICE from 56.78% to 71.61%).”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The ground-truth masks used for evaluation accurately represent the true segmentation targets.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It can be observed from Fig. 6 that, under Everything mode ( S 1 ), ViT-H exceeds ViT-B with 7.47% at DICE and is 10.61 pixels lower than ViT-B on HD.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The ground-truth masks used for evaluation accurately represent the true segmentation targets.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The performance of objects with small numbers, RGB color coding, etc., decreases.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“For S 2 (single point), the DICE performance dropped by 2.67%, 7.38%, and 14.62% as the shifting level increased.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The quote only partly states this claim.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Our study also preliminarily verified that SAM can greatly shorten the annotation time and improve the annotation quality.”
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“With the increase in the number of point prompts ( S 3 and S 4), the decline of DICE could be alleviated, and the stability of the model could be improved.”
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Based on the DICE results, our conclusion is that: 1) SAM outperformed FocalClick and SimpleClick in the first interaction with a single point; 2) As iterations progressed, SAM's performance increased slowly, or even declined, while the performance of the interactive method could be improved steadily; 3) Using 10 points, SAM performed worse than the interactive methods.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The selected 10 typical organs/tumors are representative of medical segmentation tasks for comparing SAM with interactive methods.
In the paper
“Notably, the text prompt has not been released yet on the official GitHub repository 4 .”
What the sources say
- Not checked yetcites [4] Bernal, 2015, WM-DOVA maps for accurate polyp highlighting in colonoscopy: Validation vs. saliency maps from physicians, Comput. Med. Imaging Graph. source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“Thus, we conducted a fair comparison by limiting the image size to 256 × 256 and setting the number of targets to 1.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“Retain slices with the sum of the pixel values of their labels greater than 50 for any 3D image and label volumes. This ensures that each slice has the corresponding correct label.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Since the box actually tells the exact location of the target and also its potential intensity features given the limited region. However, points only represent the part features of the target, which can lead to confusion.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
- Rests on an assumption: The ground-truth masks used for evaluation accurately represent the true segmentation targets.
In the paper
“As the Fourier order (FO) increases, the contour decoded from the Fourier series gets closer to the original contour (see Fig. 12)”
What the sources say
- Not checked yetcites Kuhl and Giardina, 1982
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“While evaluating SAM in medical images, one study used six medical datasets and found that there is no obvious winner among ViT-B , ViT-L , and ViT-H (Mattjie et al., 2023).”
What the sources say
- Not checked yetcites Mattjie et al., 2023
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“In Kirillov et al. (2023), ViT-H shows substantial performance improvements over ViT-B . However, the former requires multiplied testing time due to its increased complexity.”
What the sources say
- Not checked yetcites Kirillov et al. (2023)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Compared to the point prompt, the box prompt contains more region information about the object. Thus, it can better guide SAM with different models to achieve better segmentation performance.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
It may state two separate claims as one.
- Rests on an assumption: The ground-truth masks used for evaluation accurately represent the true segmentation targets.
In the paper
“It is noted that SAM can only find multiple targets in the image, without outputting their detailed categories (i.e., single-label: object or not).”
What the sources say
- Not checked yetcites Kirillov et al. (2023)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
In the paper
“SAM diverges from traditional segmentation frameworks by introducing a novel promptable segmentation task , which is supported by a flexible prompting-enabled model architecture and vast and diverse sources of training data .”
What the sources say
- Not checked yetcites Kirillov et al. (2023)
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“The above strategy can ensure the repeatability of the experiments.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“We further established a unified rule for point selection to ensure randomness, repeatability, and accuracy.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“Thus, for a fair comparison, we test the one-round interaction performance of SAM under the above five prompt strategies ( S 2 -S 6 ).”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This can simply be extended to other multi-strategy testing scenarios of SAM.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
- Rests on an assumption: The image embeddings computed once can be reused across different testing strategies for the same image without affecting segmentation results.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“1) SAM showed remarkable performance in some specific objects but was unstable, imperfect, or even totally failed in other situations.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
In the paper
“They shorten the annotation time to three weeks with the help of multiple pre-trained segmentation models for generating pseudo labels, and other useful strategies.”
What the sources say
- Not checked yetcites Qu et al., 2023
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“However, most medical images are gray-scale, with unclear and complex object boundaries, similar back- and fore-ground, and wide-range image size (especially containing some very small objects).”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
It may state two separate claims as one.
In the paper
“The training dataset of SAM, i.e., SA-1B 5 contains 11 million photos, including natural locations, objects, and scenes, but without any medical images.”
What the sources say
- Not checked yetcites [5] Bernard, 2018, Deep learning techniques for automatic MRI cardiac multi-structures segmentation and diagnosis: is the problem solved?, IEEE Trans. Med. Imaging source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It means that a single SAM network can be used to achieve the annotation of large-scale multi-modal, multi-class medical datasets, rather than using multiple task-specific models.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper may give no reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Finally, we believe that, although SAM has the potential to become a general MIS model, its performance in the MIS task is not stable at present.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
It may state two separate claims as one.
The paper relies on this without support.
Different medical objects often exhibit different characteristics and thus may have their own suitable modes for testing.
“However, different medical objects often exhibit different characteristics and thus may have their own suitable modes for testing.”
The paper relies on this without support.
The ground-truth masks used for evaluation accurately represent the true segmentation targets.
“we first calculated the center of mass of the ground truth (GT) mask”
“we calculated a set of dice scores { DICE n } N n = 1 between N binary predicted masks { P n } N n = 1 and the GT G .”
The paper relies on this without support.
The selected datasets and anatomical structures are representative of medical images, so the evaluation results can provide general guidance.
“We hope that a comprehensive evaluation of models of different sizes on large COSMOS 1050K can provide more inspirations for researchers.”
“we recorded the size, aspect ratio, intensity difference between foreground and background, modality, and boundary complexity of 191,779 anatomical structures.”
The paper relies on this without support.
The official GitHub implementation of SAM is correct and faithful.
“we implemented the testing pipeline of SAM basically following the official GitHub repository 4 .”
The paper relies on this without support.
The image embeddings computed once can be reused across different testing strategies for the same image without affecting segmentation results.
“since the same embedding could be reused for different testing strategies, we sought to optimize and accelerate this multi-extraction process.”
The paper relies on this without support.
The center of mass and tight bounding box of the ground truth represent the theoretical optimum prompts for SAM.
“Besides, we tend to test the theoretical optimum performance of SAM via the selection of the center of mass and tight box. Since they may include the most representative features of the target.”
The paper relies on this without support.
The 100 images randomly sampled from COSMOS 1050K with average DICE performance are representative of the broader medical image dataset for evaluating annotation time and quality.
“We randomly sampled 100 images with average DICE performance from COSMOS 1050K , to construct an evaluation subset comprising 55 objects and 620 masks across 9 modalities, including instances of the same object in different modalities.”
The paper relies on this without support.
The three doctors with 10 years of experience are representative of human annotators for evaluating annotation time and quality.
“We then invited three doctors with 10 years of experience to evaluate whether SAM's prediction under box prompts could improve annotation speed and quality.”
The paper relies on this without support.
The five attributes (size, aspect ratio, intensity difference, modality, Fourier order) adequately capture the factors that affect SAM's segmentation performance.
“We conducted an analysis of the partial correlation between the five attributes of the target objects mentioned above and the DICE score using Spearman's rank partial correlation coefficient under different testing strategies.”
The paper relies on this without support.
The Spearman partial correlation analysis assumptions (e.g., monotonic relationships, no unmeasured confounders) hold for the data.
“We conducted an analysis of the partial correlation between the five attributes of the target objects mentioned above and the DICE score using Spearman's rank partial correlation coefficient under different testing strategies.”
The paper relies on this without support.
The selected 10 typical organs/tumors are representative of medical segmentation tasks for comparing SAM with interactive methods.
“We selected 10 typical organs / tumors, covering various modalities, shapes, sizes, and intensity distributions.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“For example, the concurrent studies22,23 conducted a comprehensive assessment of SAM across a diverse array of medical images, underscoring that SAM achieved satisfactory segmentation outcomes primarily on targets characterized by distinct boundaries. However, the model exhibited substantial limitations in segmenting typical medical targets with weak boundaries or low contrast.”
From Segment anything in medical images (2024) · cites this paper as [23]
Citations
Cites
Learning Transferable Visual Models From Natural Language Supervision
Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al. (2021) · arXiv · doi:10.48550/arxiv.2103.00020 · Cited by 15,293
Measures of the Amount of Ecologic Association Between Species
Dice, L. R. (1945) · Ecology · doi:10.2307/1932409 · Cited by 12,187
Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A. C., Lo, W.-Y., et al. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00371 · Cited by 11,953
nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation
Isensee, F., Jaeger, P. F., Kohl, S. A. A., Petersen, J., & Maier-Hein, K. H. (2021) · Nature Methods · doi:10.1038/s41592-020-01008-z · Cited by 10,477 · cited 2×
“We describe the details of COSMOS 1050K in the following two aspects, including image collection and preprocessing specification.”
Masked Autoencoders Are Scalable Vision Learners
He, K., Chen, X., Xie, S., Li, Y., Dollar, P., & Girshick, R. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52688.2022.01553 · Cited by 8,489 · cited 1×
“Subsequently, a filtering mechanism is applied to remove duplicate and low-quality masks using the techniques of confidence score, stability evaluation based on threshold jitter, and non-maximal suppression (NMS).”
The multimodal brain tumor image segmentation benchmark (BRATS)
Menze, B. H., Jakab, A., Bauer, S., Kalpathy-Cramer, J., Farahani, K., Kirby, J., Burren, Y., Porz, N., Slotboom, J., Wiest, R., et al. (2015) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2014.2377694 · Cited by 6,781 · cited 1×
“AbdomenCT-1K [22] Liver, kidney, spleen and pancreas CT ACDC [23] Left and right ventricle and left ventricular myocardium cine MRI AMOS 2022 [24] Abdominal multi-organ segmentation CT, MRI AutoLaparo [25] Integrated dataset with multiple image-based perception tasks Colonoscopy BrainPTM 2021 [26] white matter tracts…”
Tschandl, P., Rosendahl, C., & Kittler, H. (2018) · Scientific Data · doi:10.1038/sdata.2018.161 · Cited by 3,628
Bakas, S., Akbari, H., Sotiras, A., Bilello, M., Rozycki, M., Kirby, J. S., Freymann, J. B., Farahani, K., & Davatzikos, C. (2017) · Scientific Data · doi:10.1038/sdata.2017.117 · Cited by 3,049
Segment anything in medical images
Ma, J., He, Y., Li, F., Han, L., You, C., & Wang, B. (2024) · Nature Communications · doi:10.1038/s41467-024-44824-z · Cited by 2,913
Codella, N. C. F., Gutman, D., Celebi, M. E., Helba, B., Marchetti, M. A., Dusza, S. W., Kalloo, A., Liopyris, K., Mishra, N., Kittler, H., et al. (2018) · 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018) · doi:10.1109/isbi.2018.8363547 · Cited by 2,620
Research works printed but not linked: 3
- Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al., 2020. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. In: International Conference on Learning Representations.
- Huazhu, 2019, PALM: PAthoLogic myopia challenge, Comput. Vis. Med. Imaging
- Lee, S., Shim, H., Park, S.H., Yun, I.D., Lee, S.U., 2010. Learning local shape and appearance for segmentation of knee cartilage in 3D MRI. In: Medical Image Analysis for the Clinic: A Grand Challenge. in Proceedings of the 13th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2010), Beijing, China. pp. 231–240.
Cited by
Sorted by citations
Items 401–500 of 511
Guo, J., Chen, G., Sun, Y., Wu, H., & Qin, J. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3754876
Cárdenas, M.-I., Tabash, S. C., Sandoval, P. M., et al. (2025) · IEEE Central America and Panama Student Conference (CONESCAPAN) · doi:10.1109/conescapan67936.2025.11229542
SISAM: An Enhanced Promptable Segmentation Model with Expanded Dataset for Surgical Instruments
Liu, Y., & Xiao, J. (2025) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn64981.2025.11227665
Polyp-GSAM: A Graph Attention-Guided Hybrid SAM Architecture Model for Precise Polyp Segmentation
Xie, K., Wang, B., Liu, S., et al. (2025) · doi:10.1109/ijcnn64981.2025.11228774
Self-citationMartinc, M., Manojlovska, A., & Džeroski, S. (2025) · European Workshop on Visual Information Processing · doi:10.1109/euvip66349.2025.11238970
Universal medical image segmentation via in-context cross-attention
Ciusdel, C. F., Serban, A., & Passerini, T. (2025) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2025.1698324
Domain-randomised instance-segmentation benchmark for soot in PIV images
Jose, B., Geigle, K. P., & Hampp, F. (2025) · Machine Learning Science and Technology · doi:10.1088/2632-2153/ae2565
Quaternion-based inter-channel correlation learning for fine-grained medical image segmentation
Zhan, B., Song, E., & Liu, H. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.132348
Löwenstein, K., Rehrl, J., Schuster, A., et al. (2025) · Computer Methods and Programs in Biomedicine Update · doi:10.1016/j.cmpbup.2025.100224
A brain-to-population graph learning framework for diagnosing brain disorders
Liao, Q., Cai, W., Sun, H., Liu, D., Chen, D., Yao, D., & Guo, D. (2025) · Neural Networks · doi:10.1016/j.neunet.2025.108526 · Cited by 3
Synthetic data enables human-grade microtubule analysis with foundation models for segmentationPreprint
Koddenbrock, M., Westerhoff, J., Fachet, D., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.01.09.698597
DBNet: A Dual-Branch Network with Boundary Awareness for Few-Shot 3D Medical Image Segmentation
Li, Y., He, H., Wang, X., Zhang, Z., He, W., & Shi, W. (2025) · IEEE International Conference on Parallel and Distributed Systems (ICPADS) · doi:10.1109/icpads67057.2025.11323007
Bao, G., Xia, W., Wang, H., Liao, Z., Wu, T., & Zhou, Y. H. (2026) · Sensors · doi:10.3390/s26020656
CSP-SAM: CNN-Enhanced and Self-prompting SAM for Ultrasound Anatomical Structure Segmentation
Yin, C., Dong, X., Tan, Y., Liang, B., Pu, B., & Jin, Z. (2026) · Lecture notes in computer science · doi:10.1007/978-981-95-5634-2_7
Juan-Ferrer, P., Pérez-Sánchez, N., Pla, F., Mollineda, R. A., Roselló-Sastre, E., & Tajahuerce, E. (2026) · BMC Medical Imaging · doi:10.1186/s12880-026-02178-6 · Cited by 1
SAM-BrainFocus: A Multi-scale Boundary-Refined Framework for Precise Brain Tumor Segmentation
Chen, Y., Zhang, Z., & Dong, Y. (2025) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc58881.2025.11343591
Huang, L., Zhao, Z., Sun, Y., Wang, X., Yang, Q., Xie, X., Hou, W., Guo, C., & Wang, Y. (2025) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm66473.2025.11356022
Brain Tumor Segmentation and Classification Using Multi-Scale SAM and VGG16
Raza, T., Haq, M. A. U., Naqvi, S. Q., Ramzan, H. A., Rehman, A., & Ramzan, S. (2025) · International Conference on Digital Futures and Transformative Technologies (ICoDT2) · doi:10.1109/icodt269104.2025.11360712 · Cited by 1
Chen, J., Hu, M., Safari, M., et al. (2026) · Physics in Medicine and Biology · doi:10.1088/1361-6560/ae4287
Self-citationPractical and Flexible Backdoor Attack Against Deep Learning Models via Shell Code Injection
Wang, J., Liu, H., Tao, R., Sun, J., Liu, X., & Zhao, Y. (2026) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2026.3662587
Depth-guided multi-modal early optimization for 6D pose estimation
Li, Y., & Lyu, E. (2026) · Neurocomputing · doi:10.1016/j.neucom.2026.134802
Liu, J., Zhou, Y., Hao, R., et al. (2026) · Bioengineering · doi:10.3390/bioengineering13020217
“The fundamental differences in imaging modalities, contrast distribution, and feature spaces often result in unstable segmentation across medical datasets [6].”
Yin, C., Wang, L., Dong, X., & Jin, Z. (2026) · Lecture notes in computer science · doi:10.1007/978-981-95-6123-0_38
SAM-SPJunc: Self-Prompting for Junction Detection in Retinal Images via Radius-Based Representations
Attari, M., Palaniappan, K., & Bunyak, F. (2025) · IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) · doi:10.1109/iccvw69036.2025.00070
Yao, X., Lai, Y., Liu, X., et al. (2026) · Expert Systems with Applications · doi:10.1016/j.eswa.2026.131758
QuEST: Quantization-Conditioned Efficient Stealthy Trojan
Lu, L., Pang, S., Wang, J., Gu, X., Liu, Y., Liu, X., & Zhou, Y. (2026) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2026.3671079 · Cited by 3
Parameter-efficient Finetuning of Foundational Models for Text-guided X-ray Image Segmentation
Alikarrar, M., Syben, C., Scheuplein, J., et al. (2026) · Informatik aktuell · doi:10.1007/978-3-658-51100-5_10
Zhang, J., Liu, Y., Jiang, Y., et al. (2026) · Physics in Medicine and Biology · doi:10.1088/1361-6560/ae54f9
Encrypt Anything: A Content‐Aware Hierarchical Privacy Protection Method for Image Data
Han, J., Wu, B., Xu, Y., Han, P., Wang, B., & Gu, Y. (2026) · IET Computer Vision · doi:10.1049/cvi2.70057
SAM-guided semi-supervised breast lesion segmentation in ultrasound videos with a new dataset
Chen, L., Zheng, Q., Chen, Y., Lv, F., & Wang, Q. (2026) · Expert Systems with Applications · doi:10.1016/j.eswa.2026.132141
Research on Computer Vision-Based Parameter Detection Methods for Table Tennis Balls
Liu, B. (2025) · IEEE International Conference on Electrical, Automation and Computer Engineering (ICEACE) · doi:10.1109/iceace67491.2025.11439225
TaP-SAM: Text-and-Point Guided SAM for Nuclei Instance Segmentation
Li, K., Li, P., & Wei, M. (2025) · International Conference on Cyberworlds (CW) · doi:10.1109/cw68232.2025.00065
Meng, J., Zhang, X., Hou, J., Qi, S., Ma, F., Wang, Z., & Li, D. (2026) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2026.114747
Fiedler, L., Howard, I., & Beyerer, J. (2026) · at - Automatisierungstechnik · doi:10.1515/auto-2025-0082
Yang, Z., Liang, S., Liu, H., & Gu, Y. (2026) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2026.114686
Multi-sequence parotid gland lesion segmentation via expert text-guided segment anything model
Wu, Z., Ren, C.-X., Wang, Y., et al. (2026) · Expert Systems with Applications · doi:10.1016/j.eswa.2026.132368
Wang, H., Cui, L., Ran, Q., Gao, M., Dai, X., & Na, L. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.110253
PromptReg: Interactive Registration by “Corresponding Prompts” for Segment Anything Model (SAM)
Huang, S., Xu, T., Li, J., Saeed, S. U., Shen, Z., Barratt, D. C., & Hu, Y. (2026) · IEEE Transactions on Image Processing · doi:10.1109/tip.2026.3683270
“Despite these developments, SAM exhibits variable sensitivity to medical images compared to natural images, as noted by [45] and [46], occasionally delivering unexpected segmentation results for certain anatomical and pathological structures without additional training.”
LLM-Guided SAM for Left Ventricle Segmentation and Functional Analysis in Echocardiography
Deng, X., Lu, Y., Zhang, W., et al. (2026) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp55912.2026.11464908
Verma, P., Van, M.-H., & Wu, X. (2026) · npj Computational Materials · doi:10.1038/s41524-026-02069-y
“Exploring the capability of SAM on medical data, Huang et al. 25 utilized 53 open-source datasets including 18 modalities, 84 objects, and more than 1 million 2D images, observing impressive performance in some cases but dismal in others.”
Lv, H., Jiang, S., Bai, Y., et al. (2026) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2026.104789
Liu, F., Yang, J., Zhang, Y., Xing, H., Yang, J., Lü, T., Wu, F., & Deng, L. (2026) · Scientific Reports · doi:10.1038/s41598-026-51220-8
“…this advancement has catalyzed progress in segmentation networks, SAM’s success extended to various medical images applications 21 , exemplified by MedSAM 8 , a foundational model designed for universal medical images segmentation 22,23 that addresses cross-object generalizability challenges.”
Nnadozie, E. C., Merino‐Caviedes, S., Luis-Román, D. A. D., et al. (2026) · Big Data and Cognitive Computing · doi:10.3390/bdcc10050142
“Mattjie et al. [33], Mazurowski et al. [34], Huang et al. [35] evaluated the robustness of SAM for medical image segmentations in zero-shot mode, where SAM performed better when provided with bounding box prompts.”
E-SAM: Training-Free Segment Every Entity Model
Zhang, W., Xiao, D., Chen, L., & Wang, L. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.02289
“This versatility has allowed SAM to find applications in a wide range of domains, including object tracking [5, 32, 41, 46], 3D instance segmentation [3, 42, 43], and medical imaging [15, 25, 27, 45].”
Lin, W., Zhong, J., Gao, Z., et al. (2026) · Medical Image Analysis · doi:10.1016/j.media.2026.104094
Self-citationSynthetic data enables human-grade microtubule analysis with foundation models for segmentation
Koddenbrock, M., Westerhoff, J., Fachet, D., Reber, S., Gers, F. A., & Rodner, E. (2026) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1013901
Introduction to AI and Its Applications
Grau, V. (2026) · Imaging informatics for healthcare professionals · doi:10.1007/978-3-032-15693-8_2
Self-citationDecoupling Shape and Texture in SAM-2 via Controlled Texture Replacement
Cohen, I., Meivar, B., Tu, P., et al. (2026) · IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) · doi:10.1109/wacv61042.2026.00397
“The latest advancement using Transformers for this task is the Segment Anything Model (SAM) [21], which presents a significant step forward while also having limitations in professional data, especially in textural and medical imaging [14,17].”
Baldeon-Calisto, M., Pérez, N., & Maldonado, C. B. G. (2026) · Physics in Medicine and Biology · doi:10.1088/1361-6560/ae6af8
“Huang et al (2024) evaluated the performance of SAM on 53 datasets with 18 modalities and 84 objects.”
MedSAM-guided geometry-aware 2D-3D feature fusion for medical image registration
Gao, F., He, Y., Jiang, H., Yu, P., Du, S., Liu, C., Li, S., & Hao, A. (2026) · Neural Networks · doi:10.1016/j.neunet.2026.109066
Promptable image segmentation: a survey of guided input techniques
Nejabat, H., D’Asaro, F., Pecora, A. E., Monopoli, T., & Bottino, A. (2026) · Foundations and Trends® in Computer Graphics and Vision · doi:10.1108/ftcgv-03-2026-001
Zhao, L., Zhang, Y., Ji, L., et al. (2026) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2026.3692645
“Recently, Segment Anything Model (SAM) [13] has been widely applied in medical tasks [14], [15], [16], [17] by the excellent image delineation performance, where points, boxes, and masks are adopted as prompts to interactively guide the delineation.”
Shape-Aware Multi-Task Instance Segmentation for Tubules in Renal Histology
Sikouky, M., Toni, G., Ponzio, F., et al. (2026) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi61048.2026.11515720
“Recent foundation models such as Med-SAM [6] extend this paradigm to large-scale, prompt-driven segmentation, but remain strongly texture-biased and struggle to encode shape, geometry, or topology [7].”
Fine-Tune Segment Anything Model (SAM) on Custom Data
Himabindu, B. A., Tryphosa, M., Bharathi, C., et al. (2026) · International Conference on Artificial Intelligence and Machine Learning Applications Theme: Healthcare and Internet of Things (AIMLA) · doi:10.1109/aimla67915.2026.11522461
“Segment Anything Model (SAM) has established a new standard in universal image segmentation, but in many situations, it is seen to decrease its performance when it is presented with images specific to the domain, including medical scans or data on industrial defects[1].”
Multi-sensor fusion on autonomous driving perception: A comprehensive survey
Jiang, M., Chu, Z., Pan, Y., Yuan, H., Han, Q., Zou, J., Zhang, Y., Zhu, X., Zhang, C., Zhou, X., et al. (2026) · Multimodal Transportation · doi:10.1016/j.multra.2026.100325 · Cited by 1
Self-citationAbdominal body composition assessment with 2D–3D Hybrid Mean‐Teacher Network
Hu, P., Zhu, Y., Hu, J., Li, X., Yu, T., Zhou, T., Liang, T., & Li, J. (2026) · Medical Physics · doi:10.1002/mp.70485
“Neverthe-less,emerging studies 38 indicated that few-shot learning or fine-tuning with training data is needed to get comparable performance with traditional CNN models for specific tasks.”
YO-SAM: YOLO-Guided Segment Anything Model
Gazawy, Q., Bachri, M., & Shami, A. A. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-26211-0_7
DB-SAM: Boundary-aware and domain-adaptive SAM for generalizable few-shot medical image segmentation
Wen, Y., Li, Y., Zhang, Z., Miao, Y., Shi, W., & Jiang, Z. (2026) · Journal of King Saud University - Computer and Information Sciences · doi:10.1007/s44443-026-00873-6
“Moreover, universal image features are captured by SAM through a self-supervised pre-training strategy, providing a foundation for its transferability and generalization to natural images (Mazurowski et al. 2023; Huang et al. 2024; Wu et al. 2025; Cheng et al. 2023).”
Bo, W., He, A., Xue, T., Zhang, Y., Xiao, Y., Liu, S., & Zhou, S. K. (2026) · Patterns · doi:10.1016/j.patter.2026.101570
Self-citationSu, J., Zhu, T., Cai, J., Wang, Z., Li, Q., Zhang, X., Tong, T., Wang, S., & Liu, W. (2026) · Medical Image Analysis · doi:10.1016/j.media.2026.104144 · Cited by 1
“However, early investigations (e.g., (Deng et al., 2025) and (Huang et al., 2024)) revealed that while SAM performs well on certain tasks, it faces significant limitations in medical imaging tasks such as nucleus, organ, and lesion segmentation.”
Medical Image Segmentation in Limited-Label Environments Using a Barlow Twins-Based VAE
Park, M., Park, C., Ryu, H., & Park, S. (2026) · Journal of Digital Contents Society · doi:10.9728/dcs.2026.27.5.1391
Lu, C.-Y., Tu, C.-H., Hsieh, C.-Y., Feng, C.-K., & Hsieh, I.-Y. L. (2026) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2026.3701624
“Nonetheless, SAM’s performance tends to degrade for structures with weak boundaries or low contrast [50], [51], which are characteristic of spinal radiographs.”
Diana-Albelda, C., Alcover‐Couso, R., García‐Martín, Á., Bescós, J., & Escudero-Viñolo, M. (2026) · International Journal of Imaging Systems and Technology · doi:10.1002/ima.70389 · Cited by 3
“Challenges such as low contrast, poorly defined boundaries and small lesion regions limit its e ff ectiveness in this context [11, 13].”
de Carvalho, O., de Carvalho, O. A., de Albuquerque, A. O., & Silva, D. G. (2026) · Remote Sensing · doi:10.3390/rs18172973
Chen, L., Zheng, Q., Guo, Y., et al. (2026) · Computerized Medical Imaging and Graphics · doi:10.1016/j.compmedimag.2026.102790
SAM-E 3 : An Edge-Aware and Efficient Enhancement Framework for Ultrasound Image Segmentation
Liu, L., Li, Y., Hou, H., et al. (2026) · International Seminar on Artificial Intelligence, Networking and Information Technology (AINIT) · doi:10.1109/ainit70033.2026.11557951
Self-citationKolmogorov Arnold UNet with channel aware encoding for white blood cell segmentation
Le, H. V., & Huynh, H. T. (2026) · Discover Artificial Intelligence · doi:10.1007/s44163-026-01581-w
“In parallel, foundation-model-based segmentation has emerged with the introduction of the Segment Anything Model (SAM) [19], which enables prompt-driven segmentation across diverse domains.”
Impact on Cost and Expert Time of Data-Efficient Deep Learning for Medical Image Segmentation
Jaiswal, A., Rinneburger, M., Meyer, F. I., Arjune, S., Oberlinkels, L., Matos, P. A. W., Lotter-Becker, L., Akünal, Ü., Bujotzek, M., Denner, S., et al. (2026) · Radiology Artificial Intelligence · doi:10.1148/ryai.250200 · Cited by 1
Sugawara, H., Takada, A., & Kato, S. (2026) · Journal of Imaging Informatics in Medicine · doi:10.1007/s10278-026-02076-x
Li, H., Chen, K., Lei, Z., et al. (2026) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2026.3709903
“Research by [13] and [27] reveals that SAM is highly sensitive to prompt uncertainty, particularly under box prompt settings, where performance degradation becomes increasingly severe as the shift magnitude increases.”
Language-Guided Segmentation of Medical Images: A Review of Foundation Models
Qamar, S. (2026) · Bioengineering · doi:10.3390/bioengineering13070803 · Cited by 1
“…and compute; PEFT supports many tasks from one base model and suits multi-task, cross-institutional, and continual deployment, since adapters can be switched or added without disturbing the frozen base and federated learning preserves privacy [8,75]; and prompt engineering suits rapid prototyping.”
Image-based geometric fidelity assessment of 3D food printed inks and designs
Habib, A. K. M. A., Khalil, M. I., Christopher, G. F., Maleky, F., Pal, R., & Egan, P. (2026) · Applied Food Research · doi:10.1016/j.afres.2026.102424
Hasan, U., & Nayeem, M. A. (2026) · Mathematics · doi:10.3390/math14142646 · Cited by 3
“Recent medical-imaging studies, however, show that zero-shot SAM performance is inconsistent in endoscopic and other specialized domains, where low-contrast boundaries and domain-specific textures differ substantially from common natural-image objects [15–17].”
Wei, J. (2026) · Scientific Journal of Technology · doi:10.54691/85dt0r63
“[9, 10] Domain adaptation has been the main response.”
Explainable Scenic Beauty Estimation of Urban Mountain Parks Using SAM2-Based Visual IndicatorsPreprint
He, X., Sun, H., Zhang, D., et al. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7202762
Luan, J., Jiang, T., Sun, X., et al. (2026) · Tomography · doi:10.3390/tomography12080110
Naidu, J. S., Naidu, S., & Baskaradoss, V. K. (2026) · Cureus · doi:10.7759/cureus.113726
“More recent highly cited publications included foundation models, general-purpose segmentation approaches, diffusion-based models, and multimodal synthesis reviews [17,20-22, 30,”
Lightweight Medical SAM with Self-prompting for Ultrasound Breast Tumor Segmentation
Zheng, Y., Dong, Y., Yao, Z., et al. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.111184
Self-citationPFPSAM: parameter-efficient fine-tuning for polyp segmentation with SAM
Handa, C. A. I., Chen, W., Xu, Z., et al. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.111123
MorVess: Morphology-aware pulmonary vessel segmentation network
Mao, F., Chen, Y., Wu, B., et al. (2026) · Pattern Recognition · doi:10.1016/j.patcog.2026.114550
“Several studies have systematically quantified this domain gap: Huang et al. [27] demonstrated under various prompt interaction settings that unadapted SAM exhibits a significant drop in Dice scores for medical tasks, while He et al. [28] reported up to a 70% performance decline across 12 medical…”
Self-citationFang, Y., Jiang, B., Huo, H., et al. (2026) · Expert systems with applications · doi:10.1016/j.eswa.2026.133883
Kang, H., Xing, F., Gilbert, I. R., Kim, J., Kostolansky, A. R., Sutton, B. P., Woo, J., & Perry, J. L. (2026) · Journal of Speech Language and Hearing Research · doi:10.1044/2026_jslhr-25-00495
Ping, S., Lin, Z., Lin, L., et al. (2026) · Bioengineering · doi:10.3390/bioengineering13080914
“Independent evaluations have shown that SAM-style models behave unevenly across modalities, target sizes, and prompt types [14,15].”
An Evidence-based Analysis of Mask Selection in Zero-Shot Medical Image Segmentation
Acharjee, T., Naidu, R. C. A., & Sangeetha.V (2026) · International Conference on Sustainable Computing and Smart Systems (ICSCSS) · doi:10.1109/icscss69635.2026.11646018
“When there is ambiguity in the quality of these masks, the similarity of confidence scores does not provide guidance on which mask is most likely to be clinically relevant[7].”
Liu, Y., Bian, Y., Zhuang, M., et al. (2026) · International Conference on Computer Vision, Image and Deep Learning (CVIDL) · doi:10.1109/cvidl70130.2026.11637394
Kinalioğlu, İ. H. (2026) · Pattern Recognition Letters · doi:10.1016/j.patrec.2026.08.026
Arefin, M. S., Islam, M. S., Nabi, M. S., Sumon, R. I., Lapina, M., Lapin, V. G., & Muthanna, M. S. A. (2026) · Diagnostics · doi:10.3390/diagnostics16172752
“General-purpose foundation segmentation models, including the Segment Anything Model, have also been evaluated in medical applications [16–20].”
Kai, C., Nakaya, M., Kasai, S., Tamori, H., Fujiwara, H., Kobayashi, T., Hashimoto, M., Kitamura, Y., Ohba, S., Nakada, M., et al. (2026) · Journal of Imaging Informatics in Medicine · doi:10.1007/s10278-026-02222-5
Zhang, Z., & He, L. (2026) · International Conference on Applied Mathematics, Modelling and Intelligent Computing (CAMMIC) · doi:10.1145/3817124.3817353
“Studies have demonstrated that SAM is highly sensitive to prompt strategies, yielding differentiated performance with different prompt types (i.e., point, box, mask) [3-5].”
Aktas, H. E., Tasci, E. S., Peng, L., et al. (2026) · Cancers · doi:10.3390/cancers18172836
“Although SAM has demonstrated strong zero-shot performance in natural images, its performance in medical imaging has been more variable, particularly for specialized targets characterized by low contrast, complex anatomy, or poorly defined boundaries [16,17].”
Bamwenda, J., Özerdem, M. S., Ayyıldız, O., et al. (2026) · Journal of Clinical Medicine · doi:10.3390/jcm15176846
Noaman, N. F., Jiao, L., Kanber, B. M., Liu, X., & Smadi, A. A. (2026) · Applied Soft Computing · doi:10.1016/j.asoc.2026.116375
FSD-Net: Foundation-Guided Spatiotemporal Distillation for Video Polyp Segmentation
Leng, Y., Min, S., & Li, C. B. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37447-9_31
PO-IoU: Point-Offset-Guided Zero-Shot Knee Joint Segmentation for Osteoarthritis Assessment
Yoo, J.-C., & Yoo, J. (2026) · Journal of the Korea Computer Graphics Society · doi:10.15701/kcgs.2026.32.4.67
Deep Learning-based Intelligent Diagnosis of Congenital Uterine Anomalies in 3-D Ultrasound
Xu, Y., Huang, Y., Deng, J., Zhang, Y., Zhang, H., Qu, J., Zheng, S., Tang, X., Chen, H., Chen, C., et al. (2026) · Ultrasound in Medicine & Biology · doi:10.1016/j.ultrasmedbio.2026.08.015
“In recent years, artificial intelligence and deep learning (DL) have made significant advances in the field of medical image analysis [12, 13, 14, 15], especially for intelligent US [16, 17, 18, 19, 20].”
Self-citationZhu, C., Zhu, H., Wang, Q., Cong, H., Huang, H., Shao, T., Zhu, H., & Cheng, Q. (2026) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2026.103533
Foundation Models in Medical Imaging: A Systematic Review, Taxonomy, and Future Research Agenda
Ghanim, R. (2026) · Academia Open · doi:10.21070/acopen.12.2027.15324
FST-SAM3: Taming SAM 3 with Frequency-Spatio-Temporal Refinement for Video Polyp Segmentation
Wu, G., Chen, G., Wu, H., & Qin, J. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37032-7_13
Sui, F., Hu, Y., Chi, M., & Du, Y. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.111508
Metadata
- Authors
- Huang, Y., Yang, X., Liu, L., Zhou, H., Chang, A., Zhou, X., Chen, R., Yu, J., Chen, J., Chen, C., Liu, S., Chi, H., Hu, X., Yue, K., Li, L., Grau, V., Fan, D.-P., Dong, F., & Ni, D.
- Affiliations
- Shenzhen University (CN)
- Zhejiang University (CN)
- Hunan First Normal University (CN)
- University of Oxford (GB)
- ETH Zurich (CH)
- Jinan University (CN)
- Corresponding author
- Dong, Fajin
- Ni, Dong
- ORCID
- Venue
- Medical Image Analysis
- Publisher
- Elsevier BV
- Published
- 2024-02
- Volume
- 92
- Pages
- 103061
- ISSN
- 1361-8415 (ISSN-L)
- OpenAlex
- W4389430914
- Semantic Scholar
- 057ddecf9b9dedd35f824c8b289c30316b4fa22e
- Cited by
- 536
- References
- 101
- Citations by year
- 2026: 134 · 2025: 230 · 2024: 154
- FWCI
- 48.56
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- Medical Image Analysis (publishedVersion)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- ePrints Soton (University of Southampton) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Natural Science Foundation of China
- Shenzhen Science and Technology Innovation Committee
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Medical Image Segmentation Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- Segment Anything Model, medical image segmentation, foundation models, point prompts, interactive segmentation, fine-tuning, boundary complexity
- MeSH terms
- Diagnostic Imaging, Humans, Image Processing, Computer-Assisted
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-08, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-08 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
