{"doi":"10.1002/mp.17785","title":"A segment anything model‐guided and match‐based semi‐supervised segmentation framework for medical imaging","abstract":"<jats:title>Abstract</jats:title>\n                  <jats:sec>\n                    <jats:title>Background</jats:title>\n                    <jats:p>Semi‐supervised segmentation leverages sparse annotation information to learn rich representations from combined labeled and label‐less data for segmentation tasks. The Match‐based framework, by using the consistency constraint of segmentation results from different models/augmented label‐less inputs, is found effective in semi‐supervised learning. This approach, however, is challenged by the low quality of pseudo‐labels generated as intermediate products for training the network, due to the lack of the ‘‘ground‐truth’’ reference.</jats:p>\n                  </jats:sec>\n                  <jats:sec>\n                    <jats:title>Purpose</jats:title>\n                    <jats:p>This study aims to leverage the foundation model, segment anything model (SAM), to assist unsupervised learning of Match‐based frameworks. Trained with an extremely large dataset, SAM‐based methods generalize better than traditional models to various imaging domains, allow it to serve as an assistant to Match‐based frameworks to improve the quality of intermediate pseudo‐labels for semi‐supervised learning.</jats:p>\n                  </jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods</jats:title>\n                    <jats:p>We propose SAM‐Match, a SAM‐guided and Match‐based framework for semi‐supervised medical image segmentation. Our approach involves two main steps: First, we use pretrained Match‐based models to extract high‐confidence predictions for prompt generation. Second, these prompts and unlabeled images are input into a fine‐tuned SAM‐based method to produce high‐quality masks as pseudo‐labels. And the refined pseudo‐labels are further fed back to train the Match‐based framework. SAM‐Match can be trained in an end‐to‐end manner, facilitating interactions between the SAM‐ and Match‐based models.</jats:p>\n                  </jats:sec>\n                  <jats:sec>\n                    <jats:title>Results</jats:title>\n                    <jats:p>SAM‐Match demonstrates robust performance across multiple medical imaging datasets, including the ACDC cardiac MRI dataset, the BUSI breast ultrasound dataset, and an in‐house liver MRI dataset (MRLiver). We partitioned the datasets into training, validation, and test sets (70%, 10%, and 20% for ACDC; 60%, 9%, and 31% for BUSI; and 62%, 12%, and 25% for MRLiver). On ACDC, with only 3 labeled cases, we achieved a Dice score of 89.36% ± 0.06% on 20 test cases. For BUSI, using just 30 labeled samples for training, we attained a Dice score of 59.35% ± 0.12% on 170 test samples. On MRLiver, training with only 3 labeled cases resulted in a Dice score of 80.04% ± 0.11% on 12 test scans. Wilcoxon signed‐rank tests with Bonferroni corrections between the SAM‐Match framework and the other comparison methods further demonstrated the statistical significance of SAM‐Match's improvement in segmentation accuracy.</jats:p>\n                  </jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions</jats:title>\n                    <jats:p>\n                      Our SAM‐Match framework shows promising results in semi‐supervised semantic segmentation, effectively tackling the challenges of automatic prompt generation for SAM and high‐quality pseudo‐label generation for Match‐based models. It can help accelerate the adoption of semi‐supervised learning in segmentation tasks, particularly in data‐scarce scenarios. Our data and code will be made available at\n                      <jats:ext-link xmlns:xlink=\"http://www.w3.org/1999/xlink\" xlink:href=\"https://github.com/apple1986/SAMatch\">https://github.com/apple1986/SAMatch</jats:ext-link>\n                      .\n                    </jats:p>\n                  </jats:sec>","journal":"Medical Physics","year":2025,"id":680825,"datarank":0.38474240361923057,"base_score":2.5649493574615367,"endowment":2.5649493574615367,"self_citation_contribution":0.38474240361923057,"citation_network_contribution":0.0,"self_endowment_contribution":0.38474240361923057,"citer_contribution":0.0,"corpus_percentile":null,"corpus_rank":null,"citation_count":12,"citer_count":0,"citers_with_citation_signal":0,"citers_with_endowment":0,"datacite_reuse_total":0,"is_dataset":false,"is_dataset_confidence":null,"is_data_producer":false,"deposit_databanks":null,"is_oa":false,"file_count":0,"downloads":0,"has_version_chain":false,"published_date":null,"fair_score":null,"fair_percentile":null,"algorithm_id":"datarank_citation_only_1hop_v6","ranking_scope":"data_only","authors":[{"id":1289910,"name":"Xiaoxue Qian","orcid":"0000-0002-6185-1074","position":1,"is_corresponding":false},{"id":881958,"name":"Hua‐Chieh Shao","orcid":"0009-0003-7664-7257","position":2,"is_corresponding":false},{"id":1778795,"name":"Jax Luo","orcid":null,"position":3,"is_corresponding":false},{"id":473763,"name":"Weiguo Lu","orcid":"0000-0003-3036-7287","position":4,"is_corresponding":false},{"id":141187,"name":"You Zhang","orcid":null,"position":5,"is_corresponding":false},{"id":694805,"name":"Guoping Xu","orcid":"0000-0003-1431-7191","position":0,"is_corresponding":false}],"reference_count":0,"raw_metadata":{"has_enrichment":true,"resolved":true,"title":"A segment anything model‐guided and match‐based semi‐supervised segmentation framework for medical imaging","abstract":"<jats:title>Abstract</jats:title>\n                  <jats:sec>\n                    <jats:title>Background</jats:title>\n                    <jats:p>Semi‐supervised segmentation leverages sparse annotation information to learn rich representations from combined labeled and label‐less data for segmentation tasks. The Match‐based framework, by using the consistency constraint of segmentation results from different models/augmented label‐less inputs, is found effective in semi‐supervised learning. This approach, however, is challenged by the low quality of pseudo‐labels generated as intermediate products for training the network, due to the lack of the ‘‘ground‐truth’’ reference.</jats:p>\n                  </jats:sec>\n                  <jats:sec>\n                    <jats:title>Purpose</jats:title>\n                    <jats:p>This study aims to leverage the foundation model, segment anything model (SAM), to assist unsupervised learning of Match‐based frameworks. Trained with an extremely large dataset, SAM‐based methods generalize better than traditional models to various imaging domains, allow it to serve as an assistant to Match‐based frameworks to improve the quality of intermediate pseudo‐labels for semi‐supervised learning.</jats:p>\n                  </jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods</jats:title>\n                    <jats:p>We propose SAM‐Match, a SAM‐guided and Match‐based framework for semi‐supervised medical image segmentation. Our approach involves two main steps: First, we use pretrained Match‐based models to extract high‐confidence predictions for prompt generation. Second, these prompts and unlabeled images are input into a fine‐tuned SAM‐based method to produce high‐quality masks as pseudo‐labels. And the refined pseudo‐labels are further fed back to train the Match‐based framework. SAM‐Match can be trained in an end‐to‐end manner, facilitating interactions between the SAM‐ and Match‐based models.</jats:p>\n                  </jats:sec>\n                  <jats:sec>\n                    <jats:title>Results</jats:title>\n                    <jats:p>SAM‐Match demonstrates robust performance across multiple medical imaging datasets, including the ACDC cardiac MRI dataset, the BUSI breast ultrasound dataset, and an in‐house liver MRI dataset (MRLiver). We partitioned the datasets into training, validation, and test sets (70%, 10%, and 20% for ACDC; 60%, 9%, and 31% for BUSI; and 62%, 12%, and 25% for MRLiver). On ACDC, with only 3 labeled cases, we achieved a Dice score of 89.36% ± 0.06% on 20 test cases. For BUSI, using just 30 labeled samples for training, we attained a Dice score of 59.35% ± 0.12% on 170 test samples. On MRLiver, training with only 3 labeled cases resulted in a Dice score of 80.04% ± 0.11% on 12 test scans. Wilcoxon signed‐rank tests with Bonferroni corrections between the SAM‐Match framework and the other comparison methods further demonstrated the statistical significance of SAM‐Match's improvement in segmentation accuracy.</jats:p>\n                  </jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions</jats:title>\n                    <jats:p>\n                      Our SAM‐Match framework shows promising results in semi‐supervised semantic segmentation, effectively tackling the challenges of automatic prompt generation for SAM and high‐quality pseudo‐label generation for Match‐based models. It can help accelerate the adoption of semi‐supervised learning in segmentation tasks, particularly in data‐scarce scenarios. Our data and code will be made available at\n                      <jats:ext-link xmlns:xlink=\"http://www.w3.org/1999/xlink\" xlink:href=\"https://github.com/apple1986/SAMatch\">https://github.com/apple1986/SAMatch</jats:ext-link>\n                      .\n                    </jats:p>\n                  </jats:sec>","is_dataset_classified":null,"base_score":2.5649493574615367,"endowment":2.5649493574615367,"datacite_reuse_total":0,"file_count":0,"downloads":0,"views":0,"has_version_chain":false,"is_dataset":false,"is_oa":false,"pmid":"40156370","pmcid":"PMC12149692","openalex_id":"https://openalex.org/W4408997492","authors":[],"funders":[{"funder_name":"NCI NIH HHS","grant_id":"R01 CA240808","title":null},{"funder_name":"NIBIB NIH HHS","grant_id":"R01 EB034691","title":null},{"funder_name":"NCI NIH HHS","grant_id":"R01 CA280135","title":null},{"funder_name":"NCI NIH HHS","grant_id":"R01 CA258987","title":null}],"total_grants":4,"fwci":5.6995,"citation_percentile":0.96565881,"influential_citations":0,"citation_trend":[{"year":2025,"count":4},{"year":2026,"count":8}],"oa_status":"hybrid","license":"cc-by-nc","oa_locations":[{"url":"https://onlinelibrary.wiley.com/doi/pdfdirect/10.1002/mp.17785","host_type":"journal"},{"url":"https://onlinelibrary.wiley.com/doi/pdfdirect/10.1002/mp.17785","host_type":"publisher"},{"url":"https://aapm.onlinelibrary.wiley.com/doi/pdf/10.1002/mp.17785","host_type":"publisher"},{"url":"https://doi.org/10.1002/mp.17785","host_type":"journal"},{"url":"https://pubmed.ncbi.nlm.nih.gov/40156370","host_type":"repository"},{"url":"https://www.ncbi.nlm.nih.gov/pmc/articles/12149692","host_type":"repository"},{"url":"https://europepmc.org/articles/PMC12149692","host_type":"Europe_PMC"},{"url":"https://europepmc.org/articles/PMC12149692?pdf=render","host_type":"Europe_PMC"}],"fields_of_study":["Advanced Neural Network Applications","Medical Image Segmentation Techniques","Domain Adaptation and Few-Shot Learning","Image Processing, Computer-Assisted","Humans","Supervised Machine Learning","Diagnostic Imaging"],"mesh_terms":["Supervised Machine Learning","Diagnostic Imaging","Humans","Image Processing, Computer-Assisted"],"keywords":["Computer science","Segmentation","Artificial intelligence","Leverage (statistics)","Ground truth","Medical imaging","Annotation","Pattern recognition (psychology)","Consistency (knowledge bases)","Supervised learning","Machine learning","Image segmentation","Deep learning","Artificial neural network","Segment Anything Model","Semi‐supervised Segmentation","Match‐based Framework"],"sdg_mappings":[],"linked_datasets":[],"clinical_trials":[],"software_tools":[],"database_accessions":[],"source":"live","citation_network_status":"fetched"},"created_at":"2026-08-17T16:12:31.632725Z","pmid":null,"pmcid":null,"fwci":null,"citation_percentile":null,"influential_citations":0,"oa_status":null,"license":null,"views":0,"total_file_size_bytes":0,"version_count":0,"fair_f":null,"fair_a":null,"fair_i":null,"fair_r":null,"fair_zscore":null,"fair_rationale":null,"fair_model":null,"fair_agent_version":null,"fair_fulltext_source":null,"fair_has_llm":null,"fair_computed_at":null,"clinical_trials":[],"software_tools":[],"db_accessions":[],"linked_datasets":[],"topics":[]}