Exploring scalable medical image encoders beyond text supervision Article Swipe
Fernando Pérez‐García
,
Harshita Sharma
,
Sam Bond-Taylor
,
Kenza Bouzid
,
Valentina Salvatelli
,
Maximilian Ilse
,
Shruthi Bannur
,
Daniel C. Castro
,
Anton Schwaighofer
,
Matthew P. Lungren
,
Maria Teodora Wetscherek
,
Noel Codella
,
Stephanie L. Hyland
,
Javier Alvarez-Valle
,
Ozan Oktay
·
YOU?
·
· 2024
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2401.10815
YOU?
·
· 2024
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2401.10815
Language-supervised pre-training has proven to be a valuable method for extracting semantically meaningful features from images, serving as a foundational element in multimodal systems within the computer vision and medical imaging domains. However, the computed features are limited by the information contained in the text, which is particularly problematic in medical imaging, where the findings described by radiologists focus on specific observations. This challenge is compounded by the scarcity of paired imaging-text data due to concerns over leakage of personal health information. In this work, we fundamentally challenge the prevailing reliance on language supervision for learning general-purpose biomedical imaging encoders. We introduce RAD-DINO, a biomedical image encoder pre-trained solely on unimodal biomedical imaging data that obtains similar or greater performance than state-of-the-art biomedical language-supervised models on a diverse range of benchmarks. Specifically, the quality of learned representations is evaluated on standard imaging tasks (classification and semantic segmentation), and a vision-language alignment task (text report generation from images). To further demonstrate the drawback of language supervision, we show that features from RAD-DINO correlate with other medical records (e.g., sex or age) better than language-supervised models, which are generally not mentioned in radiology reports. Finally, we conduct a series of ablations determining the factors in RAD-DINO's performance; notably, we observe that RAD-DINO's downstream performance scales well with the quantity and diversity of training data, demonstrating that image-only supervision is a scalable approach for training a foundational biomedical image encoder. Model weights of RAD-DINO trained on publicly available datasets are available at https://huggingface.co/microsoft/rad-dino.
Related Topics
Concepts
Metadata
- Type
- preprint
- Language
- en
- Landing Page
- http://arxiv.org/abs/2401.10815
- https://arxiv.org/pdf/2401.10815
- OA Status
- green
- Cited By
- 8
- Related Works
- 10
- OpenAlex ID
- https://openalex.org/W4391124801
All OpenAlex metadata
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4391124801Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2401.10815Digital Object Identifier
- Title
-
Exploring scalable medical image encoders beyond text supervisionWork title
- Type
-
preprintOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2024Year of publication
- Publication date
-
2024-01-19Full publication date if available
- Authors
-
Fernando Pérez‐García, Harshita Sharma, Sam Bond-Taylor, Kenza Bouzid, Valentina Salvatelli, Maximilian Ilse, Shruthi Bannur, Daniel C. Castro, Anton Schwaighofer, Matthew P. Lungren, Maria Teodora Wetscherek, Noel Codella, Stephanie L. Hyland, Javier Alvarez-Valle, Ozan OktayList of authors in order
- Landing page
-
https://arxiv.org/abs/2401.10815Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2401.10815Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://arxiv.org/pdf/2401.10815Direct OA link when available
- Concepts
-
Computer science, Encoder, Scalability, Segmentation, Artificial intelligence, Medical imaging, Natural language processing, Focus (optics), Information retrieval, Database, Optics, Physics, Operating systemTop concepts (fields/topics) attached by OpenAlex
- Cited by
-
8Total citation count in OpenAlex
- Citations by year (recent)
-
2025: 5, 2024: 2Per-year citation counts (last 5 years)
- Related works (count)
-
10Other works algorithmically related by OpenAlex
Full payload
| id | https://openalex.org/W4391124801 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2401.10815 |
| ids.doi | https://doi.org/10.48550/arxiv.2401.10815 |
| ids.openalex | https://openalex.org/W4391124801 |
| fwci | 1.70379797 |
| type | preprint |
| title | Exploring scalable medical image encoders beyond text supervision |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| topics[0].id | https://openalex.org/T11636 |
| topics[0].field.id | https://openalex.org/fields/27 |
| topics[0].field.display_name | Medicine |
| topics[0].score | 0.9898999929428101 |
| topics[0].domain.id | https://openalex.org/domains/4 |
| topics[0].domain.display_name | Health Sciences |
| topics[0].subfield.id | https://openalex.org/subfields/2718 |
| topics[0].subfield.display_name | Health Informatics |
| topics[0].display_name | Artificial Intelligence in Healthcare and Education |
| topics[1].id | https://openalex.org/T12422 |
| topics[1].field.id | https://openalex.org/fields/27 |
| topics[1].field.display_name | Medicine |
| topics[1].score | 0.9861999750137329 |
| topics[1].domain.id | https://openalex.org/domains/4 |
| topics[1].domain.display_name | Health Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/2741 |
| topics[1].subfield.display_name | Radiology, Nuclear Medicine and Imaging |
| topics[1].display_name | Radiomics and Machine Learning in Medical Imaging |
| topics[2].id | https://openalex.org/T11775 |
| topics[2].field.id | https://openalex.org/fields/27 |
| topics[2].field.display_name | Medicine |
| topics[2].score | 0.9699000120162964 |
| topics[2].domain.id | https://openalex.org/domains/4 |
| topics[2].domain.display_name | Health Sciences |
| topics[2].subfield.id | https://openalex.org/subfields/2741 |
| topics[2].subfield.display_name | Radiology, Nuclear Medicine and Imaging |
| topics[2].display_name | COVID-19 diagnosis using AI |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| concepts[0].id | https://openalex.org/C41008148 |
| concepts[0].level | 0 |
| concepts[0].score | 0.7174358367919922 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[0].display_name | Computer science |
| concepts[1].id | https://openalex.org/C118505674 |
| concepts[1].level | 2 |
| concepts[1].score | 0.6276537179946899 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q42586063 |
| concepts[1].display_name | Encoder |
| concepts[2].id | https://openalex.org/C48044578 |
| concepts[2].level | 2 |
| concepts[2].score | 0.6063056588172913 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q727490 |
| concepts[2].display_name | Scalability |
| concepts[3].id | https://openalex.org/C89600930 |
| concepts[3].level | 2 |
| concepts[3].score | 0.5472951531410217 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q1423946 |
| concepts[3].display_name | Segmentation |
| concepts[4].id | https://openalex.org/C154945302 |
| concepts[4].level | 1 |
| concepts[4].score | 0.5390595197677612 |
| concepts[4].wikidata | https://www.wikidata.org/wiki/Q11660 |
| concepts[4].display_name | Artificial intelligence |
| concepts[5].id | https://openalex.org/C31601959 |
| concepts[5].level | 2 |
| concepts[5].score | 0.521580159664154 |
| concepts[5].wikidata | https://www.wikidata.org/wiki/Q931309 |
| concepts[5].display_name | Medical imaging |
| concepts[6].id | https://openalex.org/C204321447 |
| concepts[6].level | 1 |
| concepts[6].score | 0.46744024753570557 |
| concepts[6].wikidata | https://www.wikidata.org/wiki/Q30642 |
| concepts[6].display_name | Natural language processing |
| concepts[7].id | https://openalex.org/C192209626 |
| concepts[7].level | 2 |
| concepts[7].score | 0.41132932901382446 |
| concepts[7].wikidata | https://www.wikidata.org/wiki/Q190909 |
| concepts[7].display_name | Focus (optics) |
| concepts[8].id | https://openalex.org/C23123220 |
| concepts[8].level | 1 |
| concepts[8].score | 0.3219284415245056 |
| concepts[8].wikidata | https://www.wikidata.org/wiki/Q816826 |
| concepts[8].display_name | Information retrieval |
| concepts[9].id | https://openalex.org/C77088390 |
| concepts[9].level | 1 |
| concepts[9].score | 0.08568498492240906 |
| concepts[9].wikidata | https://www.wikidata.org/wiki/Q8513 |
| concepts[9].display_name | Database |
| concepts[10].id | https://openalex.org/C120665830 |
| concepts[10].level | 1 |
| concepts[10].score | 0.0 |
| concepts[10].wikidata | https://www.wikidata.org/wiki/Q14620 |
| concepts[10].display_name | Optics |
| concepts[11].id | https://openalex.org/C121332964 |
| concepts[11].level | 0 |
| concepts[11].score | 0.0 |
| concepts[11].wikidata | https://www.wikidata.org/wiki/Q413 |
| concepts[11].display_name | Physics |
| concepts[12].id | https://openalex.org/C111919701 |
| concepts[12].level | 1 |
| concepts[12].score | 0.0 |
| concepts[12].wikidata | https://www.wikidata.org/wiki/Q9135 |
| concepts[12].display_name | Operating system |
| keywords[0].id | https://openalex.org/keywords/computer-science |
| keywords[0].score | 0.7174358367919922 |
| keywords[0].display_name | Computer science |
| keywords[1].id | https://openalex.org/keywords/encoder |
| keywords[1].score | 0.6276537179946899 |
| keywords[1].display_name | Encoder |
| keywords[2].id | https://openalex.org/keywords/scalability |
| keywords[2].score | 0.6063056588172913 |
| keywords[2].display_name | Scalability |
| keywords[3].id | https://openalex.org/keywords/segmentation |
| keywords[3].score | 0.5472951531410217 |
| keywords[3].display_name | Segmentation |
| keywords[4].id | https://openalex.org/keywords/artificial-intelligence |
| keywords[4].score | 0.5390595197677612 |
| keywords[4].display_name | Artificial intelligence |
| keywords[5].id | https://openalex.org/keywords/medical-imaging |
| keywords[5].score | 0.521580159664154 |
| keywords[5].display_name | Medical imaging |
| keywords[6].id | https://openalex.org/keywords/natural-language-processing |
| keywords[6].score | 0.46744024753570557 |
| keywords[6].display_name | Natural language processing |
| keywords[7].id | https://openalex.org/keywords/focus |
| keywords[7].score | 0.41132932901382446 |
| keywords[7].display_name | Focus (optics) |
| keywords[8].id | https://openalex.org/keywords/information-retrieval |
| keywords[8].score | 0.3219284415245056 |
| keywords[8].display_name | Information retrieval |
| keywords[9].id | https://openalex.org/keywords/database |
| keywords[9].score | 0.08568498492240906 |
| keywords[9].display_name | Database |
| language | en |
| locations[0].id | pmh:oai:arXiv.org:2401.10815 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4306400194 |
| locations[0].source.issn | |
| locations[0].source.type | repository |
| locations[0].source.is_oa | True |
| locations[0].source.issn_l | |
| locations[0].source.is_core | False |
| locations[0].source.is_in_doaj | False |
| locations[0].source.display_name | arXiv (Cornell University) |
| locations[0].source.host_organization | https://openalex.org/I205783295 |
| locations[0].source.host_organization_name | Cornell University |
| locations[0].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[0].license | |
| locations[0].pdf_url | https://arxiv.org/pdf/2401.10815 |
| locations[0].version | submittedVersion |
| locations[0].raw_type | text |
| locations[0].license_id | |
| locations[0].is_accepted | False |
| locations[0].is_published | False |
| locations[0].raw_source_name | |
| locations[0].landing_page_url | http://arxiv.org/abs/2401.10815 |
| locations[1].id | doi:10.48550/arxiv.2401.10815 |
| locations[1].is_oa | True |
| locations[1].source.id | https://openalex.org/S4306400194 |
| locations[1].source.issn | |
| locations[1].source.type | repository |
| locations[1].source.is_oa | True |
| locations[1].source.issn_l | |
| locations[1].source.is_core | False |
| locations[1].source.is_in_doaj | False |
| locations[1].source.display_name | arXiv (Cornell University) |
| locations[1].source.host_organization | https://openalex.org/I205783295 |
| locations[1].source.host_organization_name | Cornell University |
| locations[1].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[1].license | |
| locations[1].pdf_url | |
| locations[1].version | |
| locations[1].raw_type | article-journal |
| locations[1].license_id | |
| locations[1].is_accepted | False |
| locations[1].is_published | |
| locations[1].raw_source_name | |
| locations[1].landing_page_url | https://doi.org/10.48550/arxiv.2401.10815 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5054482170 |
| authorships[0].author.orcid | https://orcid.org/0000-0001-9090-3024 |
| authorships[0].author.display_name | Fernando Pérez‐García |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Pérez-García, Fernando |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5056301773 |
| authorships[1].author.orcid | https://orcid.org/0000-0003-4683-2606 |
| authorships[1].author.display_name | Harshita Sharma |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Sharma, Harshita |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5083332082 |
| authorships[2].author.orcid | https://orcid.org/0000-0003-1538-7909 |
| authorships[2].author.display_name | Sam Bond-Taylor |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Bond-Taylor, Sam |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5080583641 |
| authorships[3].author.orcid | https://orcid.org/0000-0001-5864-2792 |
| authorships[3].author.display_name | Kenza Bouzid |
| authorships[3].author_position | middle |
| authorships[3].raw_author_name | Bouzid, Kenza |
| authorships[3].is_corresponding | False |
| authorships[4].author.id | https://openalex.org/A5025795882 |
| authorships[4].author.orcid | https://orcid.org/0000-0002-3232-4101 |
| authorships[4].author.display_name | Valentina Salvatelli |
| authorships[4].author_position | middle |
| authorships[4].raw_author_name | Salvatelli, Valentina |
| authorships[4].is_corresponding | False |
| authorships[5].author.id | https://openalex.org/A5112870737 |
| authorships[5].author.orcid | |
| authorships[5].author.display_name | Maximilian Ilse |
| authorships[5].author_position | middle |
| authorships[5].raw_author_name | Ilse, Maximilian |
| authorships[5].is_corresponding | False |
| authorships[6].author.id | https://openalex.org/A5081683131 |
| authorships[6].author.orcid | https://orcid.org/0000-0001-5750-7628 |
| authorships[6].author.display_name | Shruthi Bannur |
| authorships[6].author_position | middle |
| authorships[6].raw_author_name | Bannur, Shruthi |
| authorships[6].is_corresponding | False |
| authorships[7].author.id | https://openalex.org/A5037843387 |
| authorships[7].author.orcid | https://orcid.org/0000-0002-6829-7045 |
| authorships[7].author.display_name | Daniel C. Castro |
| authorships[7].author_position | middle |
| authorships[7].raw_author_name | Castro, Daniel C. |
| authorships[7].is_corresponding | False |
| authorships[8].author.id | https://openalex.org/A5010707054 |
| authorships[8].author.orcid | https://orcid.org/0000-0003-1557-0527 |
| authorships[8].author.display_name | Anton Schwaighofer |
| authorships[8].author_position | middle |
| authorships[8].raw_author_name | Schwaighofer, Anton |
| authorships[8].is_corresponding | False |
| authorships[9].author.id | https://openalex.org/A5038969629 |
| authorships[9].author.orcid | https://orcid.org/0000-0002-8591-5861 |
| authorships[9].author.display_name | Matthew P. Lungren |
| authorships[9].author_position | middle |
| authorships[9].raw_author_name | Lungren, Matthew P. |
| authorships[9].is_corresponding | False |
| authorships[10].author.id | https://openalex.org/A5008697579 |
| authorships[10].author.orcid | https://orcid.org/0000-0003-2924-7587 |
| authorships[10].author.display_name | Maria Teodora Wetscherek |
| authorships[10].author_position | middle |
| authorships[10].raw_author_name | Wetscherek, Maria |
| authorships[10].is_corresponding | False |
| authorships[11].author.id | https://openalex.org/A5073027533 |
| authorships[11].author.orcid | https://orcid.org/0000-0001-6735-9067 |
| authorships[11].author.display_name | Noel Codella |
| authorships[11].author_position | middle |
| authorships[11].raw_author_name | Codella, Noel |
| authorships[11].is_corresponding | False |
| authorships[12].author.id | https://openalex.org/A5020430024 |
| authorships[12].author.orcid | https://orcid.org/0000-0002-4812-792X |
| authorships[12].author.display_name | Stephanie L. Hyland |
| authorships[12].author_position | middle |
| authorships[12].raw_author_name | Hyland, Stephanie L. |
| authorships[12].is_corresponding | False |
| authorships[13].author.id | https://openalex.org/A5050558499 |
| authorships[13].author.orcid | https://orcid.org/0000-0003-0906-4177 |
| authorships[13].author.display_name | Javier Alvarez-Valle |
| authorships[13].author_position | middle |
| authorships[13].raw_author_name | Alvarez-Valle, Javier |
| authorships[13].is_corresponding | False |
| authorships[14].author.id | https://openalex.org/A5038400548 |
| authorships[14].author.orcid | https://orcid.org/0000-0003-2976-0874 |
| authorships[14].author.display_name | Ozan Oktay |
| authorships[14].author_position | last |
| authorships[14].raw_author_name | Oktay, Ozan |
| authorships[14].is_corresponding | False |
| has_content.pdf | True |
| has_content.grobid_xml | False |
| is_paratext | False |
| open_access.is_oa | True |
| open_access.oa_url | https://arxiv.org/pdf/2401.10815 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2024-01-23T00:00:00 |
| display_name | Exploring scalable medical image encoders beyond text supervision |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T06:51:31.235846 |
| primary_topic.id | https://openalex.org/T11636 |
| primary_topic.field.id | https://openalex.org/fields/27 |
| primary_topic.field.display_name | Medicine |
| primary_topic.score | 0.9898999929428101 |
| primary_topic.domain.id | https://openalex.org/domains/4 |
| primary_topic.domain.display_name | Health Sciences |
| primary_topic.subfield.id | https://openalex.org/subfields/2718 |
| primary_topic.subfield.display_name | Health Informatics |
| primary_topic.display_name | Artificial Intelligence in Healthcare and Education |
| related_works | https://openalex.org/W2389214306, https://openalex.org/W4390516098, https://openalex.org/W4235240664, https://openalex.org/W2965083567, https://openalex.org/W1838576100, https://openalex.org/W2095886385, https://openalex.org/W2181948922, https://openalex.org/W2889616422, https://openalex.org/W2089704382, https://openalex.org/W4379141755 |
| cited_by_count | 8 |
| counts_by_year[0].year | 2025 |
| counts_by_year[0].cited_by_count | 5 |
| counts_by_year[1].year | 2024 |
| counts_by_year[1].cited_by_count | 2 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2401.10815 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4306400194 |
| best_oa_location.source.issn | |
| best_oa_location.source.type | repository |
| best_oa_location.source.is_oa | True |
| best_oa_location.source.issn_l | |
| best_oa_location.source.is_core | False |
| best_oa_location.source.is_in_doaj | False |
| best_oa_location.source.display_name | arXiv (Cornell University) |
| best_oa_location.source.host_organization | https://openalex.org/I205783295 |
| best_oa_location.source.host_organization_name | Cornell University |
| best_oa_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| best_oa_location.license | |
| best_oa_location.pdf_url | https://arxiv.org/pdf/2401.10815 |
| best_oa_location.version | submittedVersion |
| best_oa_location.raw_type | text |
| best_oa_location.license_id | |
| best_oa_location.is_accepted | False |
| best_oa_location.is_published | False |
| best_oa_location.raw_source_name | |
| best_oa_location.landing_page_url | http://arxiv.org/abs/2401.10815 |
| primary_location.id | pmh:oai:arXiv.org:2401.10815 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4306400194 |
| primary_location.source.issn | |
| primary_location.source.type | repository |
| primary_location.source.is_oa | True |
| primary_location.source.issn_l | |
| primary_location.source.is_core | False |
| primary_location.source.is_in_doaj | False |
| primary_location.source.display_name | arXiv (Cornell University) |
| primary_location.source.host_organization | https://openalex.org/I205783295 |
| primary_location.source.host_organization_name | Cornell University |
| primary_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| primary_location.license | |
| primary_location.pdf_url | https://arxiv.org/pdf/2401.10815 |
| primary_location.version | submittedVersion |
| primary_location.raw_type | text |
| primary_location.license_id | |
| primary_location.is_accepted | False |
| primary_location.is_published | False |
| primary_location.raw_source_name | |
| primary_location.landing_page_url | http://arxiv.org/abs/2401.10815 |
| publication_date | 2024-01-19 |
| publication_year | 2024 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 6, 18, 103, 126, 148, 195, 227, 232 |
| abstract_inverted_index.In | 82 |
| abstract_inverted_index.To | 157 |
| abstract_inverted_index.We | 100 |
| abstract_inverted_index.as | 17 |
| abstract_inverted_index.at | 248 |
| abstract_inverted_index.be | 5 |
| abstract_inverted_index.by | 38, 56, 66 |
| abstract_inverted_index.in | 21, 42, 49, 189, 202 |
| abstract_inverted_index.is | 46, 64, 137, 226 |
| abstract_inverted_index.of | 69, 78, 129, 134, 162, 197, 219, 239 |
| abstract_inverted_index.on | 59, 91, 109, 125, 139, 242 |
| abstract_inverted_index.or | 117, 178 |
| abstract_inverted_index.to | 4, 74 |
| abstract_inverted_index.we | 85, 165, 193, 206 |
| abstract_inverted_index.and | 28, 144, 147, 217 |
| abstract_inverted_index.are | 36, 185, 246 |
| abstract_inverted_index.due | 73 |
| abstract_inverted_index.for | 9, 94, 230 |
| abstract_inverted_index.has | 2 |
| abstract_inverted_index.not | 187 |
| abstract_inverted_index.sex | 177 |
| abstract_inverted_index.the | 25, 33, 39, 43, 53, 67, 88, 132, 160, 200, 215 |
| abstract_inverted_index.This | 62 |
| abstract_inverted_index.age) | 179 |
| abstract_inverted_index.data | 72, 113 |
| abstract_inverted_index.from | 14, 155, 169 |
| abstract_inverted_index.over | 76 |
| abstract_inverted_index.show | 166 |
| abstract_inverted_index.task | 151 |
| abstract_inverted_index.than | 120, 181 |
| abstract_inverted_index.that | 114, 167, 208, 223 |
| abstract_inverted_index.this | 83 |
| abstract_inverted_index.well | 213 |
| abstract_inverted_index.with | 172, 214 |
| abstract_inverted_index.(text | 152 |
| abstract_inverted_index.Model | 237 |
| abstract_inverted_index.data, | 221 |
| abstract_inverted_index.focus | 58 |
| abstract_inverted_index.image | 105, 235 |
| abstract_inverted_index.other | 173 |
| abstract_inverted_index.range | 128 |
| abstract_inverted_index.tasks | 142 |
| abstract_inverted_index.text, | 44 |
| abstract_inverted_index.where | 52 |
| abstract_inverted_index.which | 45, 184 |
| abstract_inverted_index.work, | 84 |
| abstract_inverted_index.(e.g., | 176 |
| abstract_inverted_index.better | 180 |
| abstract_inverted_index.health | 80 |
| abstract_inverted_index.method | 8 |
| abstract_inverted_index.models | 124 |
| abstract_inverted_index.paired | 70 |
| abstract_inverted_index.proven | 3 |
| abstract_inverted_index.report | 153 |
| abstract_inverted_index.scales | 212 |
| abstract_inverted_index.series | 196 |
| abstract_inverted_index.solely | 108 |
| abstract_inverted_index.vision | 27 |
| abstract_inverted_index.within | 24 |
| abstract_inverted_index.conduct | 194 |
| abstract_inverted_index.diverse | 127 |
| abstract_inverted_index.element | 20 |
| abstract_inverted_index.encoder | 106 |
| abstract_inverted_index.factors | 201 |
| abstract_inverted_index.further | 158 |
| abstract_inverted_index.greater | 118 |
| abstract_inverted_index.images, | 15 |
| abstract_inverted_index.imaging | 30, 98, 112, 141 |
| abstract_inverted_index.leakage | 77 |
| abstract_inverted_index.learned | 135 |
| abstract_inverted_index.limited | 37 |
| abstract_inverted_index.medical | 29, 50, 174 |
| abstract_inverted_index.models, | 183 |
| abstract_inverted_index.observe | 207 |
| abstract_inverted_index.obtains | 115 |
| abstract_inverted_index.quality | 133 |
| abstract_inverted_index.records | 175 |
| abstract_inverted_index.serving | 16 |
| abstract_inverted_index.similar | 116 |
| abstract_inverted_index.systems | 23 |
| abstract_inverted_index.trained | 241 |
| abstract_inverted_index.weights | 238 |
| abstract_inverted_index.Finally, | 192 |
| abstract_inverted_index.However, | 32 |
| abstract_inverted_index.RAD-DINO | 170, 240 |
| abstract_inverted_index.approach | 229 |
| abstract_inverted_index.computed | 34 |
| abstract_inverted_index.computer | 26 |
| abstract_inverted_index.concerns | 75 |
| abstract_inverted_index.datasets | 245 |
| abstract_inverted_index.domains. | 31 |
| abstract_inverted_index.drawback | 161 |
| abstract_inverted_index.encoder. | 236 |
| abstract_inverted_index.features | 13, 35, 168 |
| abstract_inverted_index.findings | 54 |
| abstract_inverted_index.images). | 156 |
| abstract_inverted_index.imaging, | 51 |
| abstract_inverted_index.language | 92, 163 |
| abstract_inverted_index.learning | 95 |
| abstract_inverted_index.notably, | 205 |
| abstract_inverted_index.personal | 79 |
| abstract_inverted_index.publicly | 243 |
| abstract_inverted_index.quantity | 216 |
| abstract_inverted_index.reliance | 90 |
| abstract_inverted_index.reports. | 191 |
| abstract_inverted_index.scalable | 228 |
| abstract_inverted_index.scarcity | 68 |
| abstract_inverted_index.semantic | 145 |
| abstract_inverted_index.specific | 60 |
| abstract_inverted_index.standard | 140 |
| abstract_inverted_index.training | 220, 231 |
| abstract_inverted_index.unimodal | 110 |
| abstract_inverted_index.valuable | 7 |
| abstract_inverted_index.RAD-DINO, | 102 |
| abstract_inverted_index.ablations | 198 |
| abstract_inverted_index.alignment | 150 |
| abstract_inverted_index.available | 244, 247 |
| abstract_inverted_index.challenge | 63, 87 |
| abstract_inverted_index.contained | 41 |
| abstract_inverted_index.correlate | 171 |
| abstract_inverted_index.described | 55 |
| abstract_inverted_index.diversity | 218 |
| abstract_inverted_index.encoders. | 99 |
| abstract_inverted_index.evaluated | 138 |
| abstract_inverted_index.generally | 186 |
| abstract_inverted_index.introduce | 101 |
| abstract_inverted_index.mentioned | 188 |
| abstract_inverted_index.radiology | 190 |
| abstract_inverted_index.RAD-DINO's | 203, 209 |
| abstract_inverted_index.biomedical | 97, 104, 111, 122, 234 |
| abstract_inverted_index.compounded | 65 |
| abstract_inverted_index.downstream | 210 |
| abstract_inverted_index.extracting | 10 |
| abstract_inverted_index.generation | 154 |
| abstract_inverted_index.image-only | 224 |
| abstract_inverted_index.meaningful | 12 |
| abstract_inverted_index.multimodal | 22 |
| abstract_inverted_index.prevailing | 89 |
| abstract_inverted_index.benchmarks. | 130 |
| abstract_inverted_index.demonstrate | 159 |
| abstract_inverted_index.determining | 199 |
| abstract_inverted_index.information | 40 |
| abstract_inverted_index.performance | 119, 211 |
| abstract_inverted_index.pre-trained | 107 |
| abstract_inverted_index.problematic | 48 |
| abstract_inverted_index.supervision | 93, 225 |
| abstract_inverted_index.foundational | 19, 233 |
| abstract_inverted_index.imaging-text | 71 |
| abstract_inverted_index.information. | 81 |
| abstract_inverted_index.particularly | 47 |
| abstract_inverted_index.performance; | 204 |
| abstract_inverted_index.pre-training | 1 |
| abstract_inverted_index.radiologists | 57 |
| abstract_inverted_index.semantically | 11 |
| abstract_inverted_index.supervision, | 164 |
| abstract_inverted_index.Specifically, | 131 |
| abstract_inverted_index.demonstrating | 222 |
| abstract_inverted_index.fundamentally | 86 |
| abstract_inverted_index.observations. | 61 |
| abstract_inverted_index.segmentation), | 146 |
| abstract_inverted_index.(classification | 143 |
| abstract_inverted_index.general-purpose | 96 |
| abstract_inverted_index.representations | 136 |
| abstract_inverted_index.vision-language | 149 |
| abstract_inverted_index.state-of-the-art | 121 |
| abstract_inverted_index.Language-supervised | 0 |
| abstract_inverted_index.language-supervised | 123, 182 |
| abstract_inverted_index.https://huggingface.co/microsoft/rad-dino. | 249 |
| cited_by_percentile_year.max | 98 |
| cited_by_percentile_year.min | 94 |
| countries_distinct_count | 0 |
| institutions_distinct_count | 15 |
| sustainable_development_goals[0].id | https://metadata.un.org/sdg/4 |
| sustainable_development_goals[0].score | 0.7699999809265137 |
| sustainable_development_goals[0].display_name | Quality Education |
| citation_normalized_percentile.value | 0.78116609 |
| citation_normalized_percentile.is_in_top_1_percent | False |
| citation_normalized_percentile.is_in_top_10_percent | False |