{"id":50503,"topic":"ai","source":"Nature","title":"Learning from routine health system data builds better neuroimaging AI models - Nature","url":"https://www.nature.com/articles/s41591-026-04567-4","url_hash":"7184d8fcf5a24f6f6dd0f0b94ae45cc6f3925cf0","author":"","summary":"<a href=\"https://news.google.com/rss/articles/CBMiX0FVX3lxTFBGeUwyWGJIUktZMjJkRWZoejZhYVpYTkhCSkloN3BKSnAwQ3E2U1F2aDk4VXZYZ0xNc2Myd3ZiZnJFbTlJWUpxOFhKTHctWW8tR2tBdFBydnhlNkVYeEE4?oc=5\" target=\"_blank\">Learning from routine health system data builds better neuroimaging AI models</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">Nature</font>","content":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.\nAccess options\nAccess Nature and 54 other Nature Portfolio journals\nGet Nature+, our best-value online-access subscription\n27,99 € / 30 days\ncancel any time\nSubscribe to this journal\nReceive 12 print issues and online access\n269,00 € per year\nonly 22,42 € per issue\nBuy this article\n39,95 €\nPrices may be subject to local taxes which are calculated during checkout\nReferences\n- Radford, A. et al. Learning transferable visual models from natural language supervision. In Proc. 38th International Conference on Machine Learning 8748–8763 (2021). This paper established contrastive image-text pretraining (CLIP) as a foundation for modern vision-language models. \n- Jiang, L. Y. et al. Health system-scale language models are all-purpose prediction engines. Nature 619, 357–362 (2023). This study demonstrated that large-scale health system data can support broadly useful clinical prediction models. \n- Assran, M. et al. Self-supervised learning from images with a joint-embedding predictive architecture. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 15619–15629 (2023). This paper introduced I-JEPA, the self-supervised representation learning method adapted here for volumetric medical imaging. \n- Lyu, Y. et al. Learning neuroimaging models from health system-scale data. Nat. Biomed. Eng. https://doi.org/10.1038/s41551-025-01608-0 (2026). This paper showed strong performance with neuroimaging learning from MRI–report pairs at health system scale. \n- Moor, M. et al. Foundation models for generalist medical artificial intelligence. Nature 616, 259–265 (2023). This perspective outlines the rationale for generalist medical AI models that learn across heterogeneous clinical data modalities and support multiple downstream tasks. \nAdditional information\nPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.\nThis is a summary of: Kondepudi, A. et al. Health system learning enables generalist neuroimaging models. Nat. Med. https://doi.org/10.1038/s41591-026-04497-1 (2026).\nA.K. and T.H. used ChatGPT to help prepare their contribution to this Research Briefing.\nAbout this article\nCite this article\nLearning from routine health system data builds better neuroimaging AI models. Nat Med (2026). https://doi.org/10.1038/s41591-026-04567-4\n- Published: \n- Version of record: \n- DOI: https://doi.org/10.1038/s41591-026-04567-4","image_url":"https://media.springernature.com/m685/springer-static/image/art%3A10.1038%2Fs41591-026-04567-4/MediaObjects/41591_2026_4567_Fig1_HTML.png","lang":"en","published_at":"2026-07-31T09:43:01+00:00","fetched_at":"2026-07-31T10:15:04+00:00","status":"read","starred":0,"extract_state":"ok","summary_auto":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.","cluster_id":null,"extract_retries":0,"extract_error":null,"contract_version":"news_item.v1","format_contract_version":"news_item_formats.v1","dedup_url":"https://www.nature.com/articles/s41591-026-04567-4","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 2904 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":2904,"summary_length":458,"usable_text_length":2904,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2904,"summary_length":458}},"news_item":{"id":50503,"canonical_url":"https://www.nature.com/articles/s41591-026-04567-4","source_url":"https://www.nature.com/articles/s41591-026-04567-4","title":"Learning from routine health system data builds better neuroimaging AI models - Nature","source_name":"Nature","author":null,"published_at":"2026-07-31T09:43:01+00:00","locale":"en","topic":"ai","tags":[],"rss_summary":"<a href=\"https://news.google.com/rss/articles/CBMiX0FVX3lxTFBGeUwyWGJIUktZMjJkRWZoejZhYVpYTkhCSkloN3BKSnAwQ3E2U1F2aDk4VXZYZ0xNc2Myd3ZiZnJFbTlJWUpxOFhKTHctWW8tR2tBdFBydnhlNkVYeEE4?oc=5\" target=\"_blank\">Learning from routine health system data builds better neuroimaging AI models</a>&nbsp;&nbsp;<font color=\"#6f6f6f\">Nature</font>","full_text":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.\nAccess options\nAccess Nature and 54 other Nature Portfolio journals\nGet Nature+, our best-value online-access subscription\n27,99 € / 30 days\ncancel any time\nSubscribe to this journal\nReceive 12 print issues and online access\n269,00 € per year\nonly 22,42 € per issue\nBuy this article\n39,95 €\nPrices may be subject to local taxes which are calculated during checkout\nReferences\n- Radford, A. et al. Learning transferable visual models from natural language supervision. In Proc. 38th International Conference on Machine Learning 8748–8763 (2021). This paper established contrastive image-text pretraining (CLIP) as a foundation for modern vision-language models. \n- Jiang, L. Y. et al. Health system-scale language models are all-purpose prediction engines. Nature 619, 357–362 (2023). This study demonstrated that large-scale health system data can support broadly useful clinical prediction models. \n- Assran, M. et al. Self-supervised learning from images with a joint-embedding predictive architecture. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 15619–15629 (2023). This paper introduced I-JEPA, the self-supervised representation learning method adapted here for volumetric medical imaging. \n- Lyu, Y. et al. Learning neuroimaging models from health system-scale data. Nat. Biomed. Eng. https://doi.org/10.1038/s41551-025-01608-0 (2026). This paper showed strong performance with neuroimaging learning from MRI–report pairs at health system scale. \n- Moor, M. et al. Foundation models for generalist medical artificial intelligence. Nature 616, 259–265 (2023). This perspective outlines the rationale for generalist medical AI models that learn across heterogeneous clinical data modalities and support multiple downstream tasks. \nAdditional information\nPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.\nThis is a summary of: Kondepudi, A. et al. Health system learning enables generalist neuroimaging models. Nat. Med. https://doi.org/10.1038/s41591-026-04497-1 (2026).\nA.K. and T.H. used ChatGPT to help prepare their contribution to this Research Briefing.\nAbout this article\nCite this article\nLearning from routine health system data builds better neuroimaging AI models. Nat Med (2026). https://doi.org/10.1038/s41591-026-04567-4\n- Published: \n- Version of record: \n- DOI: https://doi.org/10.1038/s41591-026-04567-4","excerpt":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.","extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 2904 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.nature.com/articles/s41591-026-04567-4","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 2904 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":2904,"summary_length":458,"usable_text_length":2904,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2904,"summary_length":458}}},"display_formats":["compact","card","full","digest_section","json"]},"daily_stack_record":{"title":"Learning from routine health system data builds better neuroimaging AI models - Nature","url":"https://www.nature.com/articles/s41591-026-04567-4","summary":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.","source":"Nature","date":"2026-07-31T09:43:01+00:00","content":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.\nAccess options\nAccess Nature and 54 other Nature Portfolio journals\nGet Nature+, our best-value online-access subscription\n27,99 € / 30 days\ncancel any time\nSubscribe to this journal\nReceive 12 print issues and online access\n269,00 € per year\nonly 22,42 € per issue\nBuy this article\n39,95 €\nPrices may be subject to local taxes which are calculated during checkout\nReferences\n- Radford, A. et al. Learning transferable visual models from natural language supervision. In Proc. 38th International Conference on Machine Learning 8748–8763 (2021). This paper established contrastive image-text pretraining (CLIP) as a foundation for modern vision-language models. \n- Jiang, L. Y. et al. Health system-scale language models are all-purpose prediction engines. Nature 619, 357–362 (2023). This study demonstrated that large-scale health system data can support broadly useful clinical prediction models. \n- Assran, M. et al. Self-supervised learning from images with a joint-embedding predictive architecture. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 15619–15629 (2023). This paper introduced I-JEPA, the self-supervised representation learning method adapted here for volumetric medical imaging. \n- Lyu, Y. et al. Learning neuroimaging models from health system-scale data. Nat. Biomed. Eng. https://doi.org/10.1038/s41551-025-01608-0 (2026). This paper showed strong performance with neuroimaging learning from MRI–report pairs at health system scale. \n- Moor, M. et al. Foundation models for generalist medical artificial intelligence. Nature 616, 259–265 (2023). This perspective outlines the rationale for generalist medical AI models that learn across heterogeneous clinical data modalities and support multiple downstream tasks. \nAdditional information\nPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.\nThis is a summary of: Kondepudi, A. et al. Health system learning enables generalist neuroimaging models. Nat. Med. https://doi.org/10.1038/s41591-026-04497-1 (2026).\nA.K. and T.H. used ChatGPT to help prepare their contribution to this Research Briefing.\nAbout this article\nCite this article\nLearning from routine health system data builds better neuroimaging AI models. Nat Med (2026). https://doi.org/10.1038/s41591-026-04567-4\n- Published: \n- Version of record: \n- DOI: https://doi.org/10.1038/s41591-026-04567-4","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.nature.com/articles/s41591-026-04567-4","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 2904 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 2904 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":2904,"summary_length":458,"usable_text_length":2904,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2904,"summary_length":458}},"tags":[]},"fallback_formats":["markdown","json","html"],"actions":{"read":"/item/50503","export_markdown":"/api/items/50503/export?format=markdown","export_json":"/api/items/50503/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.nature.com/articles/s41591-026-04567-4"},"formats":{"full":{"id":50503,"title":"Learning from routine health system data builds better neuroimaging AI models - Nature","url":"https://www.nature.com/articles/s41591-026-04567-4","source":"Nature","author":null,"published_at":"2026-07-31T09:43:01+00:00","locale":"en","topic":"ai","tags":[],"excerpt":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.","full_text":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.\nAccess options\nAccess Nature and 54 other Nature Portfolio journals\nGet Nature+, our best-value online-access subscription\n27,99 € / 30 days\ncancel any time\nSubscribe to this journal\nReceive 12 print issues and online access\n269,00 € per year\nonly 22,42 € per issue\nBuy this article\n39,95 €\nPrices may be subject to local taxes which are calculated during checkout\nReferences\n- Radford, A. et al. Learning transferable visual models from natural language supervision. In Proc. 38th International Conference on Machine Learning 8748–8763 (2021). This paper established contrastive image-text pretraining (CLIP) as a foundation for modern vision-language models. \n- Jiang, L. Y. et al. Health system-scale language models are all-purpose prediction engines. Nature 619, 357–362 (2023). This study demonstrated that large-scale health system data can support broadly useful clinical prediction models. \n- Assran, M. et al. Self-supervised learning from images with a joint-embedding predictive architecture. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 15619–15629 (2023). This paper introduced I-JEPA, the self-supervised representation learning method adapted here for volumetric medical imaging. \n- Lyu, Y. et al. Learning neuroimaging models from health system-scale data. Nat. Biomed. Eng. https://doi.org/10.1038/s41551-025-01608-0 (2026). This paper showed strong performance with neuroimaging learning from MRI–report pairs at health system scale. \n- Moor, M. et al. Foundation models for generalist medical artificial intelligence. Nature 616, 259–265 (2023). This perspective outlines the rationale for generalist medical AI models that learn across heterogeneous clinical data modalities and support multiple downstream tasks. \nAdditional information\nPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.\nThis is a summary of: Kondepudi, A. et al. Health system learning enables generalist neuroimaging models. Nat. Med. https://doi.org/10.1038/s41591-026-04497-1 (2026).\nA.K. and T.H. used ChatGPT to help prepare their contribution to this Research Briefing.\nAbout this article\nCite this article\nLearning from routine health system data builds better neuroimaging AI models. Nat Med (2026). https://doi.org/10.1038/s41591-026-04567-4\n- Published: \n- Version of record: \n- DOI: https://doi.org/10.1038/s41591-026-04567-4","reading_time_min":2,"extraction":{"state":"ok","confidence":0.9,"error":null,"explanation":"High confidence: full text extraction produced 2904 characters.","diagnostics_url":"/api/diagnose?url=https%3A//www.nature.com/articles/s41591-026-04567-4","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 2904 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":2904,"summary_length":458,"usable_text_length":2904,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2904,"summary_length":458}}},"quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 2904 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":2904,"summary_length":458,"usable_text_length":2904,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2904,"summary_length":458}},"actions":{"read":"/item/50503","export_markdown":"/api/items/50503/export?format=markdown","export_json":"/api/items/50503/export?format=json","diagnose":"/api/diagnose?url=https%3A//www.nature.com/articles/s41591-026-04567-4"}},"digest":{"id":50503,"title":"Learning from routine health system data builds better neuroimaging AI models - Nature","url":"https://www.nature.com/articles/s41591-026-04567-4","source":"Nature","topic":"ai","published_at":"2026-07-31T09:43:01+00:00","excerpt":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated…","quality_bucket":"high","quality_reason":"High confidence: full text extraction produced 2904 characters.","reading_time_min":2,"cluster_id":null},"card":{"display_title":"Learning from routine health system data builds better neuroimaging AI models - Nature","subtitle":"Nature · 2026-07-31","summary":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of…","badges":["quality:high"],"links":{"read":"/item/50503","original":"https://www.nature.com/articles/s41591-026-04567-4","diagnose":"/api/diagnose?url=https%3A//www.nature.com/articles/s41591-026-04567-4"},"quality_warning":null},"export":{"title":"Learning from routine health system data builds better neuroimaging AI models - Nature","url":"https://www.nature.com/articles/s41591-026-04567-4","summary":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.","source":"Nature","date":"2026-07-31T09:43:01+00:00","content":"We trained a three-dimensional visual foundation model directly on 5.24 million routine clinical computed tomography (CT) and magnetic resonance imaging (MRI) image series so that it learned a shared representation of neuroanatomy and disease. The model demonstrated state-of-the-art diagnosis, in contrast to foundation models that are trained on public Internet and medical data, and enabled preliminary report generation and triage in real health systems.\nAccess options\nAccess Nature and 54 other Nature Portfolio journals\nGet Nature+, our best-value online-access subscription\n27,99 € / 30 days\ncancel any time\nSubscribe to this journal\nReceive 12 print issues and online access\n269,00 € per year\nonly 22,42 € per issue\nBuy this article\n39,95 €\nPrices may be subject to local taxes which are calculated during checkout\nReferences\n- Radford, A. et al. Learning transferable visual models from natural language supervision. In Proc. 38th International Conference on Machine Learning 8748–8763 (2021). This paper established contrastive image-text pretraining (CLIP) as a foundation for modern vision-language models. \n- Jiang, L. Y. et al. Health system-scale language models are all-purpose prediction engines. Nature 619, 357–362 (2023). This study demonstrated that large-scale health system data can support broadly useful clinical prediction models. \n- Assran, M. et al. Self-supervised learning from images with a joint-embedding predictive architecture. In Proc. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 15619–15629 (2023). This paper introduced I-JEPA, the self-supervised representation learning method adapted here for volumetric medical imaging. \n- Lyu, Y. et al. Learning neuroimaging models from health system-scale data. Nat. Biomed. Eng. https://doi.org/10.1038/s41551-025-01608-0 (2026). This paper showed strong performance with neuroimaging learning from MRI–report pairs at health system scale. \n- Moor, M. et al. Foundation models for generalist medical artificial intelligence. Nature 616, 259–265 (2023). This perspective outlines the rationale for generalist medical AI models that learn across heterogeneous clinical data modalities and support multiple downstream tasks. \nAdditional information\nPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.\nThis is a summary of: Kondepudi, A. et al. Health system learning enables generalist neuroimaging models. Nat. Med. https://doi.org/10.1038/s41591-026-04497-1 (2026).\nA.K. and T.H. used ChatGPT to help prepare their contribution to this Research Briefing.\nAbout this article\nCite this article\nLearning from routine health system data builds better neuroimaging AI models. Nat Med (2026). https://doi.org/10.1038/s41591-026-04567-4\n- Published: \n- Version of record: \n- DOI: https://doi.org/10.1038/s41591-026-04567-4","confidence":0.9,"diagnostics_url":"/api/diagnose?url=https%3A//www.nature.com/articles/s41591-026-04567-4","quality_bucket":"high","failure_kind":"none","retryable":false,"quality_reason":"High confidence: full text extraction produced 2904 characters.","quality_profile":{"profile_version":"extraction_quality.v2","bucket":"high","confidence":0.9,"failure_kind":"none","retryable":false,"retry_after_attempts":0,"reason":"High confidence: full text extraction produced 2904 characters.","operator_guidance":{"severity":"ok","recommended_action":"trust_full_text","next_step":"Use the extracted full text as the primary article source.","operator_label":"Ready","can_retry":false,"can_use_summary":false,"diagnostics_required":false},"content_depth":{"contract_version":"content_depth.v1","category":"full_text","label":"Full text","has_full_text":true,"has_summary":true,"content_length":2904,"summary_length":458,"usable_text_length":2904,"source_field":"content"},"legacy_collapsed":false,"signals":{"extract_state":"ok","extract_error":null,"extract_retries":0,"content_length":2904,"summary_length":458}},"tags":[],"format_contract_version":"news_item_formats.v1"}}}