• Augmented Radiology: why map artificial intelligence in imaging from and for the Spanish-speaking world
• Augmented Radiology: why map artificial intelligence in imaging from and for the Spanish-speaking world
  • Abstract
  • The starting point
  • A double gap: language and reality
  • What Radiología Aumentada is
  • Three principles
  • Where we are heading
  • Declarations
  • Copyright and license
  • Disclaimer / Publisher’s note
  • References
  • Attribution and reuse

§Augmented Radiology: why map artificial intelligence in imaging from and for the Spanish-speaking world

Authors: Federico Martín Olivera Plata

Journal publication date: 2026-10-22

DOI: https://doi.org/10.59667/sjoranm.v35i2.14

Summary: Federico Martín Olivera Plata introduces Radiología Aumentada, an independent project documenting AI adoption in imaging from the Spanish-speaking world. As of October 2026, its map records 36 institutions across 14 Argentine jurisdictions; a directory surveys 59 products, with 18 showing presence or signs of presence in Argentina.

Important limitations: These publicly documented collections are not a systematic national census or an adoption-prevalence estimate. Listing does not establish regulatory approval, clinical effectiveness or endorsement.

Editorial reproduction note: The full English article follows, including references and available source figures/tables. Repeating running matter and image-only source covers are omitted. Added SJORANM branding and introductory summary; text, tables and figures were reflowed into Markdown and source-derived spacing repaired. The original PDF remains authoritative. Image paths are local review assets, not public URLs.

§Augmented Radiology: why map artificial intelligence in imaging from and for the Spanish-speaking world

Federico Martín Olivera Plata, MD¹,²

¹ Neuroradiologist, Hospital Italiano de Buenos Aires, Argentina

² Medical Director of Innovation and Science, Centro Médico Deragopyan, Buenos Aires, Argentina

Editorial role: Editor, Swiss Journal of Radiology and Nuclear Medicine ORCID 0009-0003-8514-861X · oliveraplataf@gmail.com

Accepted: 21 October 2026 · Published: 22 October 2026 · Special / Invited contribution

§Abstract

Radiología Aumentada (Augmented Radiology) is an independent personal project documenting the adoption of artificial intelligence in diagnostic imaging in Argentina and the wider Spanish-speaking world. This invited contribution discusses gaps between the predominantly English-language evidence base and local clinical practice, and describes the project’s adoption map, vendor directory, educational resources, governance guidance and opinion articles. As of October 2026, the map records 36 institutions across 14 Argentine jurisdictions; the directory surveys 59 products, of which 18 show presence or signs of presence in Argentina. The project emphasises independence, source transparency and a radiologist’s perspective, and proposes institutional surveys and international comparisons to develop locally relevant evidence on AI adoption.

Keywords: artificial intelligence; radiology; science communication; Spanish language; Latin America; technology adoption.

§The starting point

Every week a new algorithm appears to detect nodules, measure volumes or draft reports. And each time I ask myself the same questions: Is this already being used here? Who has it? Does it work? I rarely find a precise answer. Not for lack of interest, but because that information is simply not gathered anywhere.

Radiología Aumentada (Augmented Radiology) was born from that curiosity. It is a personal, independent project that aims to document, in Spanish, where and how artificial intelligence is actually used in diagnostic imaging, starting in Argentina and extending to Latin America. In this invited contribution I want to explain why I started it, what it intends to be and, above all, what it does not intend to be.

§A double gap: language and reality

The global conversation about AI in radiology takes place almost entirely in English. Trials, guidelines, registries of approved devices and market analyses are produced in the United States, Europe and Asia. For a Spanish-speaking radiologist, keeping up means translating not only words but also regulatory, economic and clinical contexts that are not their own.

The second gap is quieter. Even if one reads everything in English, that literature does not say what happens in an imaging department in Rosario, Neuquén or Tucumán. We do not know with certainty how many centres use AI in mammography, chest or neuroradiology; which vendors actually operate in the region; or how these tools are integrated into workflows with resources, connectivity and regulatory frameworks very different from those of the countries where they were validated.

This lack of data has consequences. Colleagues decide with partial information. Institutions purchase without being able to compare. And public discourse swings between uncritical enthusiasm and rejection, with no local empirical basis to discuss. Other countries have already started to address this: initiatives such as the Swiss Healthcare AI Map (1) and the Nordic AI Health Map, which documented more than 800 AI initiatives in publicly funded care across five countries (2), show that recording real-world adoption is both possible and useful. Publicly available information on AI adoption in imaging remains fragmented across the Spanish-speaking region. Radiología Aumentada is intended to help bring that information together.

§What Radiología Aumentada is

The name is deliberate. It is not about a replaced radiology, but an augmented one: the radiologist remains the one who integrates the image with the patient, and AI is a tool that can broaden — or distort — that view. Today the project has five components, and the list may grow:

An adoption map. Institutions that have publicly reported using AI in imaging, by province, with the tool they use and the source for each data point. As of October 2026, it records 36 institutions in 14 of the country’s 24 jurisdictions. (4)

A vendor directory. Which solutions exist worldwide, for which anatomical region and modality, what their regulatory status is and whether they are present in Argentina. Of 59 products surveyed, 18 show presence or signs of presence in the country. (4)

Education. A learning path for residents, a glossary and a selection of courses, books and guides to get started.

Governance. A traffic-light guide for everyday use, how Argentina’s regulatory agency (ANMAT) regulates AI software, the pillars of an institutional AI committee and the questions to ask a vendor before buying.

Opinion and analysis. Short articles, in Spanish, on relevant publications and debates in the field, written from clinical practice.

The map includes institutions publicly associated with a specific imaging AI tool in institutional reports, news coverage, vendor communications or scientific publications. The directory includes products with at least one public source and distinguishes confirmed presence from indications requiring verification. Each entry links to its supporting source; institutions and vendors can request additions, corrections or removal through the project’s contact form. (4)

The information was collected over the 12 months ending on 21 October 2026. No formal site-counting or deduplication rules were applied. The counts at that cut-off describe this publicly documented collection, not a systematic national census or an estimate of adoption prevalence. Publicity and source availability influence coverage. Absence from the map does not establish non-use, and an indication of local presence does not by itself establish regulatory approval or clinical effectiveness. (4)

Imaging is the starting point because it is my speciality and the area of clinical practice from which I can contribute most directly. Over time, the same approach could extend to other specialities and to the administrative use of language models.

§Three principles

Independence. Radiología Aumentada is a personal project. I work at two institutions and say so openly, but the project does not represent either of them, nor does it promote products. It receives no funding from vendors. Being listed in the directory does not imply endorsement, and not being listed does not imply criticism.

Evidence before enthusiasm. AI in imaging is not a homogeneous block. Some task-specific tools have been evaluated prospectively for defined uses, including AI-supported mammography screening in the MASAI trial. (5) These findings should not be generalised to other tasks or products without appropriate evidence. General-purpose multimodal models also require careful evaluation: Asadi and colleagues describe “mirage reasoning”, in which models generate detailed image-related descriptions even when no image has been supplied. (3) This preprint illustrates a failure mode, not the clinical performance of every model. Distinguishing between evidence for specific clinical tools and broader model capabilities is central to responsible communication, and I am willing to revise these positions when the evidence changes.

The radiologist’s view. Much of the communication about medical AI is written by technologists, investors or journalists. It is valuable, but incomplete. Those who report studies every day know what a false positive means at eleven at night, how the workflow changes and where the real risk to the patient lies. That is the perspective I want to contribute, and the one I believe is missing in Spanish.

§Where we are heading

A map like this cannot be built alone. A next step is to promote an institutional survey on the use of AI in imaging, ideally together with the speciality’s scientific societies, to turn impressions into data. We are also exploring comparisons with European initiatives, to understand where we are similar and where the Latin American reality demands its own answers.

That this text appears in a Swiss journal is not a paradox, but a sign. AI in radiology is a global phenomenon, but its adoption is always local. For knowledge to flow in both directions, the region needs to produce its own data and tell it in its own language, while continuing to engage with the rest of the world.

I invite radiologists, residents, medical physicists, managers and companies in the region to join: by reporting where they use AI, correcting what the map shows wrongly and proposing topics. Radiología Aumentada does not claim to have all the answers. It aims for something more modest and, I believe, more necessary: that we start asking the questions with our own data.

Radiología Aumentada: radiologiaaumentada.com.ar

§Declarations

Conflict of interest: The author is a neuroradiologist at Hospital Italiano de Buenos Aires and Medical Director of Innovation and Science at Centro Médico Deragopyan. He is a member of the SJORANM editorial board. Radiología Aumentada is an independent personal project and receives no funding. No other conflicts of interest.

Funding: This work received no external funding.

Data availability: The resources described in this invited contribution are available through Radiología Aumentada (https://radiologiaaumentada.com.ar/). Sources for the adoption map and vendor directory are identified alongside the respective entries, as described in the manuscript.

§Copyright and license

Copyright remains with the author. Submitted for publication under the Creative Commons Attribution 4.0 International License (CC BY 4.0). License terms

§Disclaimer / Publisher’s note

The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of Swiss J. Radiol. Nucl. Med. and/or the editor(s). Swiss J. Radiol. Nucl. Med. and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

§References

1. SwissMed AI. Swiss Healthcare AI Map [Internet]. Switzerland: SwissMed AI. Available from: https://mdai.ch/ai-map/ [cited 2026 Oct 21].

2. AI Sweden, et al. Nordic AI Health Map [Internet]. Funded by the Nordic Council of Ministers; 2026. Available from: https://healthmap.newnordics.ai/ [cited 2026 Oct 21].

3. Asadi M, O’Sullivan JW, Cao F, Nedaee T, Rajabalifardi K, Li FF, Adeli E, Ashley E. MIRAGE: The Illusion of Visual Understanding

[preprint]. arXiv:2603.21687v3; 2026 Apr 2 [cited 2026 Oct 22]. https://doi.org/10.48550/arXiv.2603.21687

4. Olivera Plata FM. Radiología Aumentada [Internet]. Argentina: Radiología Aumentada; 2026 [cited 2026 Oct 22]. https://radiologiaaumentada.com.ar/

5. Lång K, Josefsson V, Larsson AM, et al. Artificial intelligence-supported screen reading versus standard double reading in the Mammography Screening with Artificial Intelligence trial (MASAI): a clinical safety analysis of a randomised, controlled, non-inferiority, single-blinded, screening accuracy study. Lancet Oncol. 2023;24(8):936–944. https://doi.org/10.1016/S1470-2045(23)00298-X

§Attribution and reuse

© 2026 Federico Martín Olivera Plata. Augmented Radiology: why map artificial intelligence in imaging from and for the Spanish-speaking world. Swiss Journal of Radiology and Nuclear Medicine. DOI: 10.59667/sjoranm.v35i2.14. Original article · Original PDF.

Reproduced and adapted under the Creative Commons Attribution 4.0 International License. Changes: SJORANM cover/logo, editorial introduction, Markdown/table formatting, removal of repeating running matter and image-only cover, source-derived spacing repair, and resizing/rasterisation of article figures. Source figure credit lines remain part of the article. No scientific abridgement was intended.