iiinova — where your digital self keeps evolving

AutoPiX project @autopix.bsky.social · 08/10/2026

Cloud computing is not only about computational power. In healthcare, it is also about building infrastructure that's #scalable, #interoperable, and #secure.

100

AutoPiX project @autopix.bsky.social · 08/10/2026

That is why #AutoPiX was designed from the start as an EU-based cloud solution to minimise these risks and support a more secure, trustworthy environment for working with sensitive medical data.

100

AutoPiX project @autopix.bsky.social · 08/10/2026

Large volumes of personal health data may need to be transferred and accessed remotely, often from multiple locations and by different users. This increases the importance of strong safeguards around security, confidentiality, access control, governance and compliance.

100

AutoPiX project @autopix.bsky.social · 08/10/2026

But #healthcare data are highly sensitive, and cloud-based environments also raise important questions.

100

AutoPiX project @autopix.bsky.social · 08/10/2026

Cloud computing can help improve: ➡️Data accessibility ➡️Interoperability between systems ➡️Collaboration across institutions ➡️Scalability of computing resources ➡️Integration of multiple data sources, including medical imaging, electronic health records and connected devices

100

AutoPiX project @autopix.bsky.social · 08/10/2026

In healthcare, this can make it possible to work with large and complex datasets without relying entirely on local infrastructure. It can also support the use of advanced AI and machine learning tools for applications such as diagnostic support, predictive analytics and personalised medicine.

100

AutoPiX project @autopix.bsky.social · 08/10/2026

📘 AutoPiX Glossary | New term Today’s term is #Cloud #Computing. Cloud computing refers to the use of remote, internet-based servers and services to store, manage and process data and applications.

111

AutoPiX project @autopix.bsky.social · 07/10/2026

The main goal is to bridge what we see under a microscope and on a scan. By directly comparing physical joint tissue samples (tissue pathology) with detailed, full-body images from FAPI-PET scans (molecular imaging), they can paint a clearer, more accurate picture of how the disease operates.

010

AutoPiX project @autopix.bsky.social · 07/10/2026

The QMUL team will guide these patients through 2 key procedures: 🔬Synovial biopsy --> study at a microscopic and molecular level. 🔅FAPI-PET Scan --> a heat map, showing us exactly where the inflammation is most active in the body and how severe it is.

110

AutoPiX project @autopix.bsky.social · 07/10/2026

They will focus on patients who have not seen improvement after receiving anti-TNF therapy, a common first-line medication used to reduce inflammation. When a patient reaches this stage, we need to look much closer to understand exactly what is happening inside their body.

110

AutoPiX project @autopix.bsky.social · 07/10/2026

Meet our Royal 👑partner, @Queen Mary University of London (QMUL)

111

AutoPiX project @autopix.bsky.social · 30/09/2026

A great opportunity to learn more about the role of AI in AutoPiX, how these tools are being developed, and how we envision their future use in rheumatology.

100

AutoPiX project @autopix.bsky.social · 30/09/2026

Through this conversation, we will bring together the patient perspective with the scientific and technical vision behind AutoPiX, helping us look beyond how #AI is built and focus on how it may ultimately support research, clinical practice and patient care.

100

AutoPiX project @autopix.bsky.social · 30/09/2026

Sebastien Vicent will interview Anders Nordell to explore two key questions: 🔹 How are we developing AI in #AutoPiX? 🔹 How will these AI tools be used in #rheumatology?

100

AutoPiX project @autopix.bsky.social · 30/09/2026

Today is the day! 📆 Join us for a new AutoPiX webinar taking place today, 30 September, at 15:00 (CEST) AutoPiX: Building AI for Rheumatology

101

AutoPiX project @autopix.bsky.social · 25/09/2026

Rising #RA cases, a shortage of #rheumatology specialists, and the need for consistent, objective assessments make innovations like #ARTHUR highly relevant. ARTHUR complements, not replaces, clinicians, improving access and efficiency while maintaining patient satisfaction.

010

AutoPiX project @autopix.bsky.social · 25/09/2026

🤖 What if a robot could carry out your joint ultrasound safely, accurately, and without discomfort, and patients loved it? A study in Advances in Rheumatology assessed ARTHUR, a fully automated ultrasound system that uses robotics and AI to scan the hands in #RA and scores disease activity.

120

AutoPiX project @autopix.bsky.social · 24/09/2026

Within AutoPiX, methods such as finger fold analysis demonstrate the potential of AI-supported visual tools to transform everyday images into meaningful information for rheumatology research and patient care.

110

AutoPiX project @autopix.bsky.social · 24/09/2026

Combined with other clinical and imaging data, this approach may contribute to more objective disease monitoring and earlier detection of changes over time.

100

AutoPiX project @autopix.bsky.social · 24/09/2026

Why does this matter for rheumatology? Finger fold analysis could provide a more accessible, non-invasive way to monitor joint changes, using simple hand images rather than relying solely on in-person examination or more complex imaging techniques.

100

AutoPiX project @autopix.bsky.social · 24/09/2026

Using AI models, including Convolutional Neural Networks (#CNNs), photographs of the hands can be processed to identify, isolate and measure finger folds. Changes in the shape, depth or dimensions of these folds may provide useful information about inflammation, swelling or disease progression.

110

AutoPiX project @autopix.bsky.social · 24/09/2026

These folds can change when joints become swollen or inflamed. In conditions such as rheumatoid arthritis, joint swelling is one of the key signs used to assess disease activity.

100

AutoPiX project @autopix.bsky.social · 24/09/2026

Finger fold analysis refers to the use of artificial intelligence to analyse the folds or creases on the backs of the fingers, especially around the joints.

100

AutoPiX project @autopix.bsky.social · 24/09/2026

📘 #AutoPiX Glossary Today’s term is Finger Fold Analysis.

120

AutoPiX project @autopix.bsky.social · 17/09/2026

Within AutoPiX, approaches such as LIME are relevant to our broader goal of developing AI-supported imaging tools that are not only accurate, but also transparent, interpretable and clinically meaningful.

110

AutoPiX project @autopix.bsky.social · 17/09/2026

Why does this matter for rheumatology? As #AI tools become increasingly integrated into imaging, diagnosis, monitoring and treatment support, explainability is essential for building trust and supporting responsible clinical use.

100

AutoPiX project @autopix.bsky.social · 17/09/2026

One of its advantages is that it is model-agnostic, meaning it can explain different types of AI models without requiring access to their internal structure.

100

AutoPiX project @autopix.bsky.social · 17/09/2026

For example, in a healthcare setting, LIME could help show which symptoms, imaging findings or clinical variables contributed most strongly to a particular AI-based prediction.

100

AutoPiX project @autopix.bsky.social · 17/09/2026

LIME works by creating a simplified explanation for an individual AI prediction. Instead of trying to explain the whole model at once, it focuses on one specific result and identifies which features were most important in reaching that output.

100

AutoPiX project @autopix.bsky.social · 17/09/2026

This is especially important in healthcare, where AI-supported tools may contribute to clinical decisions and where clinicians, patients and researchers need to understand why a model has produced a specific prediction.

100

AutoPiX project @autopix.bsky.social · 17/09/2026

LIME is a technique used to make complex artificial intelligence models more understandable to humans.

100

AutoPiX project @autopix.bsky.social · 17/09/2026

📘 #AutoPiX Glossary Local Interpretable Model-agnostic Explanations (LIME).

111

AutoPiX project @autopix.bsky.social · 02/09/2026

Within AutoPiX, harmonisation is an essential step towards developing AI-supported imaging tools that are robust, trustworthy and useful across different healthcare environments.

110

AutoPiX project @autopix.bsky.social · 02/09/2026

In rheumatology, harmonisation can help bring together imaging data, clinical records, laboratory results and patient-reported information to support better research and more personalised care.

100

AutoPiX project @autopix.bsky.social · 02/09/2026

Why does this matter for artificial intelligence? AI tools rely on consistent, comparable and high-quality information. When data are fragmented or collected in different ways, AI models may be less reliable or harder to apply across clinical settings.

100

AutoPiX project @autopix.bsky.social · 02/09/2026

By harmonising this information, researchers and healthcare professionals can create more unified and comparable datasets, allowing for more accurate analysis and more reliable insights.

100

AutoPiX project @autopix.bsky.social · 02/09/2026

This is particularly important when different healthcare organisations collect information in different ways, using different formats, definitions, systems or workflows.

100

AutoPiX project @autopix.bsky.social · 02/09/2026

Harmonisation is the process of integrating and standardising information from different sources so it can be used consistently and effectively in research and patient care.

100

AutoPiX project @autopix.bsky.social · 02/09/2026

📘 AutoPiX Glossary Today’s term is Harmonisation.

111

AutoPiX project @autopix.bsky.social · 26/08/2026

📺 Haven't caught up on #AutoPiX yet? Watch how we're using #AI to turn #imaging into biomarkers in #arthritis: our aims, our challenges, and what it means for #rheumatology in practice. A good excuse for some summer catch-up viewing. Subscribe: www.youtube.com/@AutoPiXproject

031

AutoPiX project @autopix.bsky.social · 12/08/2026

By combining IRT with artificial intelligence, AutoPiX aims to contribute to more accessible, explainable and clinically meaningful imaging tools for the future of rheumatology care.

120

AutoPiX project @autopix.bsky.social · 12/08/2026

Within AutoPiX, consortium partner Singularity Biomed is contributing to this field through its work on AI-supported thermographic analysis, helping explore how portable thermal imaging and advanced algorithms can support the assessment of joint inflammation.

110

AutoPiX project @autopix.bsky.social · 12/08/2026

Why does this matter for rheumatology? Inflammatory rheumatic diseases often involve local changes in joints and surrounding tissues. IRT has the potential to support a fast, objective and non-invasive assessment of these thermal patterns, complementing other imaging and clinical information.

110