
Validating AI in classifying colorectal cancer in real-time surgery. Funded by @HorizonEU. Tweets reflect only the views of the project owner.
Joined May 2022
- Tweets675
- Following774
- Followers934
- Likes706
Pinned Tweet
πͺπΊ CLASSICA | Validating AI in Classifying Cancer in Real-Time Surgery
π‘ Five leading cancer surgery centres across Europe will treat 500 patients during the project.
Check out our website to learn more
π classicaproject.eu/
#HorizonEU #classica #ai #colorectal #cancer
π£ Interested in seeing CLASSICA's platform in action?
πΉ Check out our new educational video from project partners @ucddublin, @IrcadFrance and Arctur.
Prof @RonanCahill15 (University College Dublin), Prof John Burke (Beaumont University Hospital), Ed Murphy (University College Dublin) and @samo_erzen (Arctur) share how CLASSICA βs next-generation AI platform can enhance detection of cancer in real-time in surgical settings.
classicaproject.eu/2026/09/cβ¦
πΉ Check out our new promotional video from project partners @ucddublin and @IrcadFrance!
The video showcases the main elements of the CLASSICA project. CLASSICA transforms fluoresecence imaging into real-time clinical intelligence that can be used in operating rooms to improve colorectal cancer treatment decisions. The innovative Artificial Intelligence tool developed by the project is now under multi-centre clinical evaluation.
classicaproject.eu/2026/09/cβ¦
CLASSICA clinical partner Prof. Felix Aigner @FelixAigner2 was at the European Federation for Colorectcal Cancerβs 14th EFR Congress 2026 in Vienna last week. The EFR Congress serves as a central meeting point for experts in surgery, oncology, and gastroenterology. The theme of the congress this year was βBeyond Boundaries: From Microbiome Research to Robotic Surgery in Rectal Cancerβ. The event hosted more than 100 participants joined by 40 speakers over 3 days.
Prof Aigner spoke about βTransanal Surgery for Rectal Cancer: From Basic Techniques to Artificial Intelligenceβ and highlighted the CLASSICA project.
π Knowledge Unwrapped: πͺπΊ CLASSICA project's publication lay summary series looks at the latest paper from the team!
A recent CLASSICA paper, Privacy Considerations of Artificial Intelligence Scribes, in NEJM AI examined how Artificial Intelligence (AI) scribes handle, store, and destroy data acquired during a patient encounter.
Authors Dr Sara Gerke and Prof David A. Simon made recommendations for hospitals to consider when adopting these tools, including training procurement teams, disclosing and obtaining consent for use of AI scribes, and developing data governance policies.
Read our capsule summary of the publication here ππ½
classicaproject.eu/2026/06/kβ¦
π15β16 June 2026, CLASSICA partners participated in two days of productive discussions, knowledge exchange, and collaboration, as well as a dedicated exploitation workshop focused on defining the next steps towards ensuring the long-term impact and sustainability of the project's results.
πA big thank you to or hosts - we look forward to the exciting work ahead!
#HealthcareInnovation #artificialintelligence #machinelearning #MedicalResearch
Training is an important part of CLASSICA's dissemination work. Coordinator Prof Ronan Cahill joined the European Society of Surgical Oncology (ESSO) - International Society for Fluorescence Guided Surgery (ISFGS) Course in Leiden on 11th June 2026, where he spoke about the kinetics of fluorescence in cancer imaging.
#HORIZONEU #SurgicalInnovation
Read our summary of a @classicaproject study in the Journal of Law and the Biosciences examining whether Europe's landmark AI law will actually make AI systems fairerβparticularly in healthcare. Authors Emilia Niemiec, Peter A E Davis (both at CeBIL, @UCPH_Research), and Mathias K Hauglid (Wikborg Rein Advokatfirma AS) take a close look at the EU AI Act's rules around training data, asking whether they're strong enough to tackle one of AI's most persistent problems: biased datasets.
πclassicaproject.eu/2026/03/kβ¦
CLASSICA retweeted
Great to see @classicaproject featured in the European Commission #Science4EU portal: research-and-innovation.ec.e⦠@UCDMedicine @ucddublin
π¬ New Surgical Endoscopy publication summary from the @classicaproject team, π·πππππππππ πππππππ π
πππππ πππππππππ πππ
πππππππ ππππππππππππ: πππ ππππππππππππ πππππππππππ ππππππππππ πππππππ ππππππππππππππ ππππππππππ?
Peritoneal opening during TEM/TEO surgery is largely predictable, and when well-managed, doesn't increase overall patient risk. Important findings for safer surgical planning. π₯
π @albertoarezzo, Ammirati, Distefano et al., @unito
#MinimallyInvasiveSurgery #Colorectal #TEM #KnowledgeUnwrapped
π classicaproject.eu/2026/03/kβ¦
CLASSICA retweeted
Now published open access online @SurgEndosc βPerioperative splanchnic perfusion variation around colorectal surgery using both indocyanine green spectrophotometry and fluorescence angiographyβ @ucdmedicine @jeffreydalli @Ashokk_s
access via the linkβ¦links.springernature.com/f/aβ¦~
π©Ί New evidence in #TEM surgery
What predicts peritoneal opening during transanal endoscopic microsurgeryβand does patient positioning matter? A Feb 2026 study in Surgical Endoscopy from @unito (@albertoarezzo et al.) shows that peritoneal opening during TEM is anatomically predictable based on lesion features and positioning.π Key takeaways:
β’ Predictable risk based on anatomy & positioning
β’ Higher conversion rates & longer LOS
β’ No significant increase in morbidity when promptly recognised and managedπ¨ββοΈ Authors: Alberto Arezzo, Carlo Alberto Ammirati, Giovanni Distefano, Michele Barbiero, Simone Arolfo, Mario Morino
π Read more ππ½ classicaproject.eu/2026/03/cβ¦
π Article ππ½ link.springer.com/article/10β¦
#ColorectalSurgery #MinimallyInvasive #SurgicalEndoscopy #PatientSafety #TEM
In an August 2025 πͺπΊ @classicaproject publication in npj Digital Medicine, 'π¬πππππππ ππππππ πππ
ππππ π
πππππππππ
ππππππ π
πππ πππππππ πππ
πππππππ ππππππππππ,' CLASSICA PI @TiMinCeBIL, @CeBIL_Center, @UCPH_Research, with co-authors Cindy Welzel, TU Dresden, Max Ostermann, Technische UniversitΓ€t Dresden, Hannah Louise Smith, University of Copenhagen (KΓΈbenhavns Universitet), @toralf_kirsten, Leipzig University, and Stephen Gilbert, Technische UniversitΓ€t Dresden, presented a conceptual framework for secure and self-determined health data sharing, discussing how emerging digital technologies can support trustworthy consent management. This research aims to explore how people can gain meaningful control over how their health data, which can be generated by apps and wearable devices, is shared and used for healthcare and research.
Why is this research important? Check out our easy-to-digest, βKnowledge Unwrappedβ summary ππ½ classicaproject.eu/2026/02/kβ¦
How does the EU AI Act address dataset quality and bias in AI systems, and what are the practical and legal implications of these provisionsβparticularly regarding data governance, the "debiasing exception" for sensitive data, and enforcement mechanisms? βοΈ
Authors of the Journal of Law and the Biosciences article, "πΎπππ πππ π¬πΌ π¨π° π¨ππ ππππ ππ ππππππππ π
ππππππ ππππ ππ πππ
ππππ π¨π°?", @classicaproject researcher Emilia Niemiec, @CeBIL_Center, University of Copenhagen, Peter Davis, CeBIL, U of Copenhagen, and Mathias K. Hauglid, Wikborg Rein Advokatfirma AS, examine how Europe's new AI law tries to make AI systems fairer by setting rules about the data used to train them, requiring representative datasets and even allowing use of highly-protected sensitive data specifically to fix bias problems. The authors analysed whether these rules will work in practice, especially in medical AI, given the tension between accessing data to check for bias and protecting patient privacy.
Read on πclassicaproject.eu/2026/02/cβ¦
π£ Publication news! @CLASSICAproject explores how AI analysis of fluorescence perfusion can support in-situ classification of significant rectal polyps.
π "π¨πππππππππ ππππππππππππ ππππππππππππππ ππ ππππππ πππππππππ ππ πππ
πππππππ ππππππππππππ πππππππππ ππππππππ" by @PtrckBlnd et al. in @SciReports classicaproject.eu/2026/02/cβ¦
#HorizonEU #AI #colorectalcancer
What can legal precedents from surgical technology adoption teach us about preserving patient autonomy through informed consent as AI diagnostic and decision-support tools become standard of care? In a Perspective titled, π³ππππ ππππππππππππ ππ π¨π° πππππ
πππ
ππ ππππ πππππππππππ ππ ππππππππβ ππππππππ
πππππππ: πππππππ ππππ πππππππ,Β by Mindy N. Duffourc @MND_Law, Maastricht University and @classicaproject Coordinator Professor Dr @RonanCahill15, @ucddublin, published in @npjDigitalMed, legal cases addressing consent and surgical innovation provide a framework for understanding how AI integration, particularly diagnostic and clinical decision-support tools, will affect the informed consent process and patient autonomy.
πclassicaproject.eu/2026/02/cβ¦
Can AI Measure Surgical Skill? CLASSICA Says Yes.
Knowledge Unwrapped: πͺπΊ @classicaproject researchers used computer vision to analyse how surgeons place stitches, investigating whether AI can objectively measure suturing quality. Testing three surgical platforms across different experience levels, the team discovered that irregular stitch spacing correlates with weaker sutures. While spatial patterns alone can't fully predict suture strength, this breakthrough moves us closer to intelligent surgical systems that provide real-time feedback during operations, transforming both training and quality assurance.
Lead author Professor @albertoarezzo and @unito colleagues just published this pivotal research in Surgical Endoscopy: "ππππ πππππππππ ππ ππππππππ: ππππππππ ππππππ πππ ππππππ πππππππ ππππππππππβππ ππ ππππ πππππ ππππ
π."
Read the open-access paper in Surgical Endoscopy, or our lay summary β‘: classicaproject.eu/2026/02/kβ¦
CLASSICA retweeted
Replying to @classicaproject
I am truly proud of this studyβnot only for its results, but for the direction it points to.
With this work, we show that computer visionβderived spatial metrics of suturing can be meaningfully linked to real mechanical performance (burst pressure). Spatial features alone are not yet sufficient as standalone predictors of suture qualityβbut that is not the main message.
The real takeaway is that precision can be quantified, compared, and ultimately automated.
This, in my view, is the real turning point for robotic surgeryβincluding laparoscopic robotics.
The future is not about vaguely claiming βgreater precision.β The real breakthrough will come when the robot itself is able to plan and execute optimal trajectories autonomously, completing a suture in true task autonomy.
Crucially, this autonomy must be human-in-the-loop:
the robot operates autonomously, but always under human supervision and control, with the surgeon retaining authority, responsibility, and the ability to intervene at any time. This is not about replacing the surgeonβit is about augmenting human skill with measurable, reproducible precision.
This is the approach robotic surgery advocates should take to demonstrate real value in the operating room:
β’quantitative metrics,
β’verifiable automation,
β’objectively demonstrable technical benefits,
rather than generic statements about precision.
At MITIC-Lab, we believe this circle can close very soon. Within the year, at least in the laboratory setting, we expect to see genuine integration of automated routines into robotic surgical gestures.
At that point, the benefit of robotic technology will no longer be a matter of opinion.
It will be evidence-based.
Research on surgical automation is now becoming truly exciting, concrete, and useful. And this is only the beginning.
Can spatial metrics of suture placement quantitatively reflect mechanical performance, and how do operator experience and surgical platform influence technical outcomes in minimally invasive surgery? In the @classicaproject publication, authors Roberto Spagnulo, Francesco Marzola, Federica Corso, Giovanni Distefano, Matteo Pescio, Kengo Hayashi, Federica Barontini, Giulio Dagnino, Kaspar Althoefer, Bruno Siciliano, Sebastien Ourselin, and PI Professor @albertoarezzo of the @unito evaluated whether computer vision-derived spatial features of suture placement could predict mechanical strength (measured by burst pressure) across three surgical platforms with surgeons of varying experience levels. While spacing irregularity showed negative correlation with burst pressure, spatial metrics alone proved insufficient as standalone predictors of suture quality, though the approach shows potential for developing real-time feedback systems for surgical training and quality assurance. This paper, ππππ πππππππππ ππ ππππππππ: ππππππππ ππππππ πππ ππππππ πππππππ ππππππππππβππ ππ ππππ πππππ ππππ
π, was published on December 4, 2025, in Surgical Endoscopy.
Read the paper πlink.springer.com/article/10β¦
CLASSICA retweeted
π΅πΉ Today we meet in the beautiful Porto for day 1 of the @CLASSICAproject plenary meeting.
We will open the meeting with discussions on our #AI classifier and results from in-situ endoscopic classification of rectal neoplasia.
π Check out our recent article in press, π¨πππππππππ ππππππππππππ ππππππππππππππ ππ ππππππ πππππππππ ππ πππ
πππππππ ππππππππππππ πππππππππ ππππππππ nature.com/articles/s41598-0β¦
@EU_HaDEA #HorizonEU
#AI is transforming healthcare, but who's liable when it gets a diagnosis wrong?
Professor @gerke_sara's @classicaproject research on "nutrition labels" for AI medical devices is sparking crucial conversations in @MedEconomics and beyond about transparency, accountability, and the future of medical malpractice.
Read a recap of the recent media flurry βοΈ here: classicaproject.eu/2026/01/pβ¦

















