LUMC tests AI that creates wound reports from photographs
The application is intended to reduce nurses’ administrative workload, but is not yet an independent tool for treatment decisions.
Leiden University Medical Center (LUMC) is testing a computer vision system with KickstartAI that is intended to create a structured report from a wound photograph. The system is in the pilot and development phase; LUMC does not use it as a replacement for professional assessment.
Nurses currently record wound care by assessing the wound, taking a photograph, adding it to the electronic health record and completing a structured TIME form. This form describes, among other things, tissue, infection, moisture balance and wound edges. A wound-care specialist can then assess the report.
The new system is intended to automatically suggest parts of that report. KickstartAI describes a model that attempts to recognise the wound area in a photograph, measures its size and different tissue types, and translates the result into a standardised report. The nurse remains responsible for checking and recording the final information.
According to KickstartAI, the current process takes approximately five to ten minutes per wound-care session. The company cites approximately one minute as the expected duration with the model’s support. That is a target, not a result from a completed clinical study. It has also not yet been demonstrated that the application works equally well with all skin types and wound types, under all lighting conditions and with different image qualities.
The project began after the National AI Challenge 2025, in which LUMC and KickstartAI developed an application for nurses’ wound reporting. A description of the project states that the workflow was first examined in surgical departments and at a wound clinic. This was followed by model development and preparation for user testing with nurses.
The potential benefits mainly lie in the time saved and greater uniformity in reporting. If reports are structured more consistently, information can be transferred more easily between healthcare providers. The counterpoint is that an erroneous image analysis could put a nurse on the wrong track. An automated suggestion is therefore safe only if the user can check the result and easily correct discrepancies.
The test says nothing yet about improved wound healing or better treatment. This requires research examining the accuracy of the reports, the consequences for work and safety in clinical practice. Patients should therefore regard the application as a tool under development, not as a digital diagnosis or personal treatment advice.
One story, several perspectives
What is established
- LUMC and partners are testing AI for structured wound reporting from photographs.
- The application is intended to reduce administrative burdens and make reports more consistent.
- The nurse remains responsible for checking the report.
- No peer-reviewed evidence has yet been found for better treatment outcomes.
Left
Arguments AI in healthcare should serve patients and professionals, not productivity pressures. Deployment is responsible only with transparency about errors, protection of patient data and input from nurses and patients.
Values Equality, human dignity, privacy and public accountability.
Consequences Slower deployment may be necessary, but it reduces the risk that vulnerable patients bear the cost of technical errors.
Centre
Arguments A controlled pilot with clear responsibilities can be useful. The system should be assessed for accuracy, workload, patient safety and integration into existing records before wider deployment takes place.
Values Evidence, proportionality, professional responsibility and feasibility.
Consequences A phased approach can deliver administrative benefits without surrendering clinical control, but it requires additional time and resources.
Right
Arguments Healthcare organisations should be given room to test technology quickly if it frees up scarce nursing time. Over-regulation can slow innovation and increase pressure on staff.
Values Efficiency, freedom of choice, innovation and institutional responsibility.
Consequences Faster deployment can free up capacity, but only if organisations arrange liability, oversight and quality control in practical terms.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The technology and pilot phase have been confirmed by the project partners. Claims about time savings and quality improvement are described as expectations, not as proven clinical effects.
- confirmed LUMC and KickstartAI are developing a system that can suggest wound reports from photographs. — Project description by KickstartAI and Leiden Bio Science Park. source
- confirmed The system uses the TIME model for wound reporting. — KickstartAI describes translation into a structured TIME report. source
- confirmed The estimated time could fall from five to ten minutes to approximately one minute. — This was cited by KickstartAI as an expected outcome. source
- confirmed The application is not a proven diagnostic or treatment tool. — The project publications describe development and testing, not completed clinical validation. source
- uncertain The application has already been shown to improve wound healing. — No clinical outcomes study was found for this. source
Editor's note
The development, intended use and estimated time savings come from project publications by LUMC partners. No peer-reviewed clinical study was found that already confirms safety, accuracy or better treatment outcomes.Sources
- AI-supported wound reporting for healthcare — KickstartAI
- LUMC and KickstartAI develop the first AI model for faster and more consistent wound reporting — Leiden Bio Science Park
- AI improves wound care at LUMC — Leiden Convention Bureau
- Jij zorgt, AI noteert — HagaZiekenhuis
More on this in Dutch media
- de Volkskrant — „lumc wondzorg”
- RTL Nieuws — „lumc wondzorg”
- NOS — „lumc wondzorg”