Thanks to the Danish start-up software developing company MelaTech, what started as an idea for better clinical feedback for the General Practitioner has now grown into 4 clinical modules for all doctors involved in skin cancer diagnostics.
The lesion registration module for the clinician
DermLoop Capture allows the clinician to take clinical and dermoscopic images of skin lesions and register lesion and patient data easily. DermLoop capture is the foundation and gateway for the other DermLoop modules. Read More
The AI enhanced educational platform
6 years from novice to expert is in the past.
Starting with a library of 20,000 skin lesions with clinical and/or dermoscopic images, the user can quiz in skin and mole cancer. The AI continually re-picks the presented cases for the user in order to optimize the learning curve. The newly expert-revised written materials will be tested so that only evidence-based learning materials are used for clinical studies. Read More
The pathologists module
The histo-pathological diagnostic decision can be highly influenced by the clinical information, and easy access to this information is therefore crucial for the pathologists to make an informed decision.
DermLoop Pathology lets the pathologist see all registered data on the patient and lesion, thereby improving the diagnostic decision. Read More
The reviewers and skin specialists’ module
Allowing the general practitioner to have patients with skin-lesions of undetermined malignancy have their skin lesions checked by the dermatologists via Teledermatology in stead of having to get an appointment saves time and shifts resources from the healthy patients to those in actual need.
The Artificial Intelligence behind the enhanced educational system
How it works and why.
Doctor and AI Collaboration
Artificial Intelligens, Machine Learning, Deep Neural Network, Embedded Neural Network. There are many names for the advanced statistical algorithms involved in AI. What AI does is find connections and correlations between what you give it and ask it to find, i.e. pictures of skin lesions and diagnosis. However, how it finds these correlations is more difficult to know, and many therefore see this type of algorithms as intelligence, as the AI can interpret amounts of data that the human mind cannot even begin to fathom.
In AISC we wish to use the incredible capabilities of AI to enhance the learning and education of the doctors. We do not wish to relie on the diagnostic capabilies of the algorithm, which is also why we do not want to release an AI algorithm for public and patient use. With only a few percentage of false diagnosis, the GP and Dermatologists office would be swarmed with healthy patients, and the patients in actual need of care risks being falsly reasured that nothing is wrong.
Having a mobile app tell you that you might have cancer is no way to get such serious news. The diagnostic process belongs to the doctor, as well as the responsibility for the patient.
When the photographic database of these research projects have grown large enough (hundreds of thousands of histopathology or expert-dermatologist verified photographs) the AISC AI can use the captured photographs of skin lesions to enhance and evolve its algorithms of what distinguishes e.g. a nevi from a melanoma.
+45 3868 9236
The Research Unit at the Department of Plastic Surgery
Ground Floor, Entrance 7
Borgmester Ib Juuls Vej 7