Melanoma, “super expert” dermatologists thanks to artificial intelligence


Early diagnosis of melanoma, a very aggressive skin cancer, thanks to Artificial Intelligence (AI) with increasingly efficient evaluations and dermatologists who become ‘super experts’. The result is that priority is given to high-risk patients for rapid treatments, reducing waiting lists and thus increasing the time for the management of all those who go to the specialist.

An improvement and speeding up of the diagnosis which is also based on the recent development of digital pathological anatomy which, by means of a scanner, acquires the entire surface of the slide, transforming it into a digital file that can be read in the smallest details, remotely transferable and sharable with other specialist pathologists. Finally, sophisticated image analysis algorithms optimize and make the signals of the tissue under examination more accurately measurable. Science fiction yesterday, which became reality today, capable of revolutionizing the secondary prevention of skin cancers, improving decision-making processes and access to dermatology services.

The novelties in this field are presented during the XXVII National Congress IMI – Italian Melanoma Intergroup which is taking place from 6 to 8 November in Turin.

“Telepathology – says Daniela Massi, member of the IMI Board of Directors and Director of the pathological anatomy service of the Careggi University Hospital – allows pathologists to be able to compare themselves in real time with other specialists from all over the world before issuing a difficult diagnosis quickly providing high quality remote consultancy “.

In practice, telepathology is the key that will open the doors to teleconsultation or second opinion, which today IMI already offers on conventional histopathological preparations for atypical spitzoid tumors and difficult to interpret ambiguous melanocyte proliferations, with the most qualified specialist consultations. The next step is deep learning or deep learning, that is a computation system that autonomously manages to classify data and structure them hierarchically, finding the most relevant for solving a problem – just as the human mind acts.

“It is precisely on artificial intelligence – explains Luigi Naldi, President of Gised and Director of the complex operating unit of dermatology of the San Bortolo Hospital in Vicenza – and on a database of 10,000 images of nevi and melanomas that the project of the Study Center is based. Gised which will start by 2022 ».

The plan, which will be operational in the areas where an agreement has been signed with local health institutions, was born as an implementation of the ‘Click the neo’ app that can be downloaded in Bergamo, Vicenza, Como, Milan and Benevento. «Objective – continues Naldi – to establish, thanks to advanced processing parameters and deep learning of the operating system, if a mole is ‘not suspicious’, ‘suspicious’, ‘highly suspicious’ or ‘not evaluable’. A secondary prevention tool that could make a difference, allowing the diagnosis of this very aggressive skin cancer at an early stage ».

In practice, an ‘augmented dermatology’ which is based on the quantitative evaluations of dermatological and dermoscopic images and which can automatically identify lesions and skin structures. “With this in mind – underlines IMI President Ignazio Stanganelli, Director of Skin Cancer Unit IRST Istituto Tumori Romagna and professor at the University of Parma – the role of the dermatologist also changes, taking advantage of artificial intelligence skills, from an ‘expert’ to a ‘super expert ‘as he will be able to combine artificially generated discoveries with other clinical data collected with the patient’s medical history, always working at the highest level. Currently, literature reviews suggest caution on the use of Artificial Intelligence in early diagnosis, are still too inaccurate and are associated with a high risk of error in detecting melanoma. It is therefore essential to validate these technologies through series and large studies with numerous case series and with solid data on the standardization of algorithms supporting the diagnostic of artificial intelligence “.

A reality that, however, as regards Italy still has strong regulatory and economic constraints. “While the digitization of radiological images is already regulated – says Massi – the regulation concerning histological images has not yet been completed, thus recording a significant delay between the research phase and full clinical implementation”.

Another sore point are the costs. Large investments are needed to manage the huge amount of digital data. Costs that thanks to improved procedures and efficiency and greater clinical commitment despite the current shortage of personnel would cancel the gaps between North and South and would save many lives, reducing the health costs necessary to move to centers of excellence. But the benefits of digital telepathology don’t stop there.

«Computational pathology is already a reality – adds Massi – which, by means of automated quantification, allows for more precise and standardized diagnoses, reducing the margin of error. The next goal will be to apply artificial intelligence algorithms to diagnosis, already developed with excellent results for the diagnosis of other cancers such as prostate and breast cancer. All this, obviously, collaborating with and not replacing the anatomopathologist “.


Source: RSS DiariodelWeb.it Salute by www.diariodelweb.it.

*The article has been translated based on the content of RSS DiariodelWeb.it Salute by www.diariodelweb.it. If there is any problem regarding the content, copyright, please leave a report below the article. We will try to process as quickly as possible to protect the rights of the author. Thank you very much!

*We just want readers to access information more quickly and easily with other multilingual content, instead of information only available in a certain language.

*We always respect the copyright of the content of the author and always include the original link of the source article.If the author disagrees, just leave the report below the article, the article will be edited or deleted at the request of the author. Thanks very much! Best regards!