10 August 2020

Artificial Intelligence Will Educate Future Doctors

Artificial intelligence

A new research and development project will ensure that doctors in a matter of months will be able to appropriate several years of experience in diagnosing skin cancer by way of Artificial Intelligence-supported feedback. Innovation Fund Denmark has invested 9.9 million DKK in the project.

AI

Skin cancer is the most prevalent type of cancer in Denmark. Each year 18,000 new cases are diagnosed and the numbers are steadily rising.

Often, skin cancer can be cured by way of surgery, if discovered in time. However, research shows that doctors need several years of training before they become sufficiently apt at recognising the signs of skin cancer.

Unfortunately, this means that many cases of skin cancer are not diagnosed until the later stages, when chances of recovery are diminished.

However, an investment of 9.9 million DKK from Innovation Fund Denmark will now enable the development and implementation of a learning programme, which with the aid of Artificial Intelligence will help improve doctors’ knowledge about skin cancer.

“There is no doubt that easy access to efficient learning will be a huge improvement for the health sector in terms of diagnosing skin cancer, but also in many other areas,” says Niels Kvorning Ternov, Project Manager and CEO of the company Mela Tech.

“In the long term, I dream of a machine that will not only be able to help doctors know more, but also help them make the right decisions,” the CEO adds. He is currently affiliated with the University of Copenhagen, where he is completing his Ph.D. at the innovation house SUND Hub.

Learning solution

The programme developed by Mela Tech is called DermLoop, and according to Niels Kvorning Ternov it is an Artificial Intelligence-based learning system for doctors who diagnose and treat skin cancer in Denmark.

The system works to ensure that each individual doctor receives automatic feedback on the elements they choose to remove, and it allows for access to a learning environment where doctors can develop their skills by diagnosing skin lesions based on a larger image library.

Simultaneously, the Artificial Intelligence will continually optimise the learning material to suit each individual doctor’s current competences. Each doctor will thus be guided by a personalised teacher, which also develops and improves its own ability to adjust the material to the individual doctor’s level of knowledge and personal manner of acquiring new knowledge.

As more data on the learning processes of doctors is gathered, the system will become smarter and better at guiding new doctors. This way, the individual doctor’s learning process will help improve the learning curve of future doctors.

Societal gain

According to Niels Kvorning Terndrup, DermLoop is expected to accelerate doctors’ learning in a way that will allow them to reach a new level of competence in only a matter of months, which, when using traditional learning methods, would otherwise take them years. 

The project will thus lead to earlier diagnosing of skin cancer and thus a fall in related deaths as well as a significant reduction of costs, which will benefit society at large.

By way example, due to uncertainty among doctors, 75 benign birthmarks are removed for each case of malignant melanoma. This level of unnecessary surgery equals a cost of over 200 million DKK for the Danish state as well as discomfort and the risk of complications for each individual patient as they undergo surgery.

Collaboration between public and private sectors

With the investment from Innovation Fund Denmark, it will be possible to bring together some of the most prominent researcher in the country, specialising in skin cancer, medical education and Artificial Intelligence.

Niels Kvorning Ternov hopes that this will enable potential ground-breaking collaborations and create a paradigm shift in medical supplementary training.

“In my experience, my medical colleagues are extremely eager and also very quick to learn, only their learning is greatly restricted by heavy workloads as well as cumbersome work processes and IT systems,” says Niels Kvorning Ternov, before adding:

“Together, these obstacles make it difficult and time-consuming for doctors to receive feedback on the results deriving from their decisions. However, with this project we see the potential for a very fruitful collaboration between the doctors and the ‘machine’, both of whom will become better and improve each other’s competences.”

The project includes collaborations between the public and the private sectors, including the company Mela Tech, the Technical University of Denmark, the Capital Region of Denmark, the Southern Region of Denmark as well as the Melanoma Institute Australia from Sydney, Australia. Over the course of three years, the partners will develop, test and implement DermLoop for the supplementary training of doctors in diagnosing skin cancer.

Contact:

Niels Kvorning Ternov, CEO i MelaTech.
Tel.: +4531789726, mail: nkt@melatech.dk