In fact, it’s true that now computers may aid dermatologists and new algorithms are able to diagnose skin cancer even better than experts in some studies using images, never in real visits in front of a patient. It’s true that these algorithms will help the performance of dermatologists and probably also for other physicians, general practitioners, or maybe in the future also for the selection of patients to go to the dermatologist. But I think that it will be very difficult to completely replace dermatologists because somebody needs to take decisions and because there are many, many limitations at this moment with computers. Computers are trained with frequent diseases and rare diseases are difficult to be trained for computers because we have very few data while doctors can have this knowledge.
It’s also true that algorithms can do a very good performance for one problem but at this moment they are not trained to solve many different problems at the same time and dermatologists can visit and diagnose skin cancer and treat psoriasis at the same time. For this reason I think that dermatologists will have jobs for a while but it’s true that also dermatologists and patients will benefit from the help of the computers.
Of course there are many human behaviours that at this moment computers cannot do – empathy with the patient; to choose a better explanation for the patient to understand what is the situation; even the application of treatments. Even it’s true that we have some rewards to help surgeons, usually the surgeon behind them doing the job. It’s true that there are many human components that are still not possible to be substituted by a computer.
We’re talking about diagnosis, moving to therapy. It will be extremely complicated to substitute a doctor but for sure the doctor will need also computers to calculate the best chance for response to therapies when we will have complex biomarkers. For sure, again, it could be aid from computers for the best decision taken together, the patient and the doctor, with the best information.
Are these systems available equally across the world?
No, in fact there are some of them that can be available for some studies, even some apps can be downloaded but without a clinical validation. It means that even though we are close I cannot say that it’s something that we can use now.
Is the potential misbalance in economic distribution a concern?
Yes, but in that point computers will help because many competitions in artificial intelligence then all the codification is open and then can allow the majority of the population to reach these algorithms. I think that it could be the contrary effect, that more people will be able to reach a good care.
Is there a potential timeline for implementation?
It started many, many years ago but now the improvement is exponential and this exponential improvement is not because algorithms are much better now, it’s because we are able to have big amounts of data, also images that can be analysed automatically by the systems and because the capacity, computational capacity, is much bigger now. It means that with this exponential growth of the computer abilities probably the implementation will be also faster than we thought.
How important is computer technology to dermatology in the future?
Computers are with dermatologists to stay and I’m pretty sure they will be necessary and in the future we cannot avoid to work with them. Some of our functions will be replaced by computers and others will be kept in the hands of dermatologists for sure.