Blog about skin health and skin cancer prevention

AI in Dermatology


Photo Werner Stieber-0925 (1)

OA Dr. Michael Tripolt, MPH, Dermatology and Venereology

Studied medicine at the Karl-Franzens-University Graz and Cambridge University (degree in Human Medicine 1999 and Public Health 2005). Residency at the Univ. Clinic Graz (graduated in surgery in 2006, graduated in dermatology in 2012).

Leading doctors and researchers agree, artificial intelligence (AI) will fundamentally improve medicine in the near future. With medical AI systems, patient care will become even more humane, equitable, and most importantly, efficient. Doctors can take even better care of patients and improve the healthcare system with the support of AI.

Artificial intelligence can revolutionize patient care, particularly by improving the sensitivity and accuracy of screening for skin lesions, including malignancies.

How exactly does AI in dermatology work?

AI algorithms can process huge amounts of data much faster and more accurately than humans can. This system works optimally for repetitive tasks such as image analysis.

Huge amounts of data are independently examined by the AI, looking for anomalies , discrepancies or commonalities. Another advantage of artificial intelligence is that the self-learning algorithm is constantly improving and optimizing. In other words, the more data the AI is fed, the more accurate it becomes.

This theory isconfirmed by two studies in which AI systems were examined in the classification of skin cancer.
The Nature magazine published a study from Stanford, in which AI algorithms detect skin cancer similarly competently or even better compared to dermatologists.

In another Study of the Heidelberg University Hospital the AI was allowed to compete against 157 dermatologists.

The result of the study is convincing:

the AI was better at detecting different skin cancers than 136 of the 157 dermatologists.

In both studies, the AI was trained with thousands of images and then had to distinguish between malignant skin cancer, non-malignant skin cancer and benign moles in a data set. In each case, the sensitivity and specificity of the AIs were examined in comparison to the dermatologists.

These studies demonstrate the tremendous potential of technology in assisting general practitioners and dermatologists with the assessing between benign or malignant skin lesions.

In technical jargon, the collaboration of doctor and algorithm is called Human-in-the-loop.

Better medical care

Basically, it can be said that patient care will be improved by AI systems. There is a huge gap between the supply and demand of doctors and the high cost of medical expenses.

This could be especially important for rural communities where referrals to dermatologists can be difficult.

In recent years, several smartphone apps have been launched to diagnose skin cancer. The apps use AI and often provide an instant risk assessment of the likelihood of malignant skin disease, so people see a doctor for a detailed assessment of the lesion.

Statement of the expert

Dermatologist and founder of the smartphone app SkinScreener, Dr. Michael Tripolt, MPH, confirms that artificial intelligence has many advantages in the early detection of skin cancer. In the video below, he talks about whether apps like SkinScreener can replace the dermatologist.


Mit dem Laden des Videos akzeptieren Sie die Datenschutzerklärung von YouTube.
Mehr erfahren

Video laden

"Over the course of my career, I have seen many tragic skin cancer cases, including those of very young patients. Numerous studies have demonstrated that artificial intelligence can significantly improve healthcare systems and patient care. AI can detect skin cancer with very high accuracy and offers people in rural places and limited financial resources the opportunity to detect malignant skin lesions early."

In conclusion, physicians should not view AI as a potential threat to their skills, but rather that it can be an addition to clinical practice in the years to come. An understanding of AI concepts will help practicing general practitioners and dermatologists provide decision support to take immediate action.
You are in your skin for life!

You might also be interested in...

Try SkinScreener for free

A simple and user-friendly app to assess the risk of skin cancer, based on the best AI algorithm. The download is free and includes 2 free scans!

SkinScreener cannot replace a dermatological examination. It is not giving a diagnosis and cannot be a substitute for visiting a healthcare professional.
Scroll to Top