Machine Learning in everyday life - More than just a gimmick due to medical use

5. August 2019

For many people, Machine Learning , as the basis of Artificial Intelligence, is only known as a gimmick to win poker games for example. However, the fields of application are much more far-reaching and can offer us great added value in daily life, especially in the medical field. 

Artificial Intelligence as a winner

It became known at the beginning of July this year: For the first time, Artificial Intelligence has managed to win more poker rounds in a tournament and thus more money than five of the best players in the world. Therefore, an artificially developed game tactic was more successful than many years of human experience. 
The astonishing thing is that the software "Pluribus" could not only win, although the opponent's cards were not or only partially known, it was even able to master the poker-typical bluff. 

In many other areas of everyday life, however, Artificial Intelligence can be used much more effectively. Artificial Intelligence is based on so-called Machine Learning algorithms. Machine Learning refers to simplified computer programs that enable machines to learn by themselves. This is done with the help of a large amount of data, which is fed in and used by the programs to independently recognize similar characteristics. The first experiments in this field have been carried out since the 1960s, but technology has recently made great progress in this field. 

Machine Learning in the fight against diseases

Apart from defeating world champions in games such as poker, chess or the strategy game Go, the medical field is another exciting field of application for Artificial Intelligence.
Machine Learning algorithms offer a great advantage, especially in early cancer detection. For example, patterns found in image data can be determined more precisely by an AI program than by diagnostics performed by trained doctors. According to a group of researchers who published their results in the journal "Annals of Oncology", Machine Learning algorithms detected 95% of dangerous skin cancer tumors. In comparison, 58 dermatologists detected only 86%. The distinction between harmless moles and malignant tumors was also much more specific in the algorithms. 

The Berlin Institute for Medical Systems Biology is currently investigating the extent to which Machine Learning algorithms can be used profitably in medicine outside of image recognition. 
The amount of complex data collected over many years during examinations of individual diseases can rarely be correctly evaluated by physicians themselves. For this reason, programs are used to recognize the same characteristics and patterns in a fraction of the time that people would need. They will soon help to better predict the development of a disease and make suggestions for possible treatment. 
However, doctors are still sceptical about this development. The program can make a recommendation, but cannot provide a satisfactory and understandable explanation. 

So, a lot more research is needed in this area, before treatments will be fully related to Machine Learning programmes by doctors.



Contact Person

Josef Günthner

Co-founder & Managing Director

+49 (0) 40 180 241 180

back to top