Artificial Intelligence in the Medical Sector
Artificial Intelligence progressively revolutionizes all economic sectors of our neoliberal societies. The medical sector is no exception: A.I and Big Data help doctors and medical staff in many aspects of their work.
Indeed, NLP, Machine Learning and Deep Learning researches have created new useful models for detecting symptoms, diseases or creating effective assistance robots, or even for drug development. Thus, A.I already helps in some medical tasks:
- - For diagnosing thanks to semantic analysis of reports and existing diagnoses (especially through digitalised medical files),
- - For helping with breast cancer by spotting symptoms,
- - For creating computer-assisted robots for surgery (to improve the precision of gestures or remote surgery),
- - For processing medical images,
- - For helping with diagnosis (see the Watson model produced by IBM).
The use of A.I in the medical sector can not only reduce error rate, but it is also a great asset in the oncology field, and it doesn’t need any rest.
However, with the use of such technologies, some problems arise: the medical sector demands perfectly clean, annotated and unbiased data for a digital approach. Yet, it is extremely cumbersome and costly to obtain that kind of data. That is even more true as medical data are mostly textual and demand NLP research to be treated. Besides, A.I research groups need to be scrupulous and pay attention to data protection, especially in the health field in which patient confidentiality needs to be absolutely protected. The last main issue raised by the use of A.I is the lack of transparency concerning technology decisions: doctors don’t have access to the decision process of the A.I, because the technologies of neural network and Deep Learning work in mysterious ways.
Advantages of A.I in the medical field are undeniable. However, it is essential to maintain relationships to the individual, because it is an essential aspect of the medical sector. A.I should not be set up with the intention to replace the medical staff and doctors, on the contrary, it should be set up to support and limit the error risks. The responsibility remains in the hand of the doctor who takes all the final decisions regarding the diagnosis and the appropriate treatment for the patient. Humans are not ready yet to be cared by robots, but they will become more and more important in the health sector in the years to come.