February 6 2023
|
Reading time 2 minutes
https://sbermed.ai/en/a-medical-expert-a-dream-profession-which-combines-it-and-health-care/
Contrary to the existing myths, development of AI-based technologies does not replace doctors. Moreover, it opens up new opportunities for health professionals. Today we would like to tell you about the profession of a medical expert, which combines the two domains and allows finding one’s place in the medicine of the future.
Firstly, the medical expert cooperates with a team of developers (analysts and testers) and answers questions which arise during their work. For instance, such an expert will tell them how to understand whether an anatomical region and an abnormal area are highlighted correctly, and what requirements to the input data should be there to ensure correct functioning of a model, and how to interpret the model results for an end user.
Secondly, to create algorithms of high quality, labeling carried out by doctors should also be analyzed. The same examination result can be labeled in different ways by two different specialists. Therefore, to improve accuracy of the data, the medical expert provides the second opinion and helps understand why the doctors’ interpretations are different. At this stage, fundamental errors can be found, or an important detail for further correct training of the algorithm can be detected. Thirdly, the expert takes part in assessing quality metrics of an IT product and provides comments on medical issues. The expert also helps determine on which images false positive or false negative results are shown, and identifies their common features (it may be necessary to take into account the age and sex of a patient, and equipment settings). Then the expert in cooperation with the developers forms a task for algorithm improvement.
Deep knowledge of the main domain and training in modern digital technologies, as well as understanding of emerging trends in the industry are necessary. The doctor should try various AI-based services and be interested in neural network training, in how experiments are made and what parameters are important for developers.
It will be useful to study the market for international medtech startups. Basic courses on artificial intelligence can be found on Coursera, and more advanced courses are provided by DeepLearning.ai. There are many books in this field. One can start reading Kai-Fu Lee, who is a world-renowned expert in AI. We also recommend lectures by Andrew Ng.