JMIR Publications

ABSTRACT

Background:

Laboratory tests almost always have their results presented separately as individual values. Physicians, however, need to analyse a set of results to propose a supposed diagnosis, which leads us to think that sets of laboratory tests may contain more information than those presented separately for each result.

Objective:

In this sense, we seek to identify scientific research that uses laboratory tests and machine learning techniques to predict hidden information and diagnose diseases.

Methods:

The methodology adopted used the PICO principles (population, intervention, comparison and outcomes), searching the main Engineering and Health Sciences databases.

Results:

Following the defined requirements, 40 works were selected and evaluated, presenting good quality in the analysis process. We found that in recent years, a significant increase in the number of works that have used this methodology, mainly due to COVID-19. In general, the works used machine learning classification models to predict new information, and the most used parameters were data from routine laboratory tests, such as the complete blood count.

Conclusions:

Finally, we conclude that laboratory tests, together with machine learning techniques, can predict new tests, thus helping search for new diagnoses. This process has proved to be advantageous and innovative for medical laboratories. They are making it possible to discover hidden information and propose additional tests, reducing the number of false negatives and helping in the early discovery of unknown diseases.