Key points are not available for this paper at this time.
The application of advanced artificial intelligence (AI) models and algorithms in clinical laboratories is a new inevitable stage of development of laboratory medicine, since in the future, diagnostic and prognostic panels specific to certain diseases will be created from a large amount of laboratory data.Thanks to machine learning (ML), it is possible to analyze a large amount of structured numerical data as well as unstructured digitized images in the field of hematology, cytology and histopathology.Numerous researches refer to the testing of ML models for the purpose of screening various diseases, detecting damage to organ systems, diagnosing malignant diseases, longitudinal monitoring of various biomarkers that would enable predicting the outcome of each patient's treatment.The main advantages of advanced AI in the clinical laboratory are: faster diagnosis using diagnostic and prognostic algorithms, individualization of treatment plans, personalized medicine, better patient treatment outcomes, easier and more precise longitudinal monitoring of biomarkers, etc. Disadvantages relate to the lack of standardization, questionable quality of the entered data and their interpretability, potential over-reliance on technology, new financial investments, privacy concerns, ethical and legal aspects.Further integration of advanced AI will gradually take place on the basis of the knowledge of specialists in laboratory and clinical medicine, experts in information technology and biostatistics, as well as on the basis of evidence-based laboratory medicine.Clinical laboratories will be ready for the full and successful integration of advanced AI once a balance has been established between its potential and the resolution of existing obstacles.
Dodig et al. (Mon,) studied this question.