Artificial intelligence (AI) applications are an area of active investigation in clinical chemistry. Numerous publications have demonstrated the promise of AI across all phases of testing including preanalytic, analytic, and postanalytic phases; this includes novel methods for detecting common specimen collection errors, predicting laboratory results and diagnoses, and enhancing autoverification workflows. Although AI applications pose several ethical and operational challenges, these technologies are expected to transform the practice of the clinical chemistry laboratory in the near future.
Investigation of artificial intelligence applications for the clinical chemistry laboratory has proliferated in recent years.
Applications for artificial intelligence spanning all phases of testing (preanalytic, analytic, postanalytic) have been documented.
Novel applications including uses in laboratory result, diagnosis, and risk prediction may transform the role of the clinical chemistry laboratory.
Although relatively few artificial intelligence applications have been fully implemented in the clinical chemistry laboratory, the field is maturing quickly.
Important regulatory and ethical considerations must be addressed as chemistry laboratories look to incorporate artificial intelligence technologies in routine practice.
Investigation of artificial intelligence (AI) applications in health care has accelerated rapidly in recent years. Accordingly, there are also rapid advancements in this area applied to the clinical laboratory. In this article, advancements in AI relevant for clinical chemistry laboratories are explored.
It is important to begin with a brief classification of AI as it will be understood for the purposes of this article. AI can be classified as general (all purpose and task independent) or narrow (focused and task targeted). Although an area of active investigation, general AI does not presently exist as defined. AI as is experienced today exists as narrow applications.
Narrow AI is a term that encompasses expert systems (ie, rules based) and machine learning (ML). Expert systems AI employs the use of rules that are explicitly programmed by a human expert (eg, defined autoverification [AV] rules). In contrast, ML-based AI achieves a similar end point but arrives there through a different approach. Relationships between narrow AI types including representative examples can be found in Fig. 1 .