AI in Infectious Diseases: The Role of Datasets

Cesar de la Fuente-Nunez

DOI: https://doi.org/10.1016/j.drup.2024.101067

IF: 22.841

2024-02-11

Drug Resistance Updates

Abstract:Artificial intelligence (AI) has revolutionized various domains of our society. By applying algorithms to robust datasets, AI has made breakthroughs in image recognition, image generation, and generative chatbots. In each case, the success of AI invariably relies on the availability of comprehensive datasets, such as ImageNet 1 , DALL·E 2 2 , and vast amounts of internet data. Similarly, in the field of biology, AI has recently made significant advances in protein structure prediction 3 through the use of high-quality datasets, including the protein data bank (PDB) 4 . Other areas, such as antibiotic discovery and infectious diseases, have also seen significant progress 5-15 . However, despite these advancements, the application of AI to the diagnosis and treatment of infectious diseases is still in its infancy. The main obstacle is the scarcity of reliable datasets necessary to train AI algorithms. Datasets containing data points collected under standardized conditions are, thus, urgently needed.

pharmacology & pharmacy