An Overview of Artificial Intelligence in Oncology

Figures & data

Figure 1. Artificial intelligence flywheel.

Graphic representation of the artificial Intelligence and data cycle for building effective and responsible machine learning models for healthcare.

Table 1. Artificial intelligence and precision oncology glossary.

Terms	Definitions
Algorithm	A set of rules for solving a problem or for performing a task
Area under curve	A measure of a classifier's accuracy for a binary classification
Artificial intelligence	Systems that display intelligent behavior by analyzing their environment and taking actions – with some degree of autonomy – to achieve specific goals
Artificial neural network	A computional model in machine learning, which is inspired by the biological structures and functions of the human brain
Computer-aided detection/diagnosis	Systems that use computer science to assist doctors in the interpretation of medical images
Deep learning	A subfield of machine learning that mimics the capacity of the human brain to perform unsupervised learning tasks using multiple layers of neural networks
Machine learning	A field in computer science that builds computational models that have the ability of ‘learning’ from data and providing predictions
Radiomics	A method that extracts and analyses large amounts of advanced quantitative image features with the intent of creating mineable databases from radiological images
Radiogenomics	A field that studies the correlation between cancer imaging features and gene expression

This table represents a summary of terms used in the areas of artificial intelligence combined with precision oncology [13].

Figure 2. Potential applications of artificial intelligence in a cancer patient's journey.

AI-based models can be used in preclinical (orange box) and in clinical scenarios, both before and after cancer diagnosis (green and blue boxes, respectively). In real-life oncology care, AI has the potential to optimize risk stratification, screening recommendations, diagnosis, prognosis, decision-making and treatment-related outcome prediction. Connecting clinical research to routine oncology practice by efficient drug repurposing, accelerated new treatment discovery and efficient patient matching to RCTs is another potential contribution of AI.

AI: Artificial intelligence; RCT: Randomized controlled trial.

AI: A Glossary of Terms (2019). https://doi.org/10.1007/978-3-319-94878-2

 Google Scholar