| Types of artificial intelligence (AI) |
Working mechanism |
Common applications in surgery |
| Machine learning |
Broad subset of AI aimed at automating analytical modeling and improving the findings without being explicitly programmed to do so. |
To calculate predictions, accuracy levels, and analysis of various outcomes and interventions. |
| Artificial neural networks |
Mimics “biological neuronal synapse” by interconnected computer processors to process information. |
Subset of machine learning. |
| Reinforcement learning |
Employs “trial and error” to come up with a solution in order to maximize the notion of cumulative reward. |
Subset of machine learning. |
| Deep learning |
Mimics “human brain” to progressively extract higher-level features from the raw input with no or minimal human supervision. |
Subset of machine learning. |
| Natural language processing |
Give machines the ability to help understand, interpret, and manipulate human language to extract information and insights contained in the data. |
Automates manual assessment process, especially in the records system and to spot quality lapses. |
| Computer vision |
Mimics “human visual system” to derive meaningful information from digital images, videos, and other visual inputs. |
Acquisition and interpretation of visual images/videos, leading to image-guided and virtual surgery. |
| Fuzzy logics |
Associates degrees of likely possibilities instead of a hard yes-no type decision to handle vagueness and imprecise information. |
Assist decision-making and performance assessment of healthcare providers. |
| Robotics and cybernetics |
Transdisciplinary approach to build automatic control systems. |
Training, navigating, and assisting healthcare professionals with a minimally invasive approach. |
