Table 6.

Functionalities with explanation and examples

Functionality	Count	Explanation and examples
Provides quantitative information about pathology	13 (12%)	Measures the characteristics of pathologic findings.
		Example: InferRead CT Stroke detects haemorrhagic stroke, marks the location and assesses the volume to assist radiologists in their diagnosis and determining the prognosis of a patient
Marks regions of interest or detects change	38 (34%)	Detects and highlights abnormal findings visually.
		Example: VIZ LVO uses artificial intelligence to automatically identify suspected large vessel occlusion strokes on CT angiogram imaging. Change Detector compares serial magnetic resonance imaging studies and presents changes in the form of a colour-coded change map.
Provides classification, diagnosis or outcome probabilities	19 (17%)	Interprets imaging findings and provides a diagnosis or a standardized classification.
		Examples: Rapid Aspect automatically generates a standardized score, based on clinically validated machine learning algorithms, that enables communication about the extent of a patient’s ischaemic changes. Deepstroke provides ASPECT scores and outcome probabilities based on different treatments.
Prepares report	15 (14%)	Organizes the diagnostic findings in a report.
		Example: Atroscan provides reports with quantitative information through comparative analysis of the same age group.
Automated derivation of brain biomarkers	12 (11%)	Compares the quantitative information about anatomy or pathology with normal findings of a particular group.
		Example: Quantib ND provides insight into the possible presence of atrophy related to Alzheimer’s disease or other types of dementia, thus supporting more accurate diagnosis, and makes use of reference centile curves of a population-based study.
Workflow optimization and triaging	12 (11%)	Facilitates the efficacy of the diagnostic process.
		Example: RadReport states that radiologists work faster and better with diagnostic decision support and standardized reports. qER includes a triage aid to prioritize and notify critical head CT scans.
Anatomical segmentation	2 (2%)	Segments the images in normal anatomical areas.
		Example: Quibim Precision is to designed to automatically calculate the volume of brain tissues and their regions and the mapping of local cortical thickness distribution.