Skip to main content

Table 3 Characteristics of included studies (N = 86)

From: Value assessment of artificial intelligence in medical imaging: a scoping review

Element in study

Categories

Frequency (%)

Country of 1.author

Australia

3 (3%)

Asia and the Middle East (China, Hong Kong, India, Israel, Japan, Saudi Arabia, Singapore, South Korea)

16 (19%)

Canada

6 (7%)

Europe (Cyprus, Denmark, Finland, France, Germany, Italy, Netherland, Portugal, Spain, Sweden)

23 (27%)

UK

8 (9%)

US

26 (30%)

Unclear

1 (1%)

Other (Norway, Switzerland)

3 (3%)

Clinical area covered

Breast cancer (mammography and digital breast tomosynthesis)

9 (10%)

Dementia/Alzheimer´s disease/neuroimaging

7 (8%)

Dermatology (melanoma diagnosis, histopathologic images)

2 (2%)

Cardiovascular disease (cardiovascular imaging, coronary artery disease)

8 (9%)

Diabetes and ophthalmology (ocular imaging, diabetic retinopathy screening, diabetic eye disease screening)

6 (7%)

Oncology and radiotherapy

8 (9%)

Radiology

15 (17%)

Radiomics

11 (13%)

Medical imaging

12 (14%)

Pathology (histopathology images)

2 (2%)

Other

6 (7%)

Study type

Actual evaluation

15 (17%)

Guidelines, statements or frameworks

10 (12%)

Reviews, surveys or papers voicing future needs or challenges

61 (71%)

Year of publication

2016

0 (0%)

2017

11 (13%)

2018

9 (10%)

2019

37 (43%)

2020 (mid-September)

29 (34%)

Domain mentioned or perceived relevant

1. The health problem and current use of technology

55 (64%)

2. Technology aspects

63 (73%)

3. Safety assessment

17 (20%)

4. Clinical effectiveness, e.g. clinical outcomes

39 (45%)

5. Economics

52 (60%)

6. Ethical analysis

25 (29%)

7. Organisational aspects

53 (62%)

8. Patients and social aspects

33 (38%)

9. Legal aspects

43 (50%)

10. Development of AI algorithm, performance metrics and validation

67 (78%)

11. Other aspects

2 (2%)