Development and Validation of a Questionnaire to Assess the Radiologists' Views on the Implementation of Artificial...

Area of Science:

• Radiology
• Medical Imaging
• Artificial Intelligence

Background:

• Artificial Intelligence (AI) is increasingly vital in modern radiology, yet its practical application faces challenges.
• Radiologists exhibit mixed acceptance of AI tools, necessitating a standardized assessment of their attitudes.
• Existing tools may not adequately capture the nuances of AI adoption in clinical practice.

Purpose of the Study:

• To develop and validate a questionnaire, the Attitude of Radiologists Toward Radiology AI (ATRAI-14), to assess radiologists' perspectives on AI.
• To provide a reliable instrument for evaluating radiologists' familiarity, trust, implementation views, and concerns regarding AI.

Main Methods:

• Items were derived from the European Society of Radiology questionnaire, refined through cognitive interviews with 20 radiologists.
• The ATRAI-14, comprising 12 scoring items across four domains (Familiarity, Trust, Implementation Perspective, Hopes and Fears), was pilot tested on 90 respondents.
• Confirmatory factor analysis (CFA) was employed to assess construct validity, alongside reliability and criterion validity testing.

Main Results:

• CFA confirmed the four-domain structure of the ATRAI-14.
• The questionnaire demonstrated acceptable internal consistency (Cronbach's Alpha = 0.78) and good test-retest reliability (ICC = 0.89).
• Acceptable criterion validity was observed (Spearman's rho = 0.73), indicating the tool's effectiveness in measuring AI attitudes.

Conclusions:

• The ATRAI-14 is a validated and reliable tool for measuring radiologists' attitudes toward AI in radiology.
• This questionnaire can inform management decisions regarding the strategic implementation of AI tools in radiological practice.
• Detailed insights into AI acceptance will facilitate smoother integration and optimize the use of AI in radiology departments.