Decoding the "black-box": explainable artificial intelligence towards trustworthy advancement in respiratory medicine

Area of Science:

• Respiratory Medicine
• Artificial Intelligence
• Medical Informatics

Background:

• Artificial intelligence (AI) offers significant potential in respiratory medicine for diagnostics, treatment, and monitoring.
• Clinical adoption of AI is hindered by the "black-box" nature of algorithms, impacting clinician trust and integration.
• Explainable AI (XAI) provides transparency into AI reasoning, addressing adoption barriers.

Purpose of the Study:

• To explore the applications of XAI in respiratory medicine.
• To highlight XAI's role in enhancing transparency and clinician engagement with AI.
• To discuss the integration of XAI into clinical workflows and patient care.

Main Methods:

• This viewpoint discusses the potential applications and implications of XAI in respiratory medicine.
• It reviews the challenges and requirements for successful XAI implementation.
• It considers the role of XAI in shared decision-making and patient understanding.

Main Results:

• XAI can improve transparency, support clinician evaluation, and facilitate informed decision-making.
• Successful XAI adoption requires addressing technical, cultural, educational, and ethical considerations.
• XAI can translate complex AI outputs into understandable information for patients and clinicians.

Conclusions:

• XAI is crucial for responsible and transparent AI implementation in respiratory medicine.
• It fosters clinician engagement and supports patient-centred care.
• Continued research, education, and collaboration are essential for realizing AI's potential in respiratory care.