医患信任对于构建和谐的医患关系至关重要，而患者对医生信任的建立是其核心。随着人民群众健康需求的日益增长及人工智能的快速发展，人工智能在医疗中的应用日渐增多。然而，人工智能医疗应用，如何影响患者对医生信任具有较强的不确定性和复杂性，针对该问题的研究亟待深入进行。 本文基于系统的文献综述及分析，梳理了人工智能在医疗领域的应用现状，总结了患者对医生信任的概念和度量工具，探究并明确了本文研究中关于人工智能医疗应用中患者对医生信任的涵义。围绕患者对医生信任这一问题，将人工智能医疗应用分为诊断、治疗和康复三个典型环节并开展研究，对比了传统的医生和患者两者关系转变为医生、患者和人工智能三者关系的异同，分析了人工智能医疗应用对医生和患者关系的影响。探讨了研究人工智能医疗应用中患者对医生信任的理论视角，综合考虑医方、患方和人工智能三个因素，基于国家、组织和个人三个层面构建了本文的研究框架，为深入开展实证检验及机制分析创造了条件。 基于充分、深入的调研访谈（调研了北京11家代表性医院；访谈了131位代表性患者，前后对照，满足样本量要求），本文开展了人工智能医疗应用影响患者对医生信任的实证检验。针对本文研究中的人工智能医疗应用，基于主题分析法的研究结果表明：在诊断环节，通过效率、非效率和技术应用三个主题，没有明显促进患者对医生信任；在治疗环节，通过偏好、信息和心理三个主题，一定程度上有利于促进患者对医生信任；在康复环节，通过症状、身份、专业知识和非专业知识四个主题，有利于促进患者对医生信任。进一步地，分析了各环节实证检验结果的差异性，探讨并给出了基于患者对医生信任的医生、患者和人工智能维度表。 基于构建的研究框架和实证检验研究结果，本文从国家、组织和个人三个层面，特别是着眼于人工智能医疗应用相关的政策类型、医院的实施方式和医生的理解，剖析了人工智能医疗应用影响患者对医生信任的机制。研究结果表明，在诊断、治疗和康复三个典型医疗环节，国家层面的政策类型和组织层面的应对方式，会影响医生在个人层面对人工智能作为辅助工具的理解，进而影响其为患者提供医疗服务时的行为，从而影响了患者对医生信任。影响机制分析合理解释并拓展了实证检验研究结果。进一步地，提出了人工智能医疗应用中，如何促进患者对医生信任的对策建议，可为和谐医患关系的建立提供参考。

Trust between patients and doctors is crucially important to achieve harmony in healthcare, while the establishment of patients’ trust in their doctors (patient-doctor trust) is the core. As the public demand for healthcare increases and artificial intelligence (AI) in healthcare develops, there is a rise in AI applications in healthcare. However, the application of AI brings uncertainty and complexity to patient-doctor trust. In-depth research on this needed to be studied promptly. In this study, on the basis of a comprehensive literature review, recent AI applications and breakthroughs in healthcare are outlined. The concepts and measuring tools of patient-doctor trust are summarized, and the meaning of patient-doctor trust in the application of AI in this study is then discussed and clarified. Regarding the research questions in this study, three typical healthcare phases - diagnosis, treatment and recovery, are identified. The similarities and differences between the traditional doctor-patient relationship (a two-actor relationship), and doctor, patient and AI relationship (a three-actor relationship) are compared. Thus, the influence of the application of AI on the relationship between doctor and patient is analyzed. Theoretical viewpoints of the study on doctor-patient trust in the application of AI are discussed and analyzed. By considering three factors of doctor, patient and AI, through the lens of the national level, the organizational level and the individual level, the theoretical framework in this research is purposed, which lays a foundation for the following empirical study and influencing mechanism analysis. On the basis of sufficient and in-depth investigation and interview (11 representative hospitals in Beijing are investigated, and 131 representative patients are interviewed, before-after test, the sample size in this empirical study can reach saturation), the empirical research is conducted on the influence of AI in healthcare on patient-doctor trust comprehensively. According to the limited AI application stated in this research, thematic analysis is applied to analyze the collected empirical data. The analysis results show that, in the phase of diagnosis, three main themes are identified: efficiency-oriented, non-efficiency oriented and technical issues, which do not significantly improve the patient-doctor trust; in the phase of treatment, three main themes are identified: patients’ preference, information and psychology, which improve the patient-doctor trust to some extent. in the phase of recovery, four main themes are identified: symptoms, identities, technical knowledge and non-technical skills, which improve the patient-doctor trust. Discrepancies among the empirical results of the three phases are compared and analyzed, and a dimension table based on patient-doctor trust in terms of patient, doctor and AI in the three phases is then discussed and proposed. On the basis of the theoretical framework proposed and the above empirical results, through the national level, the organizational level and the individual level, especially from types of government policies, organizational behaviors of hospitals and doctors’ personal perspectives on AI, the influencing mechanism of application of AI on the patient-doctor trust is analyzed regarding the phases of diagnosis, treatment and recovery respectively. The research shows that types of policies and organizational behaviors will affect how doctors treat AI in the provision of healthcare and, correspondingly, how they provide service to their patients, which thus affects the patient-doctor trust. Policy suggestions aim to improve patient-doctor trust in the application of AI in healthcare are then provided, for the purpose of improving patient-doctor relationship.