【Objective:】 To understand the differences in motivation of different groups to participate in phase I clinical trials, and to provide suggestions for the recruitment and management of healthy subjects in research institutions, so as to improve the quality of clinical trials and protect the subjects’ rights and interests. 【Methods:】 Questionnaire survey was conducted among healthy subjects and potential community subjects in phase I clinical trials being carried out in a tertiary A hospital in Beijing from 2020 to 2021. SAS 9.2 was used for statistical analysis of data. 【 Results:】 There were many dimension differences between phase I ward subjects and community potential subjects in demographic and sociological characteristics, motivation to participate in the trials and influencing factors of motivation to participate in the experiment. Subjects in phase I ward paid more attention to the trial compensation, health care, social interaction and making friends, and the attitude of family or friends towards their participation in the trials. Community potential subjects were more concerned about trial quality, future potential benefits (contributing to the medical profession), and the impact on their own health and life. 【Conclusion:】 Research institutions should be aimed at different social background population stratification for recruitment, informed consent, personnel management system and the standard operating procedure, and pay attention to the subjects’ psychological and physiological needs from several aspects to standardize the management of subjects, improve the compliance of subjects to participate in the trial and the quality of clinical trials, and give more humanistic care and help to subjects.

The widespread application of medical artificial intelligence （AI） has brought technological breakthroughs to traditional diagnosis and treatment and has also altered the traditional doctor-patient interaction mode and formed a new doctor-patient relationship of “doctor-medical AI-patient” which faces a series of challenges. The intervention of AI may form a new “paternalistic style，” affecting the shared decision-making model. When its recommendations lack explanation， it may lead to clinical decision-making paralysis and affect the doctor-patient trust relationships. There may be confusion between the roles and responsibilities of AI and doctors in the process of medical practice， affecting the establishment of an emotionally responsible doctor-patient relationship. Through an in-depth analysis of the impact and causes of medical AI on the doctor-patient relationships， this paper proposed that a collaborative mechanism should be established between AI and professional doctors to complement each other， clarifying the auxiliary status of AI， reinforcing the dominant role of doctors， enhancing the regulatory mechanism， and dynamically improved the doctor-patient relationships to promote the healthy and orderly development of smart healthcare.