98 AUSTRALASIAN DENTIST FEATURE: ARTIFICIAL INTELLIGENCE Every application of artificial intelligence (AI) in medicine should improve access to medical care and its quality, as well as increase the efficiency and safety of the respective therapy. AI also enables the sharing of patient data to support medical research, thereby reinforcing and strengthening treatment recommendations and strategies.1 Digital phenotyping already plays an invisible and pervasive role beyond medical treatment, where data is typically collected via mobile devices via health and wellbeing apps and cloud-based services. The users of such phenotypic data can select what data to collect and how and where it is stored, plus the data can then be analyzed to describe individual or collective patterns. Thus, the first ethical issue in the use of AI with phenotypic data is the democratization of this data, as the data can just be selected by the user. Essentially, every new dental workflow should have a distinct, noticeable positive effect on the patient and the corresponding The use of artificial intelligence in dentistry – how should ethics be considered? By Engelschalk, Marcus 1, 2, Smeets Ralf 2 1 SlowDigitalDentistry, Private Dental Office, Munich, Germany 2 Department of Oral and Maxillofacial Surgery, University Medical Center Hamburg-Eppendorf, 20246 Hamburg, Germany form of therapy. While the application of AI for diagnosis and therapy planning can clearly offer significant improvements in treatment, its use is also associated with risks and challenges, especially related to the inclusion of soft skills (human emotions) and hard skills (technical skills). Although AI has no ability to experience human emotions (soft skills), it can recognize human emotions based on the body language, facial expressions, text, or voice expressions of the patient. For example, in the context of a psychoemotional patient survey, AI analyzes the patient’s textual interactions using a higherlevel algorithm to control the solution search of one or more dependent algorithms and machine learning classification, producing a classification rate of 90%.2 When the patient survey analysis results of the AI model were then compared with those of psychiatrists, there was no significant difference in predicting moderate mental distress, yet the AI model was significantly more accurate in predicting severe mental stress.3 Professionalism in medicine involves more than pure knowledge-based action, and can be generalized as follows: (1) the patient has a concern; (2) the doctor attends to the patient’s concern; (3) the doctor helps without patronizing; (4) the doctor looks at the patient holistically without establishing a private relationship; and (5) the doctor applies their general knowledge to the specifics of the case. When these 5 characteristics are then applied to the use of AI, Heyen et al came to the following conclusions:4 u The use of AI in medical practice requires the doctor to pay particular attention to those facts of the individual case that cannot be comprehensively considered by AI (personality, life situation or the cultural background of the patient). u The more routine the use of AI becomes in practice, the more doctors need to focus on the patient’s concerns and strengthen patient autonomy, for example through an appropriate integration of digital decision support in shared decision-making. u Due to the fact that computer-based technologies are generally considered insensitive, the use of AI in medical professions in some areas is particularly questionable. As a result, the need for doctors to apply soft skills to maintain a humane and ethically defined doctor-patient relationship becomes even more important and should be defined urgently. Accordingly, the use of AI should initiate a new discussion in the field of ethics and the associated ethical concerns and challenges must be worked out. Basic ethical principles, which can be defined as beneficence, non-maleficence, autonomy, equity, and explainability, all require concrete recommendations, as well as reimbursement policies, security and data Fig. 1: AR (Augmented Reality) in the daily routine as a intermediary between dentist and software for the use of AI Dr Marcus Engelschalk
RkJQdWJsaXNoZXIy MTc3NDk3Mw==