Brad Partridge, Monika Janda, Nicole Gillespie, Carina Vasconcelos Silva, Chris Arnold, Lisa Abbott, Tony Caccetta, H Peter Soyer
{"title":"Attitudes Towards the Use of Artificial Intelligence in Dermatology: A Survey of Australian Dermatologists.","authors":"Brad Partridge, Monika Janda, Nicole Gillespie, Carina Vasconcelos Silva, Chris Arnold, Lisa Abbott, Tony Caccetta, H Peter Soyer","doi":"10.1111/ajd.14524","DOIUrl":null,"url":null,"abstract":"<p><strong>Background/objectives: </strong>This study explored the views of dermatologists in Australia on the use of Artificial Intelligence (AI) in dermatology.</p><p><strong>Method: </strong>Fellows and Trainees of the Australasian College of Dermatologists (ACD) were invited to participate in an anonymous online survey, resulting in a sample of 122 completed surveys (response rate 16.2%).</p><p><strong>Results: </strong>Although 44% have used AI in their dermatology practice, only a minority are using AI regularly for clinical (12%) or administrative purposes (17%). A key barrier is trust, with most (69%) participants either unwilling or unsure about trusting AI for supporting the diagnosis of skin cancer. Participants identified accuracy and information on datasets, limitations, benefits, and the purpose of AI as important for trust. 52% want AI accuracy to be equivalent to or superior to the best dermatologist if it is used for clinical diagnosis. Participants perceived a range of benefits and risks of AI, with key risks around accuracy and divestment of AI to tech companies, and key benefits being reduction of monotonous tasks and improved patient access. Few (10%) are concerned about AI replacing dermatologists even though almost half (47%) expect key aspects of dermatology work will be performed by AI.</p><p><strong>Conclusions: </strong>Dermatologists in Australia are at an early stage of integrating AI into practice, with most wary or unsure about the accuracy of currently available AI tools for diagnostic purposes. Developing workflows that are acceptable to clinicians may require knowledge outreach from Dermatology Colleges, such as the ACD, to help clinicians develop well informed views on AI.</p>","PeriodicalId":8638,"journal":{"name":"Australasian Journal of Dermatology","volume":" ","pages":""},"PeriodicalIF":2.2000,"publicationDate":"2025-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Australasian Journal of Dermatology","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1111/ajd.14524","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"DERMATOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
Background/objectives: This study explored the views of dermatologists in Australia on the use of Artificial Intelligence (AI) in dermatology.
Method: Fellows and Trainees of the Australasian College of Dermatologists (ACD) were invited to participate in an anonymous online survey, resulting in a sample of 122 completed surveys (response rate 16.2%).
Results: Although 44% have used AI in their dermatology practice, only a minority are using AI regularly for clinical (12%) or administrative purposes (17%). A key barrier is trust, with most (69%) participants either unwilling or unsure about trusting AI for supporting the diagnosis of skin cancer. Participants identified accuracy and information on datasets, limitations, benefits, and the purpose of AI as important for trust. 52% want AI accuracy to be equivalent to or superior to the best dermatologist if it is used for clinical diagnosis. Participants perceived a range of benefits and risks of AI, with key risks around accuracy and divestment of AI to tech companies, and key benefits being reduction of monotonous tasks and improved patient access. Few (10%) are concerned about AI replacing dermatologists even though almost half (47%) expect key aspects of dermatology work will be performed by AI.
Conclusions: Dermatologists in Australia are at an early stage of integrating AI into practice, with most wary or unsure about the accuracy of currently available AI tools for diagnostic purposes. Developing workflows that are acceptable to clinicians may require knowledge outreach from Dermatology Colleges, such as the ACD, to help clinicians develop well informed views on AI.
期刊介绍:
Australasian Journal of Dermatology is the official journal of the Australasian College of Dermatologists and the New Zealand Dermatological Society, publishing peer-reviewed, original research articles, reviews and case reports dealing with all aspects of clinical practice and research in dermatology. Clinical presentations, medical and physical therapies and investigations, including dermatopathology and mycology, are covered. Short articles may be published under the headings ‘Signs, Syndromes and Diagnoses’, ‘Dermatopathology Presentation’, ‘Vignettes in Contact Dermatology’, ‘Surgery Corner’ or ‘Letters to the Editor’.