A comparative study of the image quality of dermoscopic images acquired in general practice versus at a dermatology department

• Main Authors: Louise Niklasson (Author), Sumangali Chandra Prasad (Author), Annette Schuster (Author), Rasa Laurinaviciene (Author), Tine Vestergaard (Author)
• Format: Book
• Published: Wiley, 2023-12-01T00:00:00Z.
• Subjects: article
• Online Access: Connect to this object online.
• Notes: 2768-6566; 10.1002/jvc2.230
• Keywords: diagnostic accuracy; diagnostic confidence; image quality; management; skin cancer; teledermoscopy (TDS); Dermatology; RL1-803; Diseases of the genitourinary system. Urology; RC870-923
• Source: JEADV Clinical Practice, Vol 2, Iss 4, Pp 802-809 (2023)

Abstract Background Skin cancer constitutes a significant and growing disease burden. To deal with this, teledermoscopy (TDS) can be useful. However, image quality is a concern for implementing TDS. High image quality can be obtained at a dermatology department, but in many occasions, it is more sensible to acquire TDS images in general practice. Objectives To compare image quality, diagnostic confidence, suggested management, sensitivity, specificity and diagnostic accuracy between two sets of dermoscopic images of the same skin tumour, where one set was acquired in general practice, and the other was obtained at a dermatology department. Methods Two sets of dermoscopic images of 192 skin lesions were evaluated by the same dermatologist at two different occasions approximately 4 years apart (2018 vs. 2022). On both occasions, the images were evaluated in regard to image quality, diagnostic confidence, whether the lesion was a benign or malignant, primary diagnosis and suggested management. Results There were statistically significant differences of diagnostic confidence and suggested management (p = 0.00 for respective parameter), but not for image quality (p = 0.37), sensitivity (p = 0.67), specificity (p = 0.66) and diagnostic accuracy (p = 0.81). The intraobserver agreements were slight for image quality (κ = 0.012), fair for diagnostic confidence (κ = 0.313) and moderate for suggested management (κ = 0.467). Conclusions The image quality of both sets of dermoscopic images was most often good. Although we did not find a statistically significant difference in image quality between the two sets of images, our minimal level of intraobserver agreement suggests otherwise. However, this should not hinder TDS image acquisition in general practice as our results indicate that with the availability of referral notes, TDS images taken in general practice can achieve a similar sensitivity, specificity and diagnostic accuracy as dermoscopic images obtained at a dermatology department.