BACKGROUND:High naevus counts and ultraviolet photodamage are strong risk factors for melanoma. However, whole-of-body measures fail to capture variability across body sites. Three-dimensional total body photography (3D-TBP) and artificial intelligence (AI) allow us the opportunity to automate the extraction of site-specific distributions of naevi and photodamage. OBJECTIVES:To identify distinct phenotypic patterns associated with melanoma in a cohort of people at high risk of advanced melanoma, using combined 3D-TBP, AI and unsupervised clustering. METHODS:Participants with a history of melanoma (diagnosed aged > 50 years) underwent 3D-TBP. Site-specific photodamage and naevus counts were assessed using density-based spatial clustering of applications with noise to identify body site-dependent phenotypic patterns. Melanoma prevalence (none, single, multiple) relative to phenotypic pattern was evaluated using population prevalence ratios (PPRs). RESULTS:Analysis of 117 individuals found four phenotypic patterns of increasing severity: moderate V-neck photodamage with few naevi [median 38; interquartile range (IQR) 27-72] in 28 patients (24%); moderate generalized photodamage with several naevi (median 155; IQR 90-259) in 31 patients (26%); moderate V-neck photodamage with many naevi (median 204) in 20 patients (17%); and severe generalized photodamage with few naevi (median 37; IQR 21-72) naevi in 38 patients (32%). No individuals had severe photodamage and several-to-many naevi. Interpattern comparisons revealed that participants with the mildest phenotypic pattern were least likely to be affected by invasive melanomas [PPR 1.51, 95% confidence interval (CI) 1.01-2.26], whereas those with the most severe phenotypic pattern were more likely to be affected by multiple invasive melanomas (PPR 2.00, 95% CI 1.06-3.77). The prevalence of melanoma in situ was consistent across patterns. Melanoma was more likely at sites of large naevi (> 5 mm; P < 0.05) in those with moderate photodamage patterns but were independent of naevi (> 2 mm) in individuals with severe photodamage patterns. From a control cohort unaffected by melanoma (n = 114), only 18 (15.8%) matched with a high-risk phenotypic pattern. CONCLUSIONS:Three-dimensional TBP phenotyping of an older Australian cohort at high risk of invasive melanoma revealed four distinct phenotypic patterns associated with risk of the disease. Individuals with severe photodamage and relatively few naevi had a significantly higher risk of developing multiple invasive melanomas. For individuals with moderate photodamage, the risk of invasive melanoma was positively associated with the number of naevi. Thus, comprehensive phenotypes may be more predictive for the diagnosis and site of invasive melanoma, which may help with nuanced risk stratification and customized surveillance.

BACKGROUND:Given Australia's high UV exposure and sun-seeking habits, public awareness is critical for early melanoma diagnosis. This study aims to assess melanoma awareness in Australian adults. METHODS:An adapted Melanoma Cancer Awareness Measure (M-CAM) survey with both prompted and unprompted questions was used to assess knowledge of melanoma signs and symptoms, targeting adults over 18 years, excluding healthcare professionals. RESULTS:Among 390 participants, the mean number of correctly identified melanoma signs and symptoms was 3.5 out of 17 (unprompted) and 11 out of 14 (prompted). For melanoma risk factors, participants recalled an average of 3.1 items out of 21 (unprompted), and 14 out of 20 items when prompted. The most commonly identified symptoms (unprompted) were changes in the colour (72%), size (41%) and shape (34%) of an existing mole. Sun exposure was the most recognised melanoma risk factor (89%) in unprompted recall, followed by family history of melanoma (64%) and fair skin colour (47%). CONCLUSION:In this Australian sample, unprompted awareness was low, even among those with a melanoma history. Targeted efforts to identify sub-groups with low awareness, alongside education highlighting less common signs, may enhance early detection and improve melanoma outcomes.

INTRODUCTION:The COVID-19 pandemic significantly disrupted melanoma care worldwide, raising concerns about diagnosis and treatment delays. OBJECTIVES:To compare changes in incidence and prognostic factors of invasive cutaneous melanomas across two regions with contrasting COVID-19 infection control strategies. METHODS:A retrospective population-based registry study using the Swedish quality registry for cutaneous melanoma (SweMR) and the Cancer Council in Victoria, Australia was performed. Incident invasive melanomas diagnosed between 2013 and 2021 were used for trend analyses considering the long-term underlying trends and seasonality. Regression models were used for analyses of prognostic variables (Breslow thickness, ulceration, lymph node status), comparing the time periods, "before COVID" and "during COVID" (using 1 March 2020 as interruption point). RESULTS:In Victoria, melanoma diagnoses declined during COVID across all age groups, most notably among patients <50 years. Median Breslow thickness remained 0.7mm, but the distribution shifted significantly (p<0.001), with an 8% increase in geometric mean ratio (95% CI, 5%-12%). In Sweden, measures of thickness remained unchanged during COVID, but the odds of ulceration increased slightly (OR 1.09; 95% CI 1.01-1.19). Lymph node metastasis rates were stable in both regions. A significant immediate drop in overall incidence followed COVID-19 onset in both regions, primarily driven by thin melanomas. In Victoria, the lower incidence level persisted without a subsequent slope change. In contrast, the Swedish melanoma incidence gradually realigned with the pre-COVID trend after the initial level change. The immediate incidence drop was most prominent among older age groups in both regions. CONCLUSIONS:COVID-19 influenced melanoma incidence and prognostic features in both Sweden and Victoria. Reduced incidence of particularly thin melanomas and modest worsening of prognostic markers were observed. Victoria showed a more sustained impact, possibly reflecting the stricter and longer-lasting pandemic control measures.

BACKGROUND:Early detection of melanoma presents a major public health challenge. Growing evidence supports targeted surveillance of individuals at high risk identified using risk stratification. Skin photodamage is the primary environmental risk factor for melanoma; however, it is inconsistently captured and often relies on self-reporting or subjective observations, resulting in poor reproducibility. The increasing use of total-body photography (TBP) in clinical skin examinations, combined with advances in artificial intelligence technology, presents new opportunities for automated skin assessment of ultraviolet damage. OBJECTIVES:To develop a clinical photonumeric scale for photodamage assessment, use the scale to build a dataset of annotated image tiles, and train a convolutional neural network (CNN) to automate photodamage assessment from three-dimensional (3D) TBP. METHODS:Our photonumeric scale was validated for assessing photodamage and pigmentation from 3D TBP by comparing inter-rater reproducibility between two dermatology research students and two lay people. A total of 24 720 cutaneous image tiles from 56 individuals at high risk and 51 at population risk for melanoma were annotated. Annotated images were used to train a CNN with a multi-task learning (MTL) strategy that incorporated pigmentation as an auxiliary task to increase the performance for photodamage. The MTL-CNN was compared with a single-task CNN that considered photodamage in isolation. RESULTS:Lay people achieved substantial-to-almost perfect agreement with dermatology research students using the photonumeric scale (κ = 0.77-0.83). The MTL-CNN design improved performance compared with the single-task CNN, with receiver operating characteristic area under the curve (ROC-AUC) increasing from 0.91 to 0.96 (P < 0.01). Class-specific accuracy improved for mild (0.96 to 0.98; P = 0.04), moderate (0.85 to 0.92; P < 0.01) and severe (0.97 to 0.99; P < 0.01) photodamage categories, and was maintained across each body site (range 0.86-0.92). Accuracy was reproduced in an external validation set with a ROC-AUC of 0.93, including class-specific accuracies of 0.97 for mild, 0.85 for moderate and 0.97 for severe photodamage. An interface was developed to display CNN-labelled photodamage as heatmaps on 3D TBP patient avatars for clinical interpretation. CONCLUSIONS:Our CNN provides a novel tool to automatically and reproducibly report an individual's photodamage phenotype from 3D TBP. Incorporating this assessment into risk prediction models may inform targeted risk prediction facilitating surveillance recommendations.

Importance:Three-dimensional (3D) total-body photography (TBP) can support clinicians in monitoring and identifying changes to skin lesions in patients at high risk of melanoma. Objective:To assess clinical outcomes between patients at high risk of melanoma receiving usual clinical care compared with those receiving usual care plus 3D TBP and sequential digital dermoscopy imaging (SDDI) every 6 months via teledermatology. Design, Setting, and Participants:This randomized clinical trial was conducted at a research hospital in Brisbane, Australia, from April 2018 to October 2021, with adult patients (≥18 years) at high risk of developing a primary or subsequent melanoma. Data analysis was conducted from March 2022 to June 2024. Intervention:Usual care plus 3D-TBP in person and SDDI via teledermatology at baseline, 6, 12, 18, and 24 months. The control group continued usual care and completed online surveys every 6 months. Main Outcome Measures:Number and rates of excisions and/or biopsies of lesions suggestive of melanoma, and results of histopathologic testing. Results:The analysis included 314 participants (mean [SD] age, 51.6 [12.8] years; 194 females [62%]) who completed all of the study procedures (158 in the intervention and 156 in the control). In all, 1527 excisions (905 intervention and 622 in the control) were performed among 226 participants (122 intervention and 104 controls), with 67 (4%) histopathologically confirmed as melanoma and 402 (26%) as keratinocyte cancer (KC). The mean (SD) number of lesions of any type excised per person was significantly higher in the intervention (5.73 [6.77]; 95% CI, 4.66-6.79) compared to the control group (3.99 [5.72]; 95% CI, 3.08-4.89; P = .02). Fewer melanomas were detected among the intervention group compared with the control (24 [35%] vs 43 [64%], respectively), and therefore, a lower incidence rate: 2.03 (95% CI, 1.30-3.02) vs 3.62 (95% CI, 2.62-4.88), respectively. After 1 year of follow-up, the intervention had a lower, but not statistically significant, rate of melanoma per person: 0.08 (95% CI, 0.03-0.13) compared with 0.16 (95% CI, 0.08-0.25) in the control; an average of 0.86 (95% CI, 0.55-1.16) vs 0.42 (95% CI, 0.24-0.59) KCs per person; and 2.01 (95% CI, 1.50-2.51) vs 1.39 (95% CI, 0.98-1.82) excisions or biopsies per person, respectively. Conclusions and Relevance:The results of this randomized clinical trial indicate that the addition of 3D-TPB and SDDI to usual care in a teledermatology setting without AI (artificial intelligence) increased the number and rate of skin excisions and biopsies performed. Further studies are required to compare teledermatology to usual care rather than adding it, and to study whether the use of AI can improve the teledermatology outcomes. Larger studies in multiple settings with a greater number of teledermatologists are needed. This study shows that conducting clinical trials in this setting is feasible. Trial Registration:anzctr.org.au Identifier: ACTRN12618000267257.

Machine learning classification algorithms have emerged as promising tools to support the early detection of skin cancers. Existing algorithms typically assess malignancy of skin lesions based on a single skin image. This is in contrast with how clinicians integrate information from their physical examination, comparing multiple skin lesions of an individual and changes in lesions over time. Including contextual information could greatly enhance machine learning algorithms. However, contextual information in skin image datasets is predominantly scarce and inconsistent. Additionally, a dataset containing images of the same lesion across multiple time points and varying resolutions is also lacking. To address these gaps, we present a comprehensive dataset derived from skin monitoring of 480 study participants recruited from a general population sample (n = 196) and a high-risk for melanoma cohort (n = 284). This dataset includes images of 250,162 skin lesions obtained from three-dimensional total body imaging (tile images), along with corresponding dermoscopic images of 9,389 lesions. For 340 of the participants, longitudinal tile and dermoscopic images (ranging from 2 to 7) are provided.

The appearance of new pigmented lesions in adults at high risk of melanoma seems to occur randomly and is not restricted to UVR-exposed areas. Most evolving lesions are benign; however, new lesions that appear on photodamaged skin should be approached with greater caution.

Although melanoma in situ has a favourable prognosis, it is associated with an increased risk of subsequent invasive melanoma. As most studies documenting overdiagnosis of melanoma in situ are derived from population-based registries, we aimed to explore the timeline and interdependence of both melanoma in situ and invasive melanoma diagnoses in a high-risk cohort. Unexpectedly, invasive melanomas arose despite regular surveillance and multiple prior melanoma in situ, underscoring the need for improved risk-stratification tools to balance timely detection with minimizing overtreatment.

Melanoma is the deadliest skin cancer, with a stark difference in survival when detected early (5 year survival rate ∼99% stage I vs. ∼26% stage IV) ( Cancer Australia, Cancer Australia. Relative survival by stage at diagnosis (melanoma) 2019. https://ncci.canceraustralia.gov.au/outcomes/relative-survival-rate/relative-survival-stage-diagnosis-melanoma (accessed September 27, 2023). Google Scholar ). While there are well-established risk factors and widely used tools for melanoma risk identification, some risk factors are overlooked. Therefore, further research into melanoma risk factors is vital to better define high risk for melanoma and identify those at high risk. One potential risk factor is a probable correlation between the number of pigmented iris freckles and their risk of developing cutaneous melanoma ( Laino et al., 2018 Laino A.M. Berry E.G. Jagirdar K. Lee K.J. Duffy D.L. Soyer H.P. et al. Iris pigmented lesions as a marker of cutaneous melanoma risk: an Australian case–control study. Br J Dermatol. 2018; 178: 1119-1127https://doi.org/10.1111/bjd.16323 Crossref PubMed Scopus (18) Google Scholar ). However, manual iris freckle counts are subjective, resource-intensive and requires skilled personnel (Figure 1). To address this drawback, we utilize an object detection deep neural network Slim-YOLO ( Naranpanawa et al., 2021 Naranpanawa D.N.U. Gu Y. Chandra S.S. Betz-Stablein B. Sturm R.A. Soyer H.P. et al. Slim-YOLO: A Simplified Object Detection Model for the Detection of Pigmented Iris Freckles as a Potential Biomarker for Cutaneous Melanoma. DICTA 2021 - 2021. Int. Conf. Digit. Image Comput. Tech. Appl. 2021; (Institute of Electrical and Electronics Engineers Inc.)https://doi.org/10.1109/DICTA52665.2021.9647150 Crossref Scopus (2) Google Scholar ) to automatically detect these pigmented iris freckles. Here, we aimed to validate the Slim-YOLO performance against expert manual annotators, and to observe the topographic distribution of iris freckle locations.

Introduction: Having many melanocytic nevi on the skin is a risk factor for melanoma. However, the reproducibility of nevus counts in previous studies is limited due to high inter- and intraobserver variation. Despite the introduction of a protocol for counting and reporting of nevi in 1990 by the International Agency for Research on Cancer (IARC), significant variations in nevus counting methods persist across studies. Objectives: We sought to review the variations in nevus counting and reporting methods, adherence and deviations from the IARC protocol, and the reproducibility of nevus counting studies. Methods: A systematic search of Embase, PubMed and Web of Science was conducted. The review was limited to nevus (>2 mm) counting studies of general population adults conducted between 2000 and 2022, and studies using skilled examiners. Results: Out of the 8 studies which were eligible for inclusion, none followed the IARC protocol. Three studies used a predefined criterion to count nevi. Five studies provided training for their observers. Three studies assessed the inter- or intraobserver variation using the correlation coefficient (>0.75), and 3 studies attempted to verify the validity and the reproducibility of the counts. There was little to no agreement in nevus counting and reporting procedures in the reviewed studies, and most studies did not report their procedures adequately. Conclusion: This review highlights the need for an easily accessible and feasible protocol for identification, counting and reporting of nevi, which also considers nevus counting from total-body imaging and automated nevus counts since these technologies are expected to become widely available for future studies.

Background. Having many melanocytic nevi is a risk factor for melanoma. The generalizability of nevus counting is limited by lack of reliable counting methodology, despite the International Agency for Research on Cancer's (IARC) protocol.