A protocol for annotation of total body photography for machine learning to analyze skin phenotype and lesion classification.

Introduction: Artificial Intelligence (AI) has proven effective in classifying skin cancers using dermoscopy images. In experimental settings, algorithms have outperformed expert dermatologists in classifying melanoma and keratinocyte cancers. However, clinical application is limited when algorithms are presented with 'untrained' or out-of-distribution lesion categories, often misclassifying benign lesions as malignant, or misclassifying malignant lesions as benign. Another limitation often raised is the lack of clinical context (e.g., medical history) used as input for the AI decision process. The increasing use of Total Body Photography (TBP) in clinical examinations presents new opportunities for AI to perform holistic analysis of the whole patient, rather than a single lesion. Currently there is a lack of existing literature or standards for image annotation of TBP, or on preserving patient privacy during the machine learning process. Methods: This protocol describes the methods for the acquisition of patient data, including TBP, medical history, and genetic risk factors, to create a comprehensive dataset for machine learning. 500 patients of various risk profiles will be recruited from two clinical sites (Australia and Spain), to undergo temporal total body imaging, complete surveys on sun behaviors and medical history, and provide a DNA sample. This patient-level metadata is applied to image datasets using DICOM labels. Anonymization and masking methods are applied to preserve patient privacy. A two-step annotation process is followed to label skin images for lesion detection and classification using deep learning models. Skin phenotype characteristics are extracted from images, including innate and facultative skin color, nevi distribution, and UV damage. Several algorithms will be developed relating to skin lesion detection, segmentation and classification, 3D mapping, change detection, and risk profiling. Simultaneously, explainable AI (XAI) methods will be incorporated to foster clinician and patient trust. Additionally, a publicly released dataset of anonymized annotated TBP images will be released for an international challenge to advance the development of new algorithms using this type of data. Conclusion: The anticipated results from this protocol are validated AI-based tools to provide holistic risk assessment for individual lesions, and risk stratification of patients to assist clinicians in monitoring for skin cancer.

A Narrative Review: Opportunities and Challenges in Artificial Intelligence Skin Image Analyses Using Total Body Photography.

Artificial intelligence (AI) algorithms for skin lesion classification have reported accuracy at par with and even outperformance of expert dermatologists in experimental settings. However, the majority of algorithms do not represent real-world clinical approach where skin phenotype and clinical background information are considered. We review the current state of AI for skin lesion classification and present opportunities and challenges when applied to total body photography (TBP). AI in TBP analysis presents opportunities for intrapatient assessment of skin phenotype and holistic risk assessment by incorporating patient-level metadata, although challenges exist for protecting patient privacy in algorithm development and improving explainable AI methods.

Genetic testing for familial melanoma.

While the average lifetime risk of melanoma worldwide is approximately 3%, those with inherited high-penetrance mutations face an increased lifetime risk of 52-84%. In countries of low melanoma incidence, such as in Southern Europe, familial melanoma genetic testing may be warranted when there are two first degree relatives with a melanoma diagnosis. Testing criteria for high incidence countries such as USA, or with very-high incidence, such as Australia and New Zealand, would require a threshold of 3 to 4 affected family members. A mutation in the most common gene associated with familial melanoma, CDKN2A, is identified in approximately 10-40% of those meeting testing criteria. However, the use of multi-gene panels covering additional less common risk genes can significantly increase the diagnostic yield. Currently, genetic testing for familial melanoma is typically conducted by qualified genetic counsellors, however with increasing demand on testing services and high incidence rate in certain countries, a mainstream model should be considered. With appropriate training, dermatologists are well placed to identify high risk individuals and offer melanoma genetic test in dermatology clinics. Genetic testing should be given in conjunction with pre- and post-test consultation. Informed patient consent should cover possible results, the limitations and implications of testing including inconclusive results, and potential for genetic discrimination. Previous studies reporting on participant outcomes of genetic testing for familial melanoma have found significant improvements in both sun protective behavior and screening frequency in mutation carriers.

The MC1R r allele does not increase melanoma risk in MITF E318K carriers.

BACKGROUND: Population-wide screening for melanoma is not cost-effective, but genetic characterization could facilitate risk stratification and targeted screening. Common Melanocortin-1 receptor (MC1R) red hair colour (RHC) variants and Microphthalmia-associated transcription factor (MITF) E318K separately confer moderate melanoma susceptibility, but their interactive effects are relatively unexplored. OBJECTIVES: To evaluate whether MC1R genotypes differentially affect melanoma risk in MITF E318K+ vs. E318K- individuals. MATERIALS AND METHODS: Melanoma status (affected or unaffected) and genotype data (MC1R and MITF E318K) were collated from research cohorts (five Australian and two European). In addition, RHC genotypes from E318K+ individuals with and without melanoma were extracted from databases (The Cancer Genome Atlas and Medical Genome Research Bank, respectively). χ2 and logistic regression were used to evaluate RHC allele and genotype frequencies within E318K+/- cohorts depending on melanoma status. Replication analysis was conducted on 200 000 general-population exomes (UK Biobank). RESULTS: The cohort comprised 1165 MITF E318K- and 322 E318K+ individuals. In E318K- cases MC1R R and r alleles increased melanoma risk relative to wild type (wt), P < 0.001 for both. Similarly, each MC1R RHC genotype (R/R, R/r, R/wt, r/r and r/wt) increased melanoma risk relative to wt/wt (P < 0.001 for all). In E318K+ cases, R alleles increased melanoma risk relative to the wt allele [odds ratio (OR) 2.04 (95% confidence interval 1.67-2.49); P = 0.01], while the r allele risk was comparable with the wt allele [OR 0.78 (0.54-1.14) vs. 1.00, respectively]. E318K+ cases with the r/r genotype had a lower but not significant melanoma risk relative to wt/wt [OR 0.52 (0.20-1.38)]. Within the E318K+ cohort, R genotypes (R/R, R/r and R/wt) conferred a significantly higher risk compared with non-R genotypes (r/r, r/wt and wt/wt) (P < 0.001). UK Biobank data supported our findings that r did not increase melanoma risk in E318K+ individuals. CONCLUSIONS: RHC alleles/genotypes modify melanoma risk differently in MITF E318K- and E318K+ individuals. Specifically, although all RHC alleles increase risk relative to wt in E318K- individuals, only MC1R R increases melanoma risk in E318K+ individuals. Importantly, in the E318K+ cohort the MC1R r allele risk is comparable with wt. These findings could inform counselling and management for MITF E318K+ individuals.

Artificial Intelligence for the Classification of Pigmented Skin Lesions in Populations with Skin of Color: A Systematic Review.

BACKGROUND: While skin cancers are less prevalent in people with skin of color, they are more often diagnosed at later stages and have a poorer prognosis. The use of artificial intelligence (AI) models can potentially improve early detection of skin cancers; however, the lack of skin color diversity in training datasets may only widen the pre-existing racial discrepancies in dermatology. OBJECTIVE: The aim of this study was to systematically review the technique, quality, accuracy, and implications of studies using AI models trained or tested in populations with skin of color for classification of pigmented skin lesions. METHODS: PubMed was used to identify any studies describing AI models for classification of pigmented skin lesions. Only studies that used training datasets with at least 10% of images from people with skin of color were eligible. Outcomes on study population, design of AI model, accuracy, and quality of the studies were reviewed. RESULTS: Twenty-two eligible articles were identified. The majority of studies were trained on datasets obtained from Chinese (7/22), Korean (5/22), and Japanese populations (3/22). Seven studies used diverse datasets containing Fitzpatrick skin type I-III in combination with at least 10% from black Americans, Native Americans, Pacific Islanders, or Fitzpatrick IV-VI. AI models producing binary outcomes (e.g., benign vs. malignant) reported an accuracy ranging from 70% to 99.7%. Accuracy of AI models reporting multiclass outcomes (e.g., specific lesion diagnosis) was lower, ranging from 43% to 93%. Reader studies, where dermatologists' classification is compared with AI model outcomes, reported similar accuracy in one study, higher AI accuracy in three studies, and higher clinician accuracy in two studies. A quality review revealed that dataset description and variety, benchmarking, public evaluation, and healthcare application were frequently not addressed. CONCLUSIONS: While this review provides promising evidence of accurate AI models in populations with skin of color, the majority of the studies reviewed were obtained from East Asian populations and therefore provide insufficient evidence to comment on the overall accuracy of AI models for darker skin types. Large discrepancies remain in the number of AI models developed in populations with skin of color (particularly Fitzpatrick type IV-VI) compared with those of largely European ancestry. A lack of publicly available datasets from diverse populations is likely a contributing factor, as is the inadequate reporting of patient-level metadata relating to skin color in training datasets.

Protocol to evaluate a pilot program to upskill clinicians in providing genetic testing for familial melanoma.

INTRODUCTION: Genetic testing for hereditary cancers can improve long-term health outcomes through identifying high-risk individuals and facilitating targeted prevention and screening/surveillance. The rising demand for genetic testing exceeds the clinical genetic workforce capacity. Therefore, non-genetic specialists need to be empowered to offer genetic testing. However, it is unknown whether patient outcomes differ depending on whether genetic testing is offered by a genetics specialist or a trained non-genetics clinician. This paper describes a protocol for upskilling non-genetics clinicians to provide genetic testing, randomise high-risk individuals to receive testing from a trained clinician or a genetic counsellor, and then determine whether patient outcomes differed depending on provider-type. METHODS: An experiential training program to upskill dermatologically-trained clinicians to offer genetic testing for familial melanoma is being piloted on 10-15 clinicians, prior to wider implementation. Training involves a workshop, comprised of a didactic learning presentation, case studies, simulated sessions, and provision of supporting documentation. Clinicians later observe a genetic counsellor led consultation before being observed leading a consultation. Both sessions are followed by debriefing with a genetic counsellor. Thereafter, clinicians independently offer genetic testing in the clinical trial. Individuals with a strong personal and/or family history of melanoma are recruited to a parallel-group trial and allocated to receive pre- and post- genetic testing consultation from a genetic counsellor, or a dermatologically-trained clinician. A mixed method approach measures psychosocial and behavioural outcomes. Longitudinal online surveys are administered at five timepoints from baseline to one year post-test disclosure. Semi-structured interviews with both patients and clinicians are qualitatively analysed. SIGNIFICANCE: This is the first program to upskill dermatologically-trained clinicians to provide genetic testing for familial melanoma. This protocol describes the first clinical trial to compare patient-reported outcomes of genetic testing based on provider type (genetic counsellors vs trained non-genetic clinicians).

Attitudes of Australian dermatologists on the use of genetic testing: A cross-sectional survey with a focus on melanoma.

Background: Melanoma genetic testing reportedly increases preventative behaviour without causing psychological harm. Genetic testing for familial melanoma risk is now available, yet little is known about dermatologists' perceptions regarding the utility of testing and genetic testing ordering behaviours. Objectives: To survey Australasian Dermatologists on the perceived utility of genetic testing, current use in practice, as well as their confidence and preferences for the delivery of genomics education. Methods: A 37-item survey, based on previously validated instruments, was sent to accredited members of the Australasian College of Dermatologists in March 2021. Quantitative items were analysed statistically, with one open-ended question analysed qualitatively. Results: The response rate was 56% (256/461), with 60% (153/253) of respondents between 11 and 30 years post-graduation. While 44% (112/252) of respondents agreed, or strongly agreed, that genetic testing was relevant to their practice today, relevance to future practice was reported significantly higher at 84% (212/251) (t = -9.82, p < 0.001). Ninety three percent (235/254) of respondents reported rarely or never ordering genetic testing. Dermatologists who viewed genetic testing as relevant to current practice were more likely to have discussed (p < 0.001) and/or offered testing (p < 0.001). Respondents indicated high confidence in discussing family history of melanoma, but lower confidence in ordering genetic tests and interpreting results. Eighty four percent (207/247) believed that genetic testing could negatively impact life insurance, while only 26% (63/244) were aware of the moratorium on using genetic test results in underwriting in Australia. A minority (22%, 55/254) reported prior continuing education in genetics. Face-to-face courses were the preferred learning modality for upskilling. Conclusion: Australian Dermatologists widely recognise the relevance of genetic testing to future practice, yet few currently order genetic tests. Future educational interventions could focus on how to order appropriate genetic tests and interpret results, as well as potential implications on insurance.

A Systematic Review on the Impact of Genetic Testing for Familial Melanoma I: Primary and Secondary Preventative Behaviours.

BACKGROUND: Increasing availability of panel testing for known high-penetrance familial melanoma genes has made it possible to improve risk awareness in those at greatest risk. Prior to wider implementation, the role of genetic testing in preventing melanoma, through influencing primary and secondary preventative behaviours, requires clarification. METHODS: Database searches of PubMed, Embase, CINAHL, PsycINFO and the Cochrane Library were conducted for studies describing preventative behaviour outcomes in response to genetic testing for melanoma risk. Publications describing original research of any study type were screened for eligibility. RESULTS: Eighteen publications describing 11 unique studies were reviewed. Outcomes assessed are based on health behaviour recommendations for those at increased risk: adherence to sun-protective behaviour (SPB); clinical skin examinations (CSE); skin self-examinations (SSE); and family discussion of risk. Overall, modest increases in adherence to primary prevention strategies of SPB were observed following genetic testing. Importantly, there were no net decreases in SPB found amongst non-carriers. For secondary preventative behaviour outcomes, including CSE and SSE, increases in post-test intentions and long-term adherence were reported across several subgroups in approximately half of the studies. While this increase reached significance in mutation carriers in some studies, one study reported a significant decline in annual CSE adherence of non-mutation carriers. CONCLUSIONS: Evidence reviewed suggests that genetic testing has a modestly positive impact on preventative behaviour in high-risk individuals. Furthermore, improvements are observed regardless of mutation carrier status, although greater adherence is found in carriers. While additional studies of more diverse cohorts would be needed to inform clinical recommendations, the findings are encouraging and suggest that genetic testing for melanoma has a positive impact on preventative behaviours.

A Systematic Review on the Impact of Genetic Testing for Familial Melanoma II: Psychosocial Outcomes and Attitudes.

BACKGROUND: Although genetic testing for known familial melanoma genes is commercially available, clinical implementation has been restrained as utility is unclear, concerns of causing psychological distress are often cited, and consumer interest and perceptions are not well understood. A review of studies exploring participant-reported psychosocial outcomes and attitudes towards genetic testing for familial melanoma will provide insight into common emotional and cognitive responses. METHODS: Database searches of PubMed, Embase, CINAHL, PsycINFO and the Cochrane Library were conducted using a date range of January 1995 to June 2020. Studies examining any psychosocial outcomes alongside genetic testing (real or hypothetical), in participants described as having a high risk of melanoma, were eligible. A narrative synthesis of results was used to describe psychosocial outcomes and summarise participant beliefs and attitudes towards genetic testing. RESULTS: Limited evidence of adverse psychosocial outcomes was found. No impacts on perceived risk or control were reported, and minimal decisional regret was recorded. Generalised distress was comparable between both genetic mutation carriers and non-carriers, often decreasing over time from pretesting levels. Melanoma-specific distress was frequently higher in carriers than non-carriers; however, this difference was present prior to testing and often associated with personal melanoma history. Overall, participants' attitudes towards testing were largely positive, with benefits more frequently described than limitations, and support for testing minors was strong. CONCLUSIONS: This review has found evidence of few adverse psychological outcomes following genetic testing. There was no indication of increased distress after genetic test results had been disclosed. If these findings were replicated in additional, larger, diverse populations over a longer follow-up period, this would be compelling evidence to guide clinical recommendations.

Dry-coated live viral vector vaccines delivered by nanopatch microprojections retain long-term thermostability and induce transgene-specific T cell responses in mice.

The disadvantages of needle-based immunisation motivate the development of simple, low cost, needle-free alternatives. Vaccine delivery to cutaneous environments rich in specialised antigen-presenting cells using microprojection patches has practical and immunological advantages over conventional needle delivery. Additionally, stable coating of vaccine onto microprojections removes logistical obstacles presented by the strict requirement for cold-chain storage and distribution of liquid vaccine, or lyophilised vaccine plus diluent. These attributes make these technologies particularly suitable for delivery of vaccines against diseases such as malaria, which exerts its worst effects in countries with poorly-resourced healthcare systems. Live viral vectors including adenoviruses and poxviruses encoding exogenous antigens have shown significant clinical promise as vaccines, due to their ability to generate high numbers of antigen-specific T cells. Here, the simian adenovirus serotype 63 and the poxvirus modified vaccinia Ankara--two vectors under evaluation for the delivery of malaria antigens to humans--were formulated for coating onto Nanopatch microprojections and applied to murine skin. Co-formulation with the stabilising disaccharides trehalose and sucrose protected virions during the dry-coating process. Transgene-specific CD8(+) T cell responses following Nanopatch delivery of both vectors were similar to intradermal injection controls after a single immunisation (despite a much lower delivered dose), though MVA boosting of pre-primed responses with Nanopatch was found to be less effective than the ID route. Importantly, disaccharide-stabilised ChAd63 could be stored for 10 weeks at 37°C with less than 1 log10 loss of viability, and retained single-dose immunogenicity after storage. These data support the further development of microprojection patches for the deployment of live vaccines in hot climates.

Microbiopsy engineered for minimally invasive and suture-free sub-millimetre skin sampling.

We describe the development of a sub-millimetre skin punch biopsy device for painless and suture-free skin sampling for molecular diagnosis and research. Conventional skin punch biopsies range from 2-4 mm in diameter. Local anaesthesia is required and sutures are usually used to close the wound. Our microbiopsy is 0.50 mm wide and 0.20 mm thick. The microbiopsy device is fabricated from three stacked medical grade stainless steel plates tapered to a point and contains a chamber within the centre plate to collect the skin sample. We observed that the application of this device resulted in a 0.21 ± 0.04 mm wide puncture site in volunteer skin using reflectance confocal microscopy. Histological sections from microbiopsied skin revealed 0.22 ± 0.12 mm wide and 0.26 ± 0.09 mm deep puncture sites. Longitudinal observation in microbiopsied volunteers showed that the wound closed within 1 day and was not visible after 7 days. Reflectance confocal microscope images from these same sites showed the formation of a tiny crust that resolved by 3 weeks and was completely undetectable by the naked eye. The design parameters of the device were optimised for molecular analysis using sampled DNA mass as the primary end point in volunteer studies. Finally, total RNA was characterized. The optimised device extracted 5.9 ± 3.4 ng DNA and 9.0 ± 10.1 ng RNA. We foresee that minimally invasive molecular sampling will play an increasingly significant role in diagnostic dermatology and skin research.