Improving Artificial Intelligence-Based Diagnosis on Pediatric Skin Lesions.

Artificial intelligence algorithms to classify melanoma are dependent on their training data, which limits generalizability. The objective of this study was to compare the performance of an artificial intelligence model trained on a standard adult-predominant dermoscopic dataset before and after the addition of additional pediatric training images. The performances were compared using held-out adult and pediatric test sets of images. We trained two models: one (model A) on an adult-predominant dataset (37,662 images from the International Skin Imaging Collaboration) and the other (model A+P) on an additional 1,536 pediatric images. We compared performance between the two models on adult and pediatric held-out test images separately using the area under the receiver operating characteristic curve. We then used Gradient-weighted Class Activation Maps and background skin masking to understand the contributions of the lesion versus background skin to algorithm decision making. Adding images from a pediatric population with different epidemiological and visual patterns to current reference standard datasets improved algorithm performance on pediatric images without diminishing performance on adult images. This suggests a way that dermatologic artificial intelligence models can be made more generalizable. The presence of background skin was important to the pediatric-specific improvement seen between models. Our study highlights the importance of carefully curated and labeled data from diverse inputs to improve the generalizability of AI models for dermatology, in this case applied to dermoscopic images of adult and pediatric lesions to improve melanoma detection.

Bias Breakers: Continuous Practice for Admissions and Selection Committees.

Introduction: Admissions and selection committees face challenges in identifying and mitigating biases in policies, processes, and discussions. Past bias training has focused on defining bias and presenting the negative impact of bias for committees. Methods: This interactive training used committee comments, reflection, and audience response to enhance the contextual recognition of bias in admissions and selection processes and practices. For each bias type, we presented specific mitigation strategies and examples. The workshop was offered at four medical schools between December 2020 and April 2021. Participants were committee members (n = 126), largely medical school faculty, involved in MD, MD/PhD, and residency program selection at participating schools. A paired pre- and postworkshop assessment was conducted for each session to determine effectiveness of the workshop. Results: Mean scores for each of the postassessment items ranged from 4.0 to 4.2 and were statistically significantly different from the preassessment scores per respective item. The results of a paired two-way t test found that these pre- to postworkshop assessment score increases were statistically significant across all assessment questions (ps < .001). Participants reported in their comments that the workshop was effective in establishing a safe and judgment-free learning environment to explore and identify biases and build skills and confidence for mitigating them. Discussion: Interactive and applied bias training can be an effective strategy to advance committee culture and practice in recognizing and mitigating bias. This workshop provides committees with ongoing tools for equity practice in selection and decision-making.

Digital skin imaging applications, part II: a comprehensive survey of post-acquisition image utilization features and technology standards.

BACKGROUND: Despite the increasing ubiquity and accessibility of teledermatology applications, few studies have comprehensively surveyed their features and technical standards. Importantly, features implemented after the point of capture are often intended to augment image utilization, while technical standards affect interoperability with existing healthcare systems. We aim to comprehensively survey image utilization features and technical characteristics found within publicly discoverable digital skin imaging applications. MATERIALS AND METHODS: Applications were identified and categorized as described in Part I. Included applications were then further assessed by three independent reviewers for post-imaging content, tools, and functionality. Publicly available information was used to determine the presence or absence of relevant technology standards and/or data characteristics. RESULTS: A total of 20 post-image acquisition features were identified across three general categories: (1) metadata attachment, (2) functional tools (i.e., those that utilized images or in-app content to perform a user-directed function), and (3) image processing. Over 80% of all applications implemented metadata features, with nearly half having metadata features only. Individual feature occurred and feature richness varied significantly by primary audience (p < 0.0001) and function (p < 0.0001). On average, each application included under three features. Less than half of all applications requested consent for user-uploaded photos and fewer than 10% provided clear data use and privacy policies. CONCLUSION: Post-imaging functionality in skin imaging applications varies significantly by primary audience and intended function, though nearly all applications implemented metadata labeling. Technical standards are often not implemented or reported consistently. Gaps in the provision of clear consent, data privacy, and data use policies should be urgently addressed.

Digital skin imaging applications, part I: Assessment of image acquisition technique features.

BACKGROUND: The rapid adoption of digital skin imaging applications has increased the utilization of smartphone-acquired images in dermatology. While this has enormous potential for scaling the assessment of concerning skin lesions, the insufficient quality of many consumer/patient-taken images can undermine clinical accuracy and potentially harm patients due to lack of diagnostic interpretability. We aim to characterize the current state of digital skin imaging applications and comprehensively assess how image acquisition features address image quality. MATERIALS AND METHODS: Publicly discoverable mobile, web, and desktop-based skin imaging applications, identified through keyword searches in mobile app stores, Google Search queries, previous teledermatology studies, and expert recommendations were independently assessed by three reviewers. Applications were categorized by primary audience (consumer-facing, nonhospital-based practice, or enterprise/health system), function (education, store-and-forward teledermatology, live-interactive teledermatology, electronic medical record adjunct/clinical imaging storage, or clinical triage), in-app connection to a healthcare provider (yes or no), and user type (patient, provider, or both). RESULTS: Just over half (57%) of 191 included skin imaging applications had at least one of 14 image acquisition technique features. Those that were consumer-facing, intended for educational use, and designed for both patient and physician users had significantly greater feature richness (p < 0.05). The most common feature was the inclusion of text-based imaging tips, followed by the requirement to submit multiple images and body area matching. CONCLUSION: Very few skin imaging applications included more than one image acquisition technique feature. Feature richness varied significantly by audience, function, and user categories. Users of digital dermatology tools should consider which applications have standardized features that improve image quality.

The Model Minority Myth, Data Aggregation, and the Role of Medical Schools in Combating Anti-Asian Sentiment.

The COVID-19 pandemic has resulted in an alarming increase in hate incidents directed toward Asian Americans and Pacific Islanders (AAPIs), including verbal harassment and physical assault, spurring the nationwide #StopAsianHate movement. This rise in anti-Asian sentiment is occurring at a critical time of racial reckoning across the United States, galvanized by the Black Lives Matter movement, and of medical student calls for the implementation of antiracist medical curricula. AAPIs are stereotyped by the model minority myth, which posits that AAPIs are educated, hardworking, and therefore able to achieve high levels of success. This myth acts as a racial wedge between minorities and perpetuates harm that is pervasive throughout the field of medicine. Critically, the frequent aggregation of all AAPI subgroups as one monolithic community obfuscates socioeconomic and cultural differences across the AAPI diaspora while reinforcing the model minority myth. Here, the authors illustrate how the model minority myth and data aggregation have negatively affected the recruitment and advancement of diverse AAPI medical students, physicians, and faculty. Additionally, the authors discuss how data aggregation obscures health disparities across the AAPI diaspora and how the model minority myth influences the illness experiences of AAPI patients. Importantly, the authors outline specific actionable policies and reforms that medical schools can implement to combat anti-Asian sentiment and support the AAPI community.

An Analysis of Skin of Color Publications in the Journal of Drugs in Dermatology from 2018 to 2020.

As the United States population becomes increasingly diverse, it is exceedingly important for dermatologists to be knowledgeable about treating patients with skin of color (SOC). The published literature is an especially valuable resource for treating SOC.

Transcriptomic Analysis of the Major Orphan Ichthyosis Subtypes Reveals Shared Immune and Barrier Signatures.

Preliminary work suggested upregulation of inflammatory pathways in patients with common forms of ichthyosis. However, a comprehensive characterization of skin from various ichthyosis subtypes is unavailable, precluding the development of targeted treatments. Thus, we sought to characterize the immune and barrier profiles of common and subtype-specific skin transcriptomes in a large group of patients with ichthyosis. We performed a global RNA-sequencing analysis in 54 patients with ichthyosis (7 with Netherton syndrome, 13 with epidermolytic ichthyosis, 16 with lamellar ichthyosis, and 18 with congenital ichthyosiform erythroderma) and 40 healthy controls. Differentially expressed genes were defined on the basis of fold changes > 2 and false discovery rate < 0.05 criteria. We found robust and significant T helper (Th) 22/Th17 skewing in all subtypes (e.g., IL-17A/C/F, S100A7/8/9/12; P < 0.001) with modest changes in Th2 pathway, primarily in Netherton syndrome, and Th1 skewing in congenital ichthyosiform erythroderma. Across all subtypes (less evident in epidermolytic ichthyosis), lipid metabolism and barrier junction markers were downregulated (e.g., FA2H, CDH10/11/12/2; P < 0.05), whereas epidermal cornification and proliferation measures were upregulated (e.g., SPRR1A/1B/2C/2G, EREG; P < 0.05). Our findings suggest that the common ichthyosis variants share aberrations in Th17/Th22 and barrier function, with minimal Th2 modulation. This may help to elucidate the pathogeneses of these subtypes and inform the development of subtype-specific treatments.

Psychosocial issues and body dysmorphic disorder in aesthetics: Review and debate.

Body dysmorphic disorder (BDD) is a serious and debilitating psychiatric condition that disproportionately presents in dermatologic and cosmetic surgery patients. BDD is currently underrecognized in cosmetic medical settings and is thought to be underdiagnosed by behavioral health professionals. The significant comorbidities associated with this disorder, as well as potential harm done to both patient and physician, raise ethical and medicolegal concerns regarding its treatment. Although cosmetic interventions have historically been discouraged in BDD, recent studies have provided controversial evidence of benefit in certain cohorts. The rise of "snapchat dysmorphia" and the proposed explanatory phenomenon of perception drift have generated further debate around the de novo development or unmasking of BDD. We critically review and summarize existing debates around the treatment of BDD in cosmetic medicine. We provide guidance for screening, clinical interviewing, and the provision of psychoeducation in cases of suspected BDD.

Deep learning prediction of axillary lymph node status using ultrasound images.

OBJECTIVE: To investigate the ability of our convolutional neural network (CNN) to predict axillary lymph node metastasis using primary breast cancer ultrasound (US) images. METHODS: In this IRB-approved study, 338 US images (two orthogonal images) from 169 patients from 1/2014-12/2016 were used. Suspicious lymph nodes were seen on US and patients subsequently underwent core-biopsy. 64 patients had metastatic lymph nodes. A custom CNN was utilized on 248 US images from 124 patients in the training dataset and tested on 90 US images from 45 patients. The CNN was implemented entirely of 3 × 3 convolutional kernels and linear layers. The 9 convolutional kernels consisted of 6 residual layers, totaling 12 convolutional layers. Feature maps were down-sampled using strided convolutions. Dropout with a 0.5 keep probability and L2 normalization was utilized. Training was implemented by using the Adam optimizer and a final SoftMax score threshold of 0.5 from the average of raw logits from each pixel was used for two class classification (metastasis or not). RESULTS: Our CNN achieved an AUC of 0.72 (SD ± 0.08) in predicting axillary lymph node metastasis from US images in the testing dataset. The model had an accuracy of 72.6% (SD ± 8.4) with a sensitivity and specificity of 65.5% (SD ± 28.6) and 78.9% (SD ± 15.1) respectively. Our algorithm is available to be shared for research use. (https://github.com/stmutasa/MetUS). CONCLUSION: It's feasible to predict axillary lymph node metastasis from US images using a deep learning technique. This can potentially aid nodal staging in patients with breast cancer.

Advances in the Etiology, Detection, and Clinical Management of Seborrheic Keratoses.

Seborrheic keratoses (SKs) are ubiquitous, generally benign skin tumors that exhibit high clinical variability. While age is a known risk factor, the precise roles of UV exposure and immune abnormalities are currently unclear. The underlying mechanisms of this benign disorder are paradoxically driven by oncogenic mutations and may have profound implications for our understanding of the malignant state. Advances in molecular pathogenesis suggest that inhibition of Akt and APP, as well as existing treatments for skin cancer, may have therapeutic potential in SK. Dermoscopic criteria have also become increasingly important to the accurate detection of SK, and other noninvasive diagnostic methods, such as reflectance confocal microscopy and optical coherence tomography, are rapidly developing. Given their ability to mimic malignant tumors, SK cases are often used to train artificial intelligence-based algorithms in the computerized detection of skin disease. These technologies are becoming increasingly accurate and have the potential to significantly augment clinical practice. Current treatment options for SK cause discomfort and can lead to adverse post-treatment effects, especially in skin of color. In light of the discontinuation of ESKATA in late 2019, promising alternatives, such as nitric-zinc and trichloroacetic acid topicals, should be further developed. There is also a need for larger, head-to-head trials of emerging laser therapies to ensure that future treatment standards address diverse patient needs.