Citations

AbstractFrozen sections are a useful pathologic tool, but variable image quality may impede the use of artificial intelligence and machine learning in their interpretation. We aimed to identify the current research on machine learning models trained or tested on frozen section images. We searched PubMed and Web of Science for articles presenting new machine learning models published in any year. Eighteen papers met all inclusion criteria. All papers presented at least one novel model trained or tested on frozen section images. Overall, convolutional neural networks tended to have the best performance. When physicians were able to view the output of the model, they tended to perform better than either the model or physicians alone at the tested task. Models trained on frozen sections performed well when tested on other slide preparations, but models trained on only formalin‐fixed tissue performed significantly worse across other modalities. This suggests not only that machine learning can be applied to frozen section image processing, but also use of frozen section images may increase model generalizability. Additionally, expert physicians working in concert with artificial intelligence may be the future of frozen section histopathology.

Abstract Evaluation of basal cell carcinoma (BCC) involves tangential biopsies of a suspicious lesion that is sent for frozen sections and evaluated by a Mohs micrographic surgeon. Advances in artificial intelligence (AI) have made possible the development of sophisticated clinical decision support systems to provide real-time feedback to clinicians that could have a role in optimizing the diagnostic workup of BCC. There were 287 annotated whole-slide images of frozen sections from tangential biopsies, of which 121 contained BCC, that were used to train and test an AI pipeline to recognize BCC. Regions of interest were annotated by a senior dermatology resident, an experienced dermatopathologist and an experienced Mohs surgeon, with concordance of annotations noted on final review. Final performance metrics included a sensitivity and specificity of 0.73 and 0.88, respectively. Our results on a relatively small dataset suggest the feasibility of developing an AI system to aid in the workup and management of BCC.

Importance Scoring systems for Stevens-Johnson syndrome and epidermal necrolysis (EN) only estimate patient prognosis and are weighted toward comorbidities and systemic features; morphologic terminology for EN lesions is inconsistent. Objectives To establish consensus among expert dermatologists on EN terminology, morphologic progression, and most-affected sites, and to build a framework for developing a skin-directed scoring system for EN. Evidence Review A Delphi consensus using the RAND/UCLA appropriateness criteria was initiated with a core group from the Society of Dermatology Hospitalists to establish agreement on the optimal design for an EN cutaneous scoring instrument, terminology, morphologic traits, and sites of involvement. Findings In round 1, the 54 participating dermatology hospitalists reached consensus on all 49 statements (30 appropriate, 3 inappropriate, 16 uncertain). In round 2, they agreed on another 15 statements (8 appropriate, 7 uncertain). There was consistent agreement on the need for a skin-specific instrument; on the most-often affected skin sites (head and neck, chest, upper back, ocular mucosa, oral mucosa); and that blanching erythema, dusky erythema, targetoid erythema, vesicles/bullae, desquamation, and erosions comprise the morphologic traits of EN and can be consistently differentiated. Conclusions and Relevance This consensus exercise confirmed the need for an EN skin-directed scoring system, nomenclature, and differentiation of specific morphologic traits, and identified the sites most affected. It also established a baseline consensus for a standardized EN instrument with consistent terminology.

Background: The Rockwood system for the classification of acute acromioclavicular (AC) joint dislocations has been associated with a lack of reliability. A novel system has been proposed (Kraus classification) that is based on dynamic posterior translation of these injuries. Purpose: To assess the interobserver and intraobserver reliability of the Rockwood and Kraus classification systems and also to examine the impact of surgeon experience on the assessments. Study Design: Cohort study (diagnosis); Level of evidence, 3. Methods: Included were 45 patients with acute AC joint injuries who underwent a radiographic examination using standard bilateral AP and modified Alexander views. For interobserver reliability, 6 shoulder surgeons (expert group) and 6 orthopaedic residents (novice group) reviewed the radiographs to classify injuries according to the Rockwood and Kraus classifications; for each categorization, the participants chose between nonoperative management and surgical treatment. The evaluation was repeated 6 weeks later to determine intraobserver reliability. Kappa (κ) coefficients and their 95% CIs were used to compare interobserver and intraobserver reliability for each classification system according to assessor experience. Statistical differences were considered significant when the upper and lower boundaries of the 95% CI did not overlap. Results: The overall interobserver agreement for diagnosis (both novice and expert groups) was moderate (κ = 0.52 [95% CI, 0.51-0.54]) for the Rockwood classification and substantial (κ = 0.62 [95% CI, 0.53-0.65]) for the Kraus classification; however, no significant differences were observed between the κ values. The overall interobserver agreement for treatment selection was substantial when using both the Rockwood (κ = 0.78 [95% CI, 0.72-0.81]) and Kraus (κ = 0.77 [95% CI, 0.66-0.87]) classifications. The overall intraobserver agreement for diagnosis was substantial using both the Rockwood (κ = 0.65 [95% CI, 0.61-0.67]) and Kraus (κ = 0.73 [95% CI, 0.69-0.75]) classifications. There were no significant differences between the novice and expert groups on any of the evaluations. Conclusion: The Kraus system was slightly more reliable than the Rockwood system for classifying AC joint injuries both between assessor groups and overall. The level of surgeon experience had no impact on the evaluations.

Background: Meniscal and chondral damage is common in the patient undergoing revision anterior cruciate ligament (ACL) reconstruction. Purpose: To determine if meniscal and/or articular cartilage pathology at the time of revision ACL surgery significantly influences a patient’s outcome at 6-year follow-up. Study Design: Cohort study; Level of evidence, 3. Methods: Patients undergoing revision ACL reconstruction were prospectively enrolled between 2006 and 2011. Data collection included baseline demographics, surgical technique, pathology, treatment, and scores from 4 validated patient-reported outcome instruments: International Knee Documentation Committee (IKDC), Knee injury and Osteoarthritis Outcome Score (KOOS), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), and Marx Activity Rating Scale. Patients were followed up at 6 years and asked to complete the identical set of outcome instruments. Regression analysis assessed the meniscal and articular cartilage pathology risk factors for clinical outcomes 6 years after revision ACL reconstruction. Results: An overall 1234 patients were enrolled (716 males, 58%; median age, 26 years). Surgeons reported the pathology at the time of revision surgery in the medial meniscus (45%), lateral meniscus (36%), medial femoral condyle (43%), lateral femoral condyle (29%), medial tibial plateau (11%), lateral tibial plateau (17%), patella (30%), and trochlea (21%). Six-year follow-up was obtained on 79% of the sample (980/1234). Meniscal pathology and articular cartilage pathology (medial femoral condyle, lateral femoral condyle, lateral tibial plateau, trochlea, and patella) were significant drivers of poorer patient-reported outcomes at 6 years (IKDC, KOOS, WOMAC, and Marx). The most consistent factors driving outcomes were having a medial meniscal excision (either before or at the time of revision surgery) and patellofemoral articular cartilage pathology. Six-year Marx activity levels were negatively affected by having either a repair/excision of the medial meniscus (odds ratio range, 1.45-1.72; P≤ .04) or grade 3-4 patellar chondrosis (odds ratio, 1.72; P = .04). Meniscal pathology occurring before the index revision surgery negatively affected scores on all KOOS subscales except for sports/recreation ( P < .05). Articular cartilage pathology significantly impaired all KOOS subscale scores ( P < .05). Lower baseline outcome scores, higher body mass index, being a smoker, and incurring subsequent surgery all significantly increased the odds of reporting poorer clinical outcomes at 6 years. Conclusion: Meniscal and chondral pathology at the time of revision ACL reconstruction has continued significant detrimental effects on patient-reported outcomes at 6 years after revision surgery.

Abstract Background Although perceived barriers to applying to dermatology have been researched among medical students, there remains a dearth of literature dedicated to understanding perceptions that medical students have of the field of dermatology and dermatologists. Methods A review of the literature in Embase, Pubmed, Scopus, Web of Science, and ScienceDirect were carried out to identify articles and abstracts between 2016 and 2021 relating to medical student perceptions of the field of dermatology. Peer-reviewed English studies measuring attitudes/level of interest in dermatology or other specialities, understanding of dermatologic topics, procedures, and/or scope of practice were included. Duplicate studies and conference abstracts were excluded. All publications were screened using the PRISMA-Sc guidelines. Findings were summarised and tabulated accordingly. Results A total of nine articles met inclusion criteria and eight are included in this review since one was not accessible online. Notable findings include non-US medical students perceiving dermatology as monotonous, stigmatized, unfamiliar, and difficult to access with a misunderstanding of the diversity and severity of the conditions dermatologists treat. No data on US medical student perceptions was found. Perceptions were found to be influential in career planning: medical students may reject specialities after exposure to negative comments on the field. Factors attracting students to dermatology include the appeal of being a dermatologist, media portrayal, and dermatologists' influence on patients' lives. Completion of dermatology-related activities improved medical student interest, comfort, and understanding of the field. Early dermatology exposure in US undergraduate premedical students led to heightened interest in the field, more confidence in ability to find dermatology mentors, and increased perception that dermatology serves the needs of underserved communities. Conclusions This review demonstrates the need to further investigate medical student perceptions of dermatology, particularly in the United States. Perceptions of medical specialities can impact medical student career choices. Understanding which misconceptions may be preventing students from exploring dermatology can inform efforts towards improving diversity, equity, and inclusion: translating to an equitable match and improving patient outcomes. Limitations include exclusion of articles published before 2016, geographic variability in studies, and limited data on evolving student perceptions over time.