Determining the clinical applicability of machine learning models through assessment of reporting across skin phototypes and rarer skin cancer types: A systematic review.

Machine learning (ML) models for skin cancer recognition may have variable performance across different skin phototypes and skin cancer types. Overall performance metrics alone are insufficient to detect poor subgroup performance. We aimed (1) to assess whether studies of ML models reported results separately for different skin phototypes and rarer skin cancers, and (2) to graphically represent the skin cancer training datasets used by current ML models. In this systematic review, we searched PubMed, Embase and CENTRAL. We included all studies in medical journals assessing an ML technique for skin cancer diagnosis that used clinical or dermoscopic images from 1 January 2012 to 22 September 2021. No language restrictions were applied. We considered rarer skin cancers to be skin cancers other than pigmented melanoma, basal cell carcinoma and squamous cell carcinoma. We identified 114 studies for inclusion. Rarer skin cancers were included by 8/114 studies (7.0%), and results for a rarer skin cancer were reported separately in 1/114 studies (0.9%). Performance was reported across all skin phototypes in 1/114 studies (0.9%), but performance was uncertain in skin phototypes I and VI from minimal representation of the skin phototypes in the test dataset (9/3756 and 1/3756, respectively). For training datasets, although public datasets were most frequently used, with the most widely used being the International Skin Imaging Collaboration (ISIC) archive (65/114 studies, 57.0%), the largest datasets were private. Our review identified that most ML models did not report performance separately for rarer skin cancers and different skin phototypes. A degree of variability in ML model performance across subgroups is expected, but the current lack of transparency is not justifiable and risks models being used inappropriately in populations in whom accuracy is low.

Connective tissue disease and sarcoidosis: a true association or a secondary granulomatous reaction?

Sarcoidosis occurs much more commonly in cohorts of connective tissue disease (1%) compared to the general population (0.01-0.04%). We present a case of concomitant connective tissue disease and cutaneous sarcoidal granulomas and discuss whether the observed granulomas represent a reactive phenomenon or true sarcoidosis. Click here for the corresponding questions to this CME article.

Subarachnoid haemorrhage due to intracranial vertebral artery dissection presenting with atypical cauda equina syndrome features: case report.

BACKGROUND: Failing to recognise the signs and symptoms of subarachnoid haemorrhage (SAH) causes diagnostic delay and may result in poorer outcomes. We report a rare case of SAH secondary to a vertebral artery dissection (VAD) that initially presented with cauda equina-like features, followed by symptoms more typical of SAH. CASE PRESENTATION: A 55-year-old man developed severe lower back pain after sudden movement. Over the next 5 days he developed paraesthesiaes in the feet, progressing to the torso gradually, and reported constipation and reduced sensation when passing urine. On day six he developed left facial palsy, and later gradual-onset headache and intermittent confusion. Magnetic resonance imaging of the brain showed diffuse subarachnoid FLAIR hyperintensity, concerning for blood, including a focus of cortical/subcortical high signal in the left superior parietal lobule, which was confirmed by computed tomography. Digital subtraction angiography demonstrated a left VAD with a fusiform aneurysm. CONCLUSION: We present a very rare case of intracranial VAD with SAH initially presenting with spinal symptoms. The majority of subsequent clinical features were consistent with a parietal focus of cortical subarachnoid blood, as observed on neuroimaging.

The impact of smoking on mortality after acute ST-segment elevation myocardial infarction treated with primary percutaneous coronary intervention: a retrospective cohort outcome study at 3 years.

The "smoker's paradox", where smokers have improved survival post-myocardial infarction, was predominantly observed in the thrombolytic era. However, evidence for the smoker's paradox in the current era of PCI therapy is both limited and inconsistent. We aimed to examine the effect of smoking status on survival in unselected ST-elevation myocardial infarction (STEMI) patients managed by primary percutaneous coronary intervention (PCI). Data were collected for all patients with acute STEMI undergoing primary PCI at The South Yorkshire Cardiothoracic Centre, UK over a 5-year period between 2009 and 2014. Differences in survival by smoking status were assessed before and after adjustment for differences in baseline variables using a Kaplan-Meier curve and a Cox regression analysis, respectively. A total of 3133 STEMI patients were included in the study. After adjustment for differences in baseline variables, smoking was associated with a significantly increased mortality (hazard ratio 1.35 (95% CI 1.04-1.74)) compared to never smokers after 3 years. The risk for ex-smokers (hazard ratio 0.99 (0.76-1.28)) was similar to never smokers. There were no significant differences in survival by smoking status at 30 days and 1 year. In this large registry of STEMI patients managed by primary PCI, smokers had a significantly higher 3-year mortality than non-smokers. This study is the first to not only dispel the existence of the smoker's paradox, but to highlight a high-risk subgroup who may warrant tailored secondary prevention treatment, including smoking cessation.

How is the term haemangioma used in the literature? An evaluation against the revised ISSVA classification.

BACKGROUND/OBJECTIVES: The International Society for the Study of Vascular Anomalies (ISSVA) classification separates vascular anomalies into vascular malformations and vascular tumors. However, misdiagnoses and misperceptions still persist around the use of the term "hemangioma." We assessed whether the term "haemangioma" (British spelling) was used as part of ISSVA terminology in the literature. METHODS: We searched PubMed for all English-language publications containing the British spelling "haemangioma" in the title or abstract from January 1, 2015 to December 31, 2016. Each paper was judged by two independent reviewers, with conflicts resolved by senior review. RESULTS: By the standard of the 2014 ISSVA classification, 126/195 (64.6%) publications used incorrect terminology for vascular anomalies. This was reduced to 118/195 (60.5%) when using the 2018 ISSVA classification. The most commonly misused terms were cavernous haemangioma (27.1%), haemangioma without further specification (26.3%), and hepatic/liver haemangioma (12.7%). Age was a significant predictor of accuracy of terminology (P = 0.01), with a higher accuracy in children. Correct usage also varied by the site of the vascular anomaly, being highest for lesions of the skin (76.5%) followed by muscle (58.3%), soft tissue (23.5%), bone (21.4%), viscera (7.7%), and eye (0.0%) (P = 0.02). CONCLUSIONS: The term "haemangioma" is frequently used incorrectly by the standards of the 2014 and 2018 ISSVA classifications. Correct terminology is important as the natural history and treatment options vary depending on the type of vascular anomaly.

Meningococcal arthritis and myopericarditis: a case report.

BACKGROUND: We report the first adult case of Neisseria meningitidis W-135 presenting with meningococcal arthritis and myopericarditis concomitantly, without other classical features of meningococcal disease. CASE PRESENTATION: A 67-year-old Caucasian man presented with acute-onset polyarthralgia, myalgia, and fever. On examination he had polyarticular synovitis. An electrocardiogram (ECG) demonstrated ST-elevation in leads I, II, III, aVF, and V2-V6 without reciprocal depression, and a high-sensitivity troponin level was significantly elevated. Cardiac magnetic resonance (CMR) imaging on day five of admission demonstrated patchy pericardial enhancement. Neisseria meningitidis W-135 was isolated from both synovial fluid and blood cultures. The clinical outcome was favourable with intravenous ceftriaxone and myopericarditis treatment (colchicine and ibuprofen). CONCLUSIONS: We conclude that this is a rare case of disseminated Neisseria meningitidis W-135 presenting with acute polyarticular septic arthritis and myopericarditis, without other classical features of systemic meningococcal disease. The earlier described entity of primary meningococcal arthritis (PMA) can present in patients with meningococcal bacteraemia, and may not be distinct from disseminated meningococcal disease, but rather an atypical presentation of this.

A retrospective cohort study of the association between smoking and mortality after acute ST-segment elevation myocardial infarction.

BACKGROUND: Several studies have shown a 'smoker's paradox', where following an acute myocardial infarction, smokers have a paradoxically lower mortality than non-smokers. To date, no large study has investigated this paradox in unselected patients with acute ST-segment elevation myocardial infarction (STEMI) managed by primary percutaneous coronary intervention (PCI) alone. OBJECTIVES: We aimed to examine the association of smoking status and 1-year mortality in patients who had STEMI managed by primary PCI. METHODS: This retrospective study included all patients admitted with acute STEMI undergoing primary PCI in a single UK centre from January 2009 to April 2012. The survival status for all patients post-STEMI was obtained. Differences in survival by smoking status were assessed using a Kaplan-Meier curve, and after adjustment for age, gender and additional cardiovascular risk factors using a Cox regression analysis. RESULTS: The 1-year mortality for patients with STEMI was 149/1796 (8.3%). There were 846/1796 (47.1%) current smokers, 476/1796 (26.5%) ex-smokers and 417/1796 (23.2%) never smokers. Current smokers were approximately 10 years younger than ex-smokers and never smokers (p=0.001). A multivariate Cox proportional hazards model found no evidence of an association between mortality and smoking status after adjustment; p=0.23. Compared with never smokers, the HR (95% CI) for 1-year mortality for current smokers was 1.47 (0.90 to 2.39) and 1.08 (0.66 to 1.77) for ex-smokers. CONCLUSIONS: In this retrospective cohort study, we found no evidence of an association between mortality and smoking status in patients with acute STEMI treated with PCI, and thus no evidence of a 'smoker's paradox'.

A retrospective cross-sectional study on the association between tobacco smoking and incidence of ST-segment elevation myocardial infarction and cardiovascular risk factors.

BACKGROUND: Cigarette smoking is a well-established risk factor for the development of coronary heart disease. However, the relationship between smoking and acute ST-segment elevation myocardial infarction (STEMI) is less well described. OBJECTIVE: To determine the relative risk of acute STEMI in smokers and ex-smokers, compared with individuals who had never smoked. METHODS: This observational study studied all patients with STEMI undergoing percutaneous coronary intervention (PCI) in South Yorkshire, UK from 1 January 2009 to 6 April 2012. Additional contemporary demographical data for the South Yorkshire population, supplied by the Office for National Statistics, allowed derivation of the incidence rate of STEMI in South Yorkshire-both overall and stratified by smoking status. Incidence rate ratios and population attributable risk (PAR) were calculated to quantify STEMI risk. RESULTS: There were 1715 STEMIs in 1680 patients during the study period. Smoking status was obtained in 96.2% patients. The prevalence of smoking was 47.3% in patients with STEMI and 22.0% in the general population. In patients with STEMI, smokers were ∼10 years younger, mean (SD) 57.2 (11.1) years, than never-smokers, 66.4 (12.1) years, and ex-smokers, 67.9 (11.9) years. The age-standardised incident rate ratio of STEMI was 5.2 (4.5-6.1) for current smokers and 1.1 (1.0-1.3) for ex-smokers, with the reference group being never-smokers for both. Almost 50% of STEMIs were attributable to smoking (PAR=48.3%). CONCLUSION: Cigarette smoking is associated with a fivefold increased risk of STEMI. Smoking cessation reduced this risk to a level similar to never-smokers.

Reliable Detection of Eczema Areas for Fully Automated Assessment of Eczema Severity from Digital Camera Images.

Assessing the severity of eczema in clinical research requires face-to-face skin examination by trained staff. Such approaches are resource-intensive for participants and staff, challenging during pandemics, and prone to inter- and intra-observer variation. Computer vision algorithms have been proposed to automate the assessment of eczema severity using digital camera images. However, they often require human intervention to detect eczema lesions and cannot automatically assess eczema severity from real-world images in an end-to-end pipeline. We developed a model to detect eczema lesions from images using data augmentation and pixel-level segmentation of eczema lesions on 1,345 images provided by dermatologists. We evaluated the quality of the obtained segmentation compared with that of the clinicians, the robustness to varying imaging conditions encountered in real-life images, such as lighting, focus, and blur, and the performance of downstream severity prediction when using the detected eczema lesions. The quality and robustness of eczema lesion detection increased by approximately 25% and 40%, respectively, compared with that of our previous eczema detection model. The performance of the downstream severity prediction remained unchanged. Use of skin segmentation as an alternative to eczema segmentation that requires specialist labeling showed the performance on par with when eczema segmentation is used.

Detecting Eczema Areas in Digital Images: An Impossible Task?

Assessing the severity of atopic dermatitis (AD, or eczema) traditionally relies on a face-to-face assessment by healthcare professionals and may suffer from inter- and intra-rater variability. With the expanding role of telemedicine, several machine learning algorithms have been proposed to automatically assess AD severity from digital images. Those algorithms usually detect and then delineate (segment) AD lesions before assessing lesional severity and are trained using the data of AD areas detected by healthcare professionals. To evaluate the reliability of such data, we estimated the inter-rater reliability of AD segmentation in digital images. Four dermatologists independently segmented AD lesions in 80 digital images collected in a published clinical trial. We estimated the inter-rater reliability of the AD segmentation using the intraclass correlation coefficient at the pixel and the area levels for different resolutions of the images. The average intraclass correlation coefficient was 0.45 ( standard error = 0.04 ) corresponding to a poor agreement between raters, whereas the degree of agreement for AD segmentation varied from image to image. The AD segmentation in digital images is highly rater dependent even among dermatologists. Such limitations need to be taken into consideration when AD segmentation data are used to train machine learning algorithms that assess eczema severity.

Full-thickness Skin Graft Fixation Techniques: A Review of the Literature.

Multiple techniques for skin graft fixation have been proposed, but the evidence underlying these techniques is unclear. This study aimed to review the literature for full-thickness graft fixation techniques. PubMed was electronically searched to identify relevant studies. The search strategy identified 91 relevant articles. These consisted of 2 randomised controlled trials (RCTs), 10 observational cohort studies (8 retrospective, 2 prospective), and 79 descriptive studies (case series, case reports, or expert opinion articles). Both identified RCTs compared the tie-over dressing against a modified tie-over dressing. The tie-over dressing was also included in all identified observational studies, and comparisons were made against quilting/mattress suturing (4 studies, 181 grafts in total), simple pressure dressings (3 studies, 528 grafts), non-tie-over dressings non-specifically (1 study, 71 grafts), hydrocolloid dressings (1 study, 62 grafts), and double-tie over dressings (1 study, 43 grafts). No significant differences were found between fixation methods for graft take, haematoma rate, and infection rate. No studies have found a significant difference between tie-over dressings and alternative graft fixation technique, with the most evidence for simple pressure dressings and quilting/mattress suturing. However, the evidence base consists mostly of small, retrospective observational studies. This article describes the current evidence base and this should be considered when planning future reports in the field.

The Proteolytic Network in Palmoplantar Keratoderma: SERPINA12 Joins the Family.

Mohamad et al. (2020) describe loss-of-function mutations in SERPINA12 as a cause of diffuse, transgradient palmoplantar keratoderma (PPK). This disorder shares similar clinical features with other PPKs caused by protease overactivity, including erythema, peeling, and exacerbation on water exposure. Understanding this disorder may shed further light on the role of proteases and their inhibitors in epidermal physiology.

Using sepsis scores in emergency department and ward patients.

Sepsis-3, published in 2016, defined sepsis as 'life-threatening organ dysfunction caused by a dysregulated host response to infection'. Instead of systemic inflammatory response syndrome (SIRS), calculating the Sequential Organ Failure Assessment (SOFA) score was recommended. The complexity of SOFA also led to the introduction of quick SOFA (qSOFA) as a bedside tool. The simultaneous removal of SIRS and introduction of qSOFA belies their significant differences, with SIRS having a high sensitivity but very low specificity, and qSOFA being very specific for a poor outcome, but having a lower sensitivity than SIRS. In the UK, the variables within qSOFA are collected on a regular and repeated basis, along with additional variables, as part of the National Early Warning Score (NEWS). A knowledge of SIRS, qSOFA and NEWS is of value in assessing patients with suspected sepsis, as discussed in this article.

Sentinel bruising as a presentation of metastatic melanoma.

A 46-year-old man presented with a 4-week history of bruising with subcutaneous nodules and weight loss. He also had a 2-week history of progressive back and hip pain. He had been diagnosed with stage Ib cutaneous melanoma 30 months previously, which had been fully excised. A sentinel lymph node biopsy was negative. On examination, there were five skin lesions at different stages. Each had spontaneously appeared as a bruise with a central subcutaneous nodule, and the bruising then faded to leave a persistent subcutaneous nodule. Excision of one of the nodules demonstrated a 4.5 mm diameter partly necrotic melanoma deposit in the dermis. CT scan of the head, chest, abdomen and pelvis showed widespread metastases. This rare presentation of cutaneous malignant melanoma metastases has been termed 'sentinel bruising'. There are fewer than 10 cases reported in the literature.