"Facekit"-Toward an Automated Facial Analysis App Using a Machine Learning-Derived Facial Recognition Algorithm.

Introduction: Multiple tools have been developed for facial feature measurements and analysis using facial recognition machine learning techniques. However, several challenges remain before these will be useful in the clinical context for reconstructive and aesthetic plastic surgery. Smartphone-based applications utilizing open-access machine learning tools can be rapidly developed, deployed, and tested for use in clinical settings. This research compares a smartphone-based facial recognition algorithm to direct and digital measurement performance for use in facial analysis. Methods: Facekit is a camera application developed for Android that utilizes ML Kit, an open-access computer vision Application Programing Interface developed by Google. Using the facial landmark module, we measured 4 facial proportions in 15 healthy subjects and compared them to direct surface and digital measurements using intraclass correlation (ICC) and Pearson correlation. Results: Measurement of the naso-facial proportion achieved the highest ICC of 0.321, where ICC > 0.75 is considered an excellent agreement between methods. Repeated measures analysis of variance of proportion measurements between ML Kit, direct and digital methods, were significantly different (F[2,14] = 6-26, P<<.05). Facekit measurements of orbital, orbitonasal, naso-oral, and naso-facial ratios had overall low correlation and agreement to both direct and digital measurements (R<<0.5, ICC<<0.75). Conclusion: Facekit is a smartphone camera application for rapid facial feature analysis. Agreement between Facekit's machine learning measurements and direct and digital measurements was low. We conclude that the chosen pretrained facial recognition software is not accurate enough for conducting a clinically useful facial analysis. Custom models trained on accurate and clinically relevant landmarks may provide better performance.

Introduction : Il existe de multiples outils pour procéder aux mesures et à l'analyse des caractéristiques faciales à l'aide des techniques d'apprentissage machine de la reconnaissance faciale. Cependant, il reste plusieurs défis à relever avant qu'ils soient utiles en contexte clinique de chirurgie reconstructive et de chirurgie plastique. Des applications pour téléphone intelligent faisant appel à des outils d'apprentissage machine en libre accès peuvent être rapidement mises au point, déployées et mises à l'essai dans un cadre clinique. Dans la présente étude, les chercheurs comparent un algorithme de reconnaissance faciale sur téléphone intelligent pour effectuer les mesures directes et numériques nécessaires lors de l'analyse faciale. Méthodologie : Facekit est une application pour appareil photo de téléphone Android qui fait appel à ML Kit, une application de vision par ordinateur en libre accès créée par Google. Au moyen du module de repères faciaux, les chercheurs ont mesuré quatre proportions faciales chez 15 sujets en santé et les ont comparées aux mesures de surface directe et aux mesures numériques à l'aide de la corrélation intraclasse et de la corrélation de Pearson. Résultats : La mesure de la proportion nasofaciale a obtenu le coefficient de corrélation intraclasse (CCI) le plus élevé, à 0,321, où un CCI supérieur à 0,75 est considéré comme une excellente corrélation entre les méthodes. Des analyses de variance répétées des mesures de proportion entre le ML Kit, la méthode directe et la méthode numérique différaient considérablement (F[2,14] = 6 à 26, p<<0,05). Les mesures Facekit des ratios entre les mesures orbitale, orbitonasale, naso-orale et naso-faciale avaient une faible corrélation globale et étaient corrélées avec les mesures directes et numériques (R<<0,5, CCI<<0,75). Conclusion : Facekit est une application pour appareil photo de téléphone intelligent visant à analyser rapidement les caractéristiques faciales. La concordance entre les mesures d'apprentissage machine de Facekit et les mesures directes et numériques était faible. Les chercheurs concluent que le logiciel de reconnaissance faciale préentraîné n'est pas assez précis pour procéder à une analyse faciale utile sur le plan clinique. Des modèles personnalisés formés à des repères précis et pertinents sur le plan clinique donneront peut-être un meilleur rendement.

Upper Extremity Infection Related to Intravenous Drug Use: Considering the True Cost of the COVID-19 Pandemic and Lockdown.

BACKGROUND: The COVID-19 pandemic caused significant morbidity and mortality in people who inject drugs (PWID). Upper extremity soft tissue infections are frequently associated with intravenous drug use (IVDU) due to poor compliance with aseptic technique. In Canada, multiple safe injection sites providing clean injection supplies closed, leaving many PWID with no alternatives to inject safely. It was hypothesized that these closures will correspond with increased morbidity and mortality among PWID. The main objective of this study was to determine the effect of the COVID-19 pandemic on the incidence of upper extremity infections in PWID. METHODS: This was a retrospective chart review study. The primary outcome of interest was the frequency of upper extremity infections in PWID. Data were filtered to include only those patients presenting to the emergency department between March to June of 2019 and 2020. Chi-squared analysis was used to compare the number of IVDU patients among patients with upper extremity skin infections between these time periods. RESULTS: The number of IVDU patients treated for upper extremity infections in Hamilton significantly increased during the pandemic, relative risk = 2.0 (95% confidence interval [CI]: 1.3-2.9, P = .0012,) while total upper extremity infections numbers have decreased overall. During the pandemic, PWID made up a larger proportion of upper extremity infections (χ2 = 10.444, P = .00123). Demographic data such as age and sex of IVDU patients presenting with upper extremity infection was not significantly affected by the pandemic. CONCLUSIONS: The effect of the pandemic on accessing harm reduction services has led to evident increases in morbidity as described by this study. Further research on the impact of closures in PWID is needed to quantify these harms and work toward mitigation strategies.

Cancer-selective metabolic vulnerabilities in MYC-amplified medulloblastoma.

MYC-driven medulloblastoma (MB) is an aggressive pediatric brain tumor characterized by therapy resistance and disease recurrence. Here, we integrated data from unbiased genetic screening and metabolomic profiling to identify multiple cancer-selective metabolic vulnerabilities in MYC-driven MB tumor cells, which are amenable to therapeutic targeting. Among these targets, dihydroorotate dehydrogenase (DHODH), an enzyme that catalyzes de novo pyrimidine biosynthesis, emerged as a favorable candidate for therapeutic targeting. Mechanistically, DHODH inhibition acts on target, leading to uridine metabolite scarcity and hyperlipidemia, accompanied by reduced protein O-GlcNAcylation and c-Myc degradation. Pyrimidine starvation evokes a metabolic stress response that leads to cell-cycle arrest and apoptosis. We further show that an orally available small-molecule DHODH inhibitor demonstrates potent mono-therapeutic efficacy against patient-derived MB xenografts in vivo. The reprogramming of pyrimidine metabolism in MYC-driven medulloblastoma represents an unappreciated therapeutic strategy and a potential new class of treatments with stronger cancer selectivity and fewer neurotoxic sequelae.

Predicted Position Error Triggers Catch-Up Saccades during Sustained Smooth Pursuit.

For humans, visual tracking of moving stimuli often triggers catch-up saccades during smooth pursuit. The switch between these continuous and discrete eye movements is a trade-off between tolerating sustained position error (PE) when no saccade is triggered or a transient loss of vision during the saccade due to saccadic suppression. de Brouwer et al. (2002b) demonstrated that catch-up saccades were less likely to occur when the target re-crosses the fovea within 40-180 ms. To date, there is no mechanistic explanation for how the trigger decision is made by the brain. Recently, we proposed a stochastic decision model for saccade triggering during visual tracking (Coutinho et al., 2018) that relies on a probabilistic estimate of predicted PE (PEpred). Informed by model predictions, we hypothesized that saccade trigger time length and variability will increase when pre-saccadic predicted errors are small or visual uncertainty is high (e.g., for blurred targets). Data collected from human participants performing a double step-ramp task showed that large pre-saccadic PEpred (>10°) produced short saccade trigger times regardless of the level of uncertainty while saccade trigger times preceded by small PEpred (<10°) significantly increased in length and variability, and more so for blurred targets. Our model also predicted increased signal-dependent noise (SDN) as retinal slip (RS) increases; in our data, this resulted in longer saccade trigger times and more smooth trials without saccades. In summary, our data supports our hypothesized predicted error-based decision process for coordinating saccades during smooth pursuit.

A Randomized Trial to Evaluate the Effect of Two Topical Anesthetics on Pain Response During Frenotomy in Young Infants.

Objective: To examine the comparative effectiveness of two topical anesthetics in controlling the pain associated with tongue-tie release (frenotomy) in young infants. Design: Randomized trial. Setting: A Pediatric Craniofacial Clinic. Subjects: Forty-two infants who were referred for frenotomy were randomly allocated to receive the topical anesthetic gel 2% tetracaine or 20% benzocaine applied prior to frenotomy. Frenotomies were videotaped. The primary outcome measure was the Neonatal Facial Coding System (NFCS) score. Secondary outcome measures included cry duration and a visual analog scale (VAS) assessed by the parents. Results: The two groups were comparable with regard to weight, age, gender, previous painful experience, and last feeding time. Median NFCS scores prior to frenotomy in the tetracaine and the benzocaine groups were 4.5 (IQR: 0.75­10.2) and 3.5 (IQR: 0­9.5), respectively (P = 0.89, 95% CI −3 to 4). During frenotomy, median NFCS score increased to 28 (IQR: 24.5­30.25) in the tetracaine group (P < 0.0001, median difference −22, 95% CI −24.5 to −19), and to 28 (IQR: 26­30) in the benzocaine group (P < 0.0001, median difference −23, 95% CI −27 to −17). Mean cry durations in the tetracaine and the benzocaine groups were 69.4 seconds and 63.9 seconds, respectively (P = 0.32, 95% CI −47 to 15), and mean VAS scores were 57.2 and 58.2, respectively (P = 0.89, 95% CI −15.2 to 13.4). Conclusions: These topical anesthetics seem ineffective in controlling the pain associated with frenotomy. Clinicians should continue to search for an effective treatment for this procedure.