Search

Deep neural networks have been widely adopted in numerous domains due to their high performance and accessibility to developers and application-specific end-users. Fundamental to image-based applications is the development of Convolutional Neural Networks (CNNs), which possess the ability to automatically extract features from data. However, comprehending these complex models and their learned representations, which typically comprise millions of parameters and numerous layers, remains a challenge for both developers and end-users. This challenge arises due to the absence of interpretable and transparent tools to make sense of black-box models. There exists a growing body of Explainable Artificial Intelligence (XAI) literature, including a collection of methods denoted Class Activation Maps (CAMs), that seek to demystify what representations the model learns from the data, how it informs a given prediction, and why it, at times, performs poorly in certain tasks. We propose a novel XAI visualization method denoted CAManim that seeks to simultaneously broaden and focus end-user understanding of CNN predictions by animating the CAM-based network activation maps through all layers, effectively depicting from end-to-end how a model progressively arrives at the final layer activation. Herein, we demonstrate that CAManim works with any CAM-based method and various CNN architectures. Beyond qualitative model assessments, we additionally propose a novel quantitative assessment that expands upon the Remove and Debias (ROAD) metric, pairing the qualitative end-to-end network visual explanations assessment with our novel quantitative "yellow brick ROAD" assessment (ybROAD). This builds upon prior research to address the increasing demand for interpretable, robust, and transparent model assessment methodology, ultimately improving an end-user's trust in a given model's predictions. Examples and source code can be found at: https://omni-ml.github.io/pytorch-grad-cam-anim/. [ABSTRACT FROM AUTHOR]

Introduction: Racial differences in adverse maternal and birth outcomes have been studied in other countries, however, there are few studies specific to the Canadian population. In this study, we sought to examine the inequities in adverse perinatal outcomes between Black and White pregnant people in Ontario, Canada. Methods: We conducted a population-based retrospective cohort study that included all Black and White pregnant people who attended prenatal screening and had a singleton birth in any Ontario hospital (April 1st, 2012-March 31st, 2019). Poisson regression with robust error variance models were used to estimate the adjusted relative risks of adverse perinatal outcomes for Black people compared with White people while adjusting for covariates. Results: Among 412,120 eligible pregnant people, 10.1% were Black people and 89.9% were White people. Black people were at an increased risk of gestational diabetes mellitus, preeclampsia, placental abruption, preterm birth (<37, <34, <32 weeks), spontaneous preterm birth, all caesarean sections, emergency caesarean section, low birth weight (<2500g, <1500g), small-for-gestational-age (<10th percentile, <3rd percentile) neonates, 5-minute Apgar score <4 and <7, neonatal intensive care unit admission, and hyperbilirubinemia requiring treatment but had lower risks of elective caesarean section, assisted vaginal delivery, episiotomy, 3rd and 4th degree perineal tears, macrosomia, large-for-gestational-age neonates, and arterial cord pH≤7.1, as compared with White people. No difference in risks of gestational hypertension and placenta previa were observed between Black and White people. Conclusion: There are differences in several adverse perinatal outcomes between Black and White people within the Ontario health care system. Findings might have potential clinical and health policy implications, although more studies are needed to further understand the mechanisms. [ABSTRACT FROM AUTHOR]

Objective: To develop and internally validate a deep-learning algorithm from fetal ultrasound images for the diagnosis of cystic hygromas in the first trimester. Methods: All first trimester ultrasound scans with a diagnosis of a cystic hygroma between 11 and 14 weeks gestation at our tertiary care centre in Ontario, Canada were studied. Ultrasound scans with normal nuchal translucency were used as controls. The dataset was partitioned with 75% of images used for model training and 25% used for model validation. Images were analyzed using a DenseNet model and the accuracy of the trained model to correctly identify cases of cystic hygroma was assessed by calculating sensitivity, specificity, and the area under the receiver-operating characteristic (ROC) curve. Gradient class activation heat maps (Grad-CAM) were generated to assess model interpretability. Results: The dataset included 289 sagittal fetal ultrasound images;129 cystic hygroma cases and 160 normal NT controls. Overall model accuracy was 93% (95% CI: 88–98%), sensitivity 92% (95% CI: 79–100%), specificity 94% (95% CI: 91–96%), and the area under the ROC curve 0.94 (95% CI: 0.89–1.0). Grad-CAM heat maps demonstrated that the model predictions were driven primarily by the fetal posterior cervical area. Conclusions: Our findings demonstrate that deep-learning algorithms can achieve high accuracy in diagnostic interpretation of cystic hygroma in the first trimester, validated against expert clinical assessment. [ABSTRACT FROM AUTHOR]

Streptococcus pyogenes (group A Streptococcus) is a globally disseminated and human-adapted bacterial pathogen that causes a wide range of infections, including scarlet fever. Scarlet fever is a toxin-mediated disease characterized by the formation of an erythematous, sandpaper-like rash that typically occurs in children aged 5 to 15. This infectious disease is caused by toxins called superantigens, a family of highly potent immunomodulators. Although scarlet fever had largely declined in both prevalence and severity since the late 19th century, outbreaks have now reemerged in multiple geographical regions over the past decade. Here, we review recent findings that address the role of superantigens in promoting a fitness advantage for S. pyogenes within human populations and discuss how superantigens may be suitable targets for vaccination strategies. [ABSTRACT FROM AUTHOR]

Background: Open appendectomy, laparoscopic appendectomy, and non-surgical treatment are three options to treat acute appendicitis during pregnancy. Previous studies on the association of different treatment methods for acute appendicitis with pregnancy outcomes have been limited by small sample sizes and residual confounding, especially with respect to hospital-level factors. This study aimed to investigate the association of treatment method for acute appendicitis with pregnancy outcomes using a multi-level analysis. Methods: A retrospective cohort study was conducted based on a large electronic health records database in the United States during the period 2000 to 2016. All pregnancies diagnosed with acute appendicitis and treated in participating hospitals during the study period were included. We conducted multi-level hierarchical logistic regression to analyze both individual- and hospital-level factors for abortion, preterm labor, and cesarean section. Results: A total of 10,271 acute appendicitis during pregnancy were identified during the study period. Of them, 5,872 (57.2%) were treated by laparoscopic appendectomy, 1,403 (13.7%) by open appendectomy, and 2,996 (29.2%) by non-surgical treatment. Compared with open appendectomy, both laparoscopic appendectomy (adjusted OR, 0.6, 95% CI, 0.4, 0.9) and non-surgical treatment (adjusted OR, 0.4; 95% CI, 0.3–0.7) showed a decreased risk of preterm labor. Other important individual-level determinants of adverse pregnancy outcomes included maternal age, gestational hypertension, and anemia during pregnancy, the hospital-level determinant included the number of beds. Conclusions: Compared with open appendectomy, both laparoscopic appendectomy and non-surgical treatment may be associated with a lower risk of preterm labor, without increased risks of abortion and cesarean section. [ABSTRACT FROM AUTHOR]

Background: Given that pregnant women taking medications are excluded from clinical trials, real-world evidence is essential. We aimed to build a Canadian Mother-Child Cohort Active Surveillance Initiative (CAMCCO) and compare frequency of prematurity, low-birth-weight (LBW), major malformations, multiplicity, and gestational medication use across four provinces., Methods: CAMCCO is a collaborative research infrastructure that uses real-world data from large provincial health care databases in Canada; developed with standardized methods to similarly construct population-based pregnancy/child cohorts with longitudinal follow-up by linking administrative/hospital/birth databases. CAMCCO also includes a common repository to i) share algorithms and case definitions based on diagnostic and procedural codes for research/training purpose, and ii) download aggregate data relevant to primary care providers, researchers, and decision makers. For this study, data from Quebec (1998-2015), Manitoba (1995-2019), Saskatchewan (1996-2020), and Alberta (2005-2018) are compared (Chi-square tests, p-values), and trends are calculated using Cochran-Armitage trend tests., Results: Almost two-thirds (61%) of women took medications during pregnancy, mostly antibiotics (26%), asthma drugs (8%), and antidepressants (4%). Differences in the prevalence of prematurity (5.9-6.8%), LBW (4.0-5.2%), and multiplicity (1.0-2.5%) were statistically significant between provinces (p<0.001). Frequency of major malformations increased over time in Quebec (7-11%; p<0.001), Saskatchewan (5-11%; p<0.001), and Alberta (from 7-8%; p<0.001), and decreased in Manitoba (5-3%; p<0.001). Cardiovascular and musculoskeletal malformations were the most prevalent., Interpretation: Medications are often used among Canadian pregnancies but adverse pregnancy outcomes vary across provinces. Digitized health data may help researchers and care providers understand the risk-benefit ratios related to gestational medication use, as well as province-specific trends., Competing Interests: The authors have declared that no competing interests exist.