Canadian Institute Research Articles

Sophisticated deep learning with on-chip optical diffractive tensor processing

Ever-growing deep-learning technologies are making revolutionary changes for modern life. However, conventional computing architectures are designed to process sequential and digital programs but are burdened with performing massive parallel and adaptive deep-learning applications. Photonic integrated circuits provide an efficient approach to mitigate bandwidth limitations and the power-wall brought on by its electronic counterparts, showing great potential in ultrafast and energy-free high-performance computation. Here, we propose an optical computing architecture enabled by on-chip diffraction to implement convolutional acceleration, termed “optical convolution unit” (OCU). We demonstrate that any real-valued convolution kernels can be exploited by the OCU with a prominent computational throughput boosting via the concept of structral reparameterization. With the OCU as the fundamental unit, we build an optical convolutional neural network (oCNN) to implement two popular deep learning tasks: classification and regression. For classification, Fashion Modified National Institute of Standards and Technology (Fashion-MNIST) and Canadian Institute for Advanced Research (CIFAR-4) data sets are tested with accuracies of 91.63% and 86.25%, respectively. For regression, we build an optical denoising convolutional neural network to handle Gaussian noise in gray-scale images with noise level σ=10, 15, and 20, resulting in clean images with an average peak signal-to-noise ratio (PSNR) of 31.70, 29.39, and 27.72 dB, respectively. The proposed OCU presents remarkable performance of low energy consumption and high information density due to its fully passive nature and compact footprint, providing a parallel while lightweight solution for future compute-in-memory architecture to handle high dimensional tensors in deep learning.

CCTG EN10: A phase II study of tailored adjuvant therapy in POLE-mutated and p53-wildtype/NSMP early-stage endometrial cancer (EC)—RAINBO BLUE and TAPER.

TPS5632 Background: Molecular classification of EC provides an opportunity to personalize treatment for patients with early disease based on prognosis. We hypothesize that de-escalated adjuvant treatment in patients with early-stage POLE-mutated ( POLEmut) or p53wt/NSMP (p53 wildtype/no specific molecular profile) EC is feasible and associated with favourable disease control, QOL and health economics. Methods: EN.10 is a Canadian Cancer Trials Group led phase II trial that enrolls two independent cohorts of patients with early stage EC based on tumour molecular status: POLEmut (substudy A) and p53wt/NSMP (substudy B). The primary objective is to estimate the 3-year rate of pelvic recurrence. Secondary objectives include estimated 3-year isolated vaginal recurrence, para-aortic recurrence, distant metastasis, survival (recurrence-free, EC-specific, overall survival) and to describe the patient decisional conflict and fear of recurrence. Tertiary objectives: QOL, economic evaluation and interrogation/discovery of additional prognostic biomarkers. Statistical Design: Registration to Substudy A or B is based on molecular classification. Treatment will be deescalated in both substudies; to surveillance-only (all early stage POLEmut, and a proportion of NSMP), or vaginal brachytherapy (proportion of NSMP), with NSMP stratification based on histological grade, and lymphovascular invasion. Substudy A can enroll endometrioid, serous, clear cell, un/dedifferentiated and carcinosarcomas whereas Substudy B is confined to endometrioid histotype. Both arms require ECOG 0-2 and hysterectomy+BSO with nodal assessment required for stage II and grade 3 EC. A total of 120 patients with POLEmut (Sub-study A) and 180 with p53wt/NSMP ER+ (Sub-study B) EC will be recruited in 3 years. If the upper 95% confidence limit is less than 5% in either of Substudy A, Cohort A1 and Sub-study B or both, it will be concluded that patients in one or both groups have an acceptable low risk of pelvic recurrence at 3 years with molecularly-tailored de-escalated adjuvant treatment An additional estimated 25 patients with higher risk POLEmut EC will be enrolled for an exploratory analysis. Conduct to Date: Study activation Dec 19 2022. Enrollment January 31, 2023: 36 Supported by Canadian Cancer Society, Canadian Institutes of Health Research, Gynecologic Cancer Initiative, BC Cancer Foundation. Clinical trial information: NCT05640999 .

F.5 Spinal column and spinal cord injuries secondary to mountain biking accidents: a 15-year review at a provincial spine referral centre

Background: Mountain biking (MTB) is an increasingly popular sport that has been associated with serious spinal injuries, which can have devastating effects on patients and significant impacts on healthcare resources. Herein, we characterized the occurrence of these MTB spinal injuries over a 15-year period and analyzed the affiliated acute-care hospital costs. Methods: Patients seen at Vancouver General Hospital for MTB spinal injuries between 2008-2022 were retrospectively reviewed. Demographics, injury details, treatments, outcomes, and resource requirements for acute hospitalization were collected. The Canadian Institute for Health Information was referenced for cost analysis. Results: Over the 15 years of analysis, 149 MTB spinal injuries occurred. The majority (87.2%) were male. 59 (39.6%) were associated with spinal cord injury; most of these were in the cervical spine (72.3%) and majority were AIS Grade A (36.1%). 102 patients (68.5%) required spine surgery; 26 (17.4%) required intensive care; 34 (22.8%) required inpatient rehabilitation. Mean length of stay was 13.5 days and acute admission costs for the healthcare system averaged $35,251 (95% CI $27,080-$43,424). Conclusions: MTB spinal injuries are associated with significant medical, personal, and financial burden. As injury prevention remains paramount, further investigation of the roles of education and safety measures is recommended.

CCTG HN11: SPECT-CT guided elective contralateral neck treatment (SELECT) for patients with lateralized oropharyngeal cancer—A phase III randomized controlled trial.

TPS6114 Background: Lymphatic mapping identifies neck lymph nodes at risk for cancer spread in patients with lateralized oropharyngeal (OPC) squamous cell carcinoma. We hypothesize that a lymphatic mapping guided approach to radiotherapy (RT) treatment of the contralateral neck will enable safe de-escalation of therapy with acceptable disease control with potential benefits in RT related toxicity, QOL, swallowing function and economics. Methods: HN.11 is a Canadian Cancer Trials Group international multi-centre, non-inferiority randomized phase III trial comparing a lymphatic mapping-guided approach for management of the contralateral neck (experimental) vs. bilateral neck radiation therapy (RT) (control). The primary objective is to determine if a lymphatic mapping-guided approach for management of the contralateral neck has a non-inferior disease-free survival (DFS) compared to bilateral neck RT. Secondary objectives: To compare swallowing related QOL, xerostomia, isolated contralateral neck failure, overall survival, loco-regional failure, distant metastases, RT-related toxicities, patient reported adverse events (PRO CTCAE), gastrostomy tube usage, and economics indicators (resource utilization, lost productivity, financial toxicity, EQ5D). Exploratory objectives: Swallowing function using video fluoroscopic swallow studies, head and neck-cancer specific QOL, patterns of lymphatic drainage, radiomic prediction of contralateral lymphatic drainage, correlation of tumour somatic mutations and ctDNA with disease recurrence. Statistical Design: The target sample size is 510 patients. The experimental arm will be considered non-inferior if the upper limit of the one sided 95% confidence interval (CI) of the estimated hazard ratio (HR) does not exceed 1.46 (non-inferiority margin of 6.5% for 2-year DFS). In the primary intention-to-treat (ITT) analysis, the study has 81% power, with a one-sided type I error rate of 5%, assuming 5 years of accrual, 3-years of follow up to observe 178 DFS events, and a 5% loss to follow up. If non-inferiority is demonstrated in the ITT analysis, a secondary per-protocol analysis will be performed, analyzing those patients treated as per their allocation, with 80% power to detect non-inferiority of the experimental arm assuming a conservative 10% cross-over from the experimental to the control arm (i.e. failed lymphatic mapping). Conduct to Date: Study activation September 29, 2022. First enrollment was February 10, 2023. Supported by Canadian Cancer Society, Canadian Institutes of Health Research, National Cancer Trials Network (NCTN). Clinical trial information: NCT05451004 .

Visual skills essential for rugby.

Keen vision is one of the most important qualities required of athletes. It enables players to perform sports-related drills and apply decision-making skills. To accurately measure the visual ability of athletes, it is important to first identify the variety of visual skills involved in the particular sport. The objectives of this novel review are to identify the most important visual skills required for rugby, and to create a reference point for further studies to include visual skills essential to rugby players. We conducted an electronic search with various combinations of relevant keywords using the following databases: Sport Discuss, Ovid's Evidence-Based Medicine Reviews, PubMed/MEDLINE, Current Contents, Science Direct, the National Research Council's Canada Institute for Scientific and Technical Information, Cochrane Database of Systematic Reviews, Google Scholar, and international electronic catalogues to assess the scientific literature related to the visual skills required for rugby. Only the records published in English were included. We extracted data on the relationship between vision and match performance, the defined problem or purpose of the study, and the inclusion of theoretical definitions of tactical behaviors. Our search yielded 80 records, 51 of which fulfilled the inclusion criteria. The most important visual skills in rugby are classified based on whether they meet the requirements for visual hardware or visual software skills. Visual hardware skills include visual acuity, depth perception, fusion flexibility, and contrast sensitivity; visual software skills include eye tracking, hand-eye coordination, eye focusing, peripheral vision, speed and span of recognition, visual response time, and visual memory. Rugby players must use both visual hardware and software skills to reliably observe their teammates' positions, understand their opponents' actions and tactics, handle the ball, analyze the immediate circumstances, and anticipate what will occur. Further studies are needed to verify the significance of each visual skill in actual competition to determine a relationship between vision and the results of a championship.

Comparing FTP and SSH Password Brute Force Attack Detection using k-Nearest Neighbour (k-NN) and Decision Tree in Cloud Computing

Cloud computing represents a new epoch in computing. From huge enterprises to individual use, cloud computing always provides an answer. Therefore, cloud computing must be readily accessible and scalable, and customers must pay only for the resources they consume rather than for the entire infrastructure. With such conveniences, come with their own threat especially brute force attacks since the resources are available publicly online for the whole world to see. In a brute force attack, the attacker attempts every possible combination of username and password to obtain access to the system. This study aims to examine the performance of the k-Nearest Neighbours (k-NN) and Decision Tree algorithms by contrasting their precision, recall, and F1 score. This research makes use of the CICIDS2017 dataset, which is a labelled dataset produced by the Canada Institute for Cybersecurity. A signature for the brute force attack is utilised with an Intrusion Detection System (IDS) to detect the attack. This strategy, however, is ineffective when a network is being attacked by a novel or unknown attack or signature. At the conclusion of the study, the performance of both algorithms is evaluated by comparing their precision, recall, and f1 score. The results show that Decision Tree performs slightly better than k-NN at classifying FTP and SSH attacks.

0932 Sleep Difficulties: Prevalence, relationship with psychological outcomes, and interindividual differences in student-athletes

Abstract Introduction Mental health issues are a leading cause of decreased academic and athletic performances when left untreated. Recently, the International Olympic Committee (IOC) and the National Collegiate Athletic Association (NCAA) have highlighted the fundamental importance of mental health among young athletes. Sleep represents a core feature of mental health, and student-athletes represent a vulnerable population for sleep restriction and disruption due to the demands of their courseload and training schedule. This investigation was performed to identify differences in sleep and mental health across gender, race/ethnicity, and sport type. Methods Data were obtained throughout the academic semester from Canadian institutions, with student-athletes completing the Athlete Sleep Screening Questionnaire (ASSQ), General Anxiety Disorder-7 (GAD-7), and Patient Health Questionnaire (PHQ-9). ASSQ captured sleep and circadian health, while GAD-7 and PHQ-9 assessed anxiety and depressive symptoms, respectively. Linear regression determined gender, race/ethnicity, and sport type differences across ASSQ sleep problems score, ASSQ sleep items, GAD-7 score, PHQ-9 score, and PHQ-9 item #9 (suicidal ideation). Results Final sample (N=559) was 53.0% female, 79.1% white, comprised of young individual (Mean Age = 18.9 ± 3.21), and included student-athletes from basketball, field-hockey, football, ice-hockey, soccer, volleyball, track & field and swimming. Females reported more overall sleep difficulties (p=0.01), medication use (p=0.04) and WASO (p=0.001) compared to males. Females also reported higher anxiety (p < 0.0001), depression (p< 0.0001), and suicidal ideation (p=0.03). Significant variance in total sleep time (TST) (p=0.001) and sleep satisfaction (p=0.04) was observed across race/ethnicity, with student-athletes identifying as Black or African American reporting the lowest TST, while Asian or Pacific Islander student-athletes reported the lowest sleep satisfaction. Significant variance in sleep difficulties score (p< 0.0001), TST (p < 0.0001), and satisfaction (p< 0.0001), as well as anxiety (p < 0.0001) and depression (p< 0.0001) was observed across sport type. Football student-athletes reported the highest sleep difficulties score, and lowest TST satisfaction. Field-hockey student-athletes reported the highest anxiety and depression. Conclusion Results suggest that sleep and mental health challenges present differently across gender, race/ethnicity, and sport type. Clarifying these unique and specific challenges to student-athlete sleep and mental health is necessary to enhance and personalize interventions. Support (if any) Mitacs #IT23469

0827 Effect of virtual cognitive behavioral therapy for insomnia on perceived cognitive functioning among cancer survivors

Abstract Introduction Comorbid insomnia and perceived cognitive impairments (PCI) affect up to 26% of individuals diagnosed with cancer. Given the association between sleep and cognition, and the lack of interventions to improve PCI, the present study examined the impact of cognitive behavioral therapy for insomnia (CBT-I) on PCI among cancer survivors with insomnia disorder and cognitive complaints (clinicaltrials.gov:NCT04026048). Methods Cancer survivors (N=122) with insomnia and PCI were randomized to 7 weekly virtual CBT-I treatment sessions (n=56) or waitlist control (WLC: n=66). Participants completed the Insomnia Severity Index and the Functional Assessment of Cancer Therapy – Cognitive Function upon entering the study (baseline; T0), and then again at 4 (T1), and 8 (T2) weeks. A series of 2 (treatment group: CBT-I and WLC) by 3 (time: T0, T1, T2) mixed ANOVAs were performed to assess changes in insomnia symptoms, PCI, perceived cognitive abilities (PCA), and impact of cognitive function on quality of life (QOL). Results Participants were, on average, aged 60 years, had 16 years of education, and 79% were female. Breast cancer was the most reported cancer type (46%). No significant differences were observed between groups at T0 for demographic or clinical variables (p>.05). There was a significant group by time interaction for insomnia [p=<.001, partial eta squared (pes)=.36]. Individuals randomized to CBT-I reported significant decreases in insomnia symptoms post treatment compared to WLC (CBT-I: -11.5; WLC: -2.7). Significant interactions were also observed for PCI (p=<.001, pes=.11), PCA (p=<.001, pes=.16), and impact on QOL (p=<.001, pes=.15). Participants randomized to CBT-I reported significantly less PCI (CBT-I: 14.7; WLC: 4.1), better PCA (CBT-I: 5.8; WLC: 1.1), and less impact on QOL (CBT-I: 4.9; WLC: 1.1) than participants in the WLC. Conclusion In addition to improving insomnia, CBT-I significantly improve perceived cognitive functioning and quality of life among cancer survivors. Further research is needed to understand the mechanisms underlying improvements in cognition following CBT-I. Support (if any) Dr. Sheila Garland is supported by a Canadian Cancer Society Emerging Scholar Award (Survivorship) (grant #707146). This project was funded through a grant from the Canadian Institutes of Health Research (CIHR) (grant number: PJT 162428) and the Beatrice Hunter Cancer Research Institute.

0324 Protective and Risk Factors for Insomnia Over 5 Years in a Population-Based Sample of Adults

Abstract Introduction Insomnia is a major public health concern and one of the most prevalent health issues among adults. Identifying protective and risk factors for insomnia could help identify vulnerable populations and design public health interventions to prevent some adverse medical and mental health consequences. The study objective was to identify protective and risk factors for insomnia in a population-based sample of adults. Methods Data is from a large epidemiological cohort study on the natural course of insomnia conducted in Canada before the COVID-19 pandemic. Insomnia and potential protective and risk factors for insomnia were measured annually over 5 years with validated questionnaires (e.g., Insomnia Severity Index, Pittsburgh Sleep Quality Index, State-Trait Anxiety Inventory, Beck Depression Inventory, Perceived Stress Scale, Coping Inventory for Stressful Situations, Ford Insomnia Response to Stress Test, Arousal Predisposition Scale, Life Experiences Survey). Risk factors for insomnia were divided as predisposing or precipitating factors. Results From a cohort of 3,413 participants 1,709 adults were identified as good sleepers at baseline and were included in the analyses. A total of 202 people developed an insomnia syndrome during the 5-year follow-up. Using survival analysis for discrete events, the following variables were significant predisposing factors for insomnia: anxiety (HR=1.037; p=0.018), depression (HR=1.082; p< 0.001), perceived stress (HR=1.073; p< 0.001), number of negative life events (HR=1.478; p=0.02), perceived worse general health (HR=1.606; p=0.006), pain (HR=1.475; p=0.002), vulnerability to stress (HR=1.089; p< 0.001), and maladaptive emotional coping (HR=1.044; p< 0.001). The following variables were identified as precipitating factors for insomnia: loss of income (HR=1.346; p=0.03), increased anxiety (HR=1.070, p< 0.001), depression (HR=1.119; p< 0.001) and perceived stress (HR=1.083; p=0.001), decline in perceived general health (HR=2.061; p=0.006), and increased pain (HR=1.353; p=0.013). Physical activity was not a significant protective factor for insomnia (p=0.54). Conclusion These results provide new information that could be helpful to prevent the onset and chronicity of insomnia by targeting specific precipitating factors, and people with predisposing factors for insomnia. This is important as insomnia tends to be a persistent condition and is known to increase adverse medical and mental health outcomes. Support (if any) This study was funded by the Canadian Institutes of Health Research.

0861 Does Cognitive Behavioral Therapy for Insomnia Reduce Productivity Loss in Cancer Survivors with Cognitive Impairment?

Abstract Introduction Insomnia and perceived cognitive impairments (PCI) affect 1 in 4 cancer survivors and negatively affects work productivity. This study explores the impact of cognitive behavioral therapy for insomnia (CBT-I) on work presenteeism and absenteeism in cancer survivors with comorbid insomnia and PCI. Methods A sub-sample of 42 Atlantic Canadian survivors with insomnia disorder and PCI who were currently working for pay were analyzed as part of a larger randomized clinical trial. Participants completed 7 virtual sessions of CBT-I. The Work Productivity and Activity Impairment Index Specific Health Problem Questionnaire (WPAI: SHP) was administered pre-and post-treatment. Absenteeism was defined as the percentage of time absent from work, while presenteeism was defined as the extent to which health problems affected productivity at work in the past 7 days. Costs were calculated by the human capital method. Significant changes in productivity loss were defined at ≥15% or 20% change in absenteeism and presenteeism, respectively. Descriptive statistics and paired samples t-tests were used to measure changes in productivity following treatment. Average net benefit was calculated by subtracting the average treatment cost from average treatment benefits after 1 year to reflect cost savings after CBT-I. The cost of treatment was estimated in terms of average cost per session and lost time at work. Results Participants (76% Female) were, on average, 50.4 years old with 16.6 years of education. Breast cancer was the most reported cancer type (43%). Overall rates of presenteeism were significantly lower at post-treatment (p=<.001, d=0.55), with 52.4% reporting significant reductions. Absenteeism rates from pre- to post-treatment were not significant (p=.11, d=0.20). Reduced productivity loss post-treatment was associated with an average savings of CAD$10,979/person/year. Average net benefits for the first-year post-treatment equaled CAD$9348/person. Conclusion CBT-I can improve work productivity in cancer survivors. Investments to increase and provide access to evidence-based treatment may lead to significant savings for individuals and businesses. Support (if any) Dr. Sheila Garland is supported by a Canadian Cancer Society Emerging Scholar Award (Survivorship) (grant #707146). This project was funded through a grant from the Canadian Institutes of Health Research (CIHR) (grant number: PJT 162428) and the Beatrice Hunter Cancer Research Institute.

Organizational and Institutional Initiatives for the Recruitment, Retention, and Advancement of Black Nurses in the Canadian Healthcare System: A Rapid Review Protocol

Introduction: In response to protests for racial justice, several organizations and institutions have made public declarations denouncing anti-Black racism. One prominent sector emphasizing their commitment to addressing anti-Black racism is health care—more specifically, nursing. To address anti-Black racism, many Canadian organizations and institutions have announced initiatives to recruit, retain, and support the advancement of Black nurses. Our team is interested in charting these initiatives to inform future policy and program decisions related to the recruitment, retention, and advancement of Black nurses. Objective: The objective of this review is to identify and chart evidence of organizational and institutional initiatives related to the recruitment, retention, and advancement of Black nurses in Canada. Inclusion criteria: This rapid review will include sources focused on Black nurses in Canada. Further, this review is focused on the organizational or institutional initiatives that support or facilitate aspects of recruitment, retention, or advancement of Black nurses in the workforce in Canada. Methods: A comprehensive search, developed in collaboration with a library scientist, will be used to locate peer-reviewed and grey literature from select databases and repositories. Databases will be searched from time of inception, and language will be restricted to English and French sources. Title and abstract screening as well as full-text review will each be completed by two independent reviewers. Sources will be included if they meet the inclusion criteria for the population, concept, and context. Data will be extracted by two reviewers using an extraction tool. Data will be reviewed and consolidated before being presented narratively and visually.Protocol Registration: The protocol has been registered with Open Science Framework (OSF) on March 1st, 2023.

(025) Disclosure of Mid-Urethral Sling and Radical Prostatectomy Risks: A Comparative Analysis of Hospital-Affiliated Online Resources

Abstract Introduction Radical prostatectomy (RP) involves the removal of the prostate gland to treat cancer. The mid-urethral sling (MUS) procedure, used to correct symptoms of stress urinary incontinence in women, affects anterior vaginal wall female periurethral tissue (AVW-FPT). Glandular structures in AVW-FPT have similar immunohistochemistry to prostatic tissue. Erectile dysfunction and female orgasmic disorder have been reported following RP and MUS, respectively. The disclosure of these, and other associated risks, in patient education materials has yet to be assessed. Further investigation is required to determine if and how risk information reported in the literature is presented to patients, as it may impact the ability of patients to make informed health-related decisions. Objective To assess the level of knowledge translation to patients of reported risks associated with MUS and RP procedures in hospital affiliated online resources (HAORs) from Canada, Australia, and the United Kingdom, and to evaluate the accessibility of this information using validated measures of health literacy. Methods A list of healthcare institutions in Canada, Australia, and the UK was generated. RP and MUS-specific HAORs were collected from all institutions. Qualitative analysis of 35 HAORs (RP n=20 and MUS n=15) was conducted using NVivo12. The Patient Education Materials Assessment Tool- Print version (PEMAT-P), was used to determine understandability and actionability, and the Flesch-Kincaid Grade Level (FKGL), Gunning Fox Index (GFI), and Simple Measure of Gobbledygook (SMOG) were used to assess the readability. Results Most RP and MUS HAORs were written higher than the recommended grade 5/6 reading level and were below the threshold values for understandability and actionability. The majority of Canadian MUS HAORs met the adequate threshold values for understandability and actionability, unlike the MUS HAORs from Australia and the UK and the RP resources from all three regions. There was inconsistency between the risks reported in the RP and MUS HAORs from Canada and Australia compared to the risks commonly reported in literature. The organization of information including the use of diagrams and medical jargon in the HAORs differed between RP and MUS documents and across the three regions. RP HAORs used diagrams more consistently than MUS HAORs and scored higher on measures of understandability and actionability. The RP and MUS HAORs from the three regions that used medical jargon scored lower on measures of readability. There was inconsistency in the reporting of sexual health risks between the RP and MUS HAORs in all three regions. MUS HAORs less consistently reported sexual health risks following surgery compared to RP HAORs. Conclusions HAORs are important means by which health-related information is translated to patients. There were inconsistencies in the risk information reported for RP versus MUS in Canada, Australia, and the UK. This discordance may impact patient decision-making and health outcomes. The development of guidelines for HAORs curation for healthcare institutions operating under similar healthcare frameworks may be beneficial in ensuring equal access to health information for all patients. Disclosure No

The impact of excessive epicardial adipose tissue accumulation on cardiorespiratory fitness in adults without cardiac disease: a cross-sectional magnetic resonance imaging study

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Canadian Institutes of Health Research. Background/Introduction Excessive epicardial adipose tissue (EAT) is associated with worse exercise capacity and is thought to play a role in the pathophysiology of heart failure with preserved ejection fraction. Few studies have explored the relationship between EAT, cardiorespiratory fitness in men and women without cardiovascular disease. Purpose To investigate the association between EAT, cardiorespiratory fitness, and physical activity levels in asymptomatic men and women of variable degree of adiposity. Methods A total of 239 participants (43% women) free from cardiovascular disease who underwent concurrent quantification of EAT volume assessed by cardiac magnetic resonance and a cardiopulmonary exercise testing (CPET) (VO2peak [ml/kg/min]) were studied. Physical activity levels/intensity were measured by questionnaire and using pedometer. Participants were categorized into low-EAT, moderate-EAT, and high-EAT groups on the basis of EAT volume. Results Mean age of participants was 49.8 ± 8.3 years, BMI 26.0 ± 3.7 kg/m2 (15% with obesity), waist circumference 89 ± 11 cm, visceral adipose tissue volume 91.3 ± 51.3 cm3, and EAT volume 9.8 ± 3.1 cm3. EAT volume decreased with increment in cardiorespiratory fitness in men and women of variable adiposity (p<0.05). Men and women with high-EAT had worse cardiometabolic profile (visceral adiposity, insulin resistance, high blood pressure, dysglycemia, atherogenic dyslipidemia), and lower cardiorespiratory fitness (VO2peak) than those with low-EAT (all p<0.05). Physical activity levels were comparable across EAT tertiles (p=NS). The association between EAT and cardiorespiratory fitness was no longer significant after controlling for visceral adiposity. Conclusions In men and women of variable degree of adiposity, excessive EAT accumulation is associated with visceral adiposity and worse cardiorespiratory fitness. Further longitudinal studies are needed to understand the interplay between EAT and cardiorespiratory fitness, and the development of obesity-related heart failure features.

FAST the Inside Track: Where We Are, Where Do We Want to Be, and How Do We Get There?

This is an overview of the development of FAST (Faceted Application of Subject Terminology) from its inception in the late 1990s, through its development and implementation to the work being undertaken by OCLC and the FAST Policy and Outreach Committee (FPOC) to develop and promote FAST. FPOC members explain how FAST is used by institutions in Canada, the United Kingdom, and the United States. They cover their experience of implementing FAST and the benefits they have derived. The final section considers the value of FAST as a faceted vocabulary and the potential for future development and linked data.

A deep learning framework for personalised dynamic diagnosis of graft fibrosis after liver transplantation: a retrospective, single Canadian centre, longitudinal study

Recurrent graft fibrosis after liver transplantation can threaten both graft and patient survival. Therefore, early detection of fibrosis is essential to avoid disease progression and the need for retransplantation. Non-invasive blood-based biomarkers of fibrosis are limited by moderate accuracy and high cost. We aimed to evaluate the accuracy of machine learning algorithms in detecting graft fibrosis using longitudinal clinical and laboratory data. In this retrospective, longitudinal study, we trained machine learning algorithms, including our novel weighted long short-term memory (LSTM) model, to predict the risk of significant fibrosis using follow-up data from 1893 adults who had a liver transplantation between Feb 1, 1987, and Dec 30, 2019, with at least one liver biopsy post transplantation. Liver biopsy samples with indefinitive fibrosis stage and those from patients with multiple transplantations were excluded. Longitudinal clinical variables were collected from transplantation to the date of last available liver biopsy. Deep learning models were trained on 70% of the patients as the training set and 30% of the patients as the test set. The algorithms were also separately tested on longitudinal data from patients in a subgroup of patients (n=149) who had transient elastography within 1 year before or after the date of liver biopsy. Weighted LSTM model performance for diagnosing significant fibrosis was compared against LSTM, other deep learning models (recurrent neural network and temporal convolutional network), and machine learning models (Random Forest, Support vector machines, Logistic regression, Lasso regression, and Ridge regression) and aspartate aminotransferase-to-platelet ratio index (APRI), fibrosis-4 index (FIB-4), and transient elastography. 1893 people who had a liver transplantation (1261 [67%] men and 632 [33%] women) with at least one liver biopsy between Jan 1, 1992, and June 30, 2020, were included in the study (591 [31%] cases and 1302 [69%] controls). The median age at liver transplantation was 53·7 years (IQR 47·3-59·0) for cases and 55·3 years (48·0 to 61·2) for controls. The median time interval between transplant and liver biopsy was 21 months (5 to 71). The weighted LSTM model (area under the curve 0·798 [95% CI 0·790 to 0·810]) consistently outperformed other methods, including unweighted LSTM (0·761 [0·750 to 0·769]; p=0·031) Recurrent Neural Network (0·736 [0·721 to 0·744]), Temporal Convolutional Networks (0·700 [0·662 to 0·747], and Random Forest 0·679 [0·652 to 0·707]), FIB-4 (0·650 [0·636 to 0·663]) and APRI (0·682 [0·671 to 0·694]) when diagnosing F2 or worse stage fibrosis. In a subgroup of patients with transient elastography results, weighted LSTM was not significantly better at detecting fibrosis (≥F2; 0·705 [0·687 to 0·724]) than transient elastography (0·685 [0·662 to 0·704]). The top ten variables predictive for significant fibrosis were recipient age, primary indication for transplantation, donor age, and longitudinal data for creatinine, alanine aminotransferase, aspartate aminotransferase, total bilirubin, platelets, white blood cell count, and weight. Deep learning algorithms, particularly weighted LSTM, outperform other routinely used non-invasive modalities and could help with the earlier diagnosis of graft fibrosis using longitudinal clinical and laboratory variables. The list of most important predictive variables for the development of fibrosis will enable clinicians to modify their management accordingly to prevent onset of graft cirrhosis. Canadian Institute of Health Research, American Society of Transplantation, Toronto General and Western Hospital Foundation, and Paladin Labs.

Geographic Virtual Pooling of Hospital Resources: Data-Driven Trade-off Between Waiting and Traveling

Problem definition: Patient-level data from 72 magnetic resonance imaging (MRI) hospitals in Ontario, Canada from 2013 to 2017 show that over 60% of patients exceeded their wait time targets. We conduct a data-driven analysis to quantify the reduction in the patient fraction exceeding (FET) target for MRI services through geographic virtual resource-sharing while limiting incremental driving time. We present a data-driven method to solve the geographic pooling problem of partitioning 72 hospitals with heterogeneous patients with different wait time targets located in a two-dimensional region into a set of clusters. Methodology/results: We propose an “augmented-priority rule,” which is a sequencing rule that balances the patient’s initial priority class and the number of days until her wait time target. We then use neural networks to predict patient arrival and service times. We combine this predicted information and the sequencing rule to implement “advance scheduling,” which informs the patient of her treatment day and location when requesting an MRI scan. We then optimize the number of geographic resource pools among the 72 hospitals using genetic algorithms. Our resource-pooling model lowers the FET from 66% to 36% while constraining the average incremental travel time below three hours. In addition, our model shows that only 10 additional scanners are needed to achieve 10% FET, whereas 50 additional scanners would be needed without resource sharing. Over 70% of the hospitals are not worse off financially. Each individual hospital, measured over at least two weeks, achieves a higher machine utilization and a lower FET. Managerial implications: Our paper provides a practical, data-driven geographical resource-sharing model that hospitals can readily implement. Our method achieves a near-optimal solution with low computational complexity. Using smart data-driven scheduling, a little extra capacity placed at the right location is all we need to achieve the desired FET under geographic resource-sharing. Funding: This paper is supported by the following grant: Canadian Institutes of Health Research (CIHR) [Grant CIHR-950-231935]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2023.1225 .

Critical Race Composite Counter Storytelling as Appropriate Methodology to “Wrestle the White Beast”

Being a Brown woman in academia remains a minority experience. Racialized students within the ivory tower consistently experience microaggressions and violence through institutionally biased university curricula, programs, and policies. Using personal storytelling and narrative to describe my experience of navigating academic dynamics in a public institution in Canada, this article seeks to demystify and dismantle the challenges of navigating graduate school as a woman of colour, specifically in relation to finding an appropriate methodology for my doctoral dissertation research. In this article I will unpack my use of critical race theory’s composite counter storytelling methodology. I look at my process for creating Beti, a composite counterstory of the ten racialized and Indigenous activists I interviewed, and some of the challenges and limitations I encountered in this process. This methodology seeks to improve access and inclusion for racialized students researching their own communities within academic institutions and beyond.

Surgical margin status in relation to surgical approach in the management of early-stage cervical Cancer: A Canadian cervical Cancer collaborative (4C) study

ObjectiveSurgical margin status in women undergoing surgery for early-stage cervical cancer is an important prognostic factor. We sought to determine whether close (<3 mm) and positive surgical margins are associated with surgical approach and survival. MethodsThis is a national retrospective cohort study of cervical cancer patients treated with radical hysterectomy. Patients with stage IA1/LVSI-Ib2(FIGO 2018) with lesions up to 4 cm at 11 Canadian institutions from 2007 to 2019 were included. Surgical approach included robotic/laparoscopic (LRH), abdominal (ARH) or combined laparoscopic-assisted vaginal/vaginal (LVRH) radical hysterectomy. Recurrence free survival(RFS) and overall survival (OS) were estimated using Kaplan-Meier analysis. Chi-square and log-rank tests were used to compare groups. Results956 patients met inclusion criteria. Surgical margins were as follows: negative (87.0%), positive (0.4%) or close <3 mm (6.8%), missing (5.8%). Most patients had squamous histology (46.9%); 34.6% had adenocarcinomas and 11.3% adenosquamous. Most were stage IB (75.1%) and 24.9% were IA. Mode of surgery included: LRH(51.8%), ARH (39.2%), LVRH (8.9%).Predictive factors for close/positive margins included stage, tumour diameter, vaginal involvement and parametrial extension. Surgical approach was not associated with margin status (p = 0.27). Close/positive margins were associated with a higher risk of death on univariate analysis (HR = non calculable for positive and HR = 1.83 for close margins, p = 0.017), but not significant for OS when adjusted for stage, histology, surgical approach and adjuvant treatment. There were 7 recurrences in patients with close margins (10.3%, p = 0.25). 71.5% with positive/close margins received adjuvant treatment. In addition, MIS was associated with a higher risk of death (OR = 2.39, p = 0.029). ConclusionSurgical approach was not associated to close or positive margins. Close surgical margins were associated with a higher risk of death. MIS was associated with worse survival, suggesting that margin status may not be the driver of worse survival in these cases.

0016 Influence of Circadian Preference on Sleep, Depression, and Anxiety in Canadian Student-Athletes

Abstract Introduction Student-athletes face unique demands that heighten vulnerability to both sleep and mental health problems. Circadian preference (CP) describes an individuals’ preferred timing for sleep and wake activity, with this conceived as a continuum between the extremes of morningness and eveningness. CP has been shown to influence both sleep and mental health, with greater eveningness preference generally associating with worse outcomes. However, these relationships are understudied in student-athletes. Thus, this investigation was designed to assess the influence of CP on sleep, depression, and anxiety in Canadian student-athletes. Methods Data were obtained throughout the academic semester from Canadian institutions, with student-athletes completing the Athlete Sleep Screening Questionnaire (ASSQ), General Anxiety Disorder-7 (GAD-7), and Patient Health Questionnaire (PHQ-9). ASSQ captured sleep health characteristics and CP, while GAD-7 and PHQ-9 assessed anxiety and depressive symptoms, respectively. ASSQ CP response options were Definitely-Morning, Morning, Evening, or Definitely-Evening as response options, and was coded on a 0-3 scale, with higher values representing increasing eveningness preference. Linear regression estimated associations of CP with sleep, depression, and anxiety. Adjusted regressions controlled for age, gender, race/ethnicity, and sport type. Results Sample included 559 Canadian student-athletes (Average Age = 18.9 ± 3.21 years; Percentage Women = 53%; Percentage White = 79.1%) from eight different sports. Regarding CP, 11.4% reported Definitely-Morning, 27.0% reported Morning, 41.8% reported Evening, and 19.8% reported Definitely-Evening. Greater eveningness preference significantly associated with worse overall sleep difficulties (padjusted &amp;lt; 0.0001), longer sleep latency (padjusted &amp;lt; 0.0001), reduced sleep satisfaction (padjusted &amp;lt; 0.0001), more prolonged sleep inertia (padjusted &amp;lt; 0.0001), more daytime naps (padjusted = 0.01), and more severe depressive symptomatology (padjusted = 0.0006). Conclusion Greater eveningness in a sample of Canadian student-athletes associated with worse sleep and psychological characteristics, including overall sleep difficulty, sleep latency, sleep satisfaction, sleep inertia, daytime napping, and depressive symptom severity. The results of this study provide compelling evidence that eveningness CP may predict poor mental and sleep health outcomes in student-athletes. Thus, further attention to circadian characteristics of student-athletes is warranted. Additionally, modifications to training schedules and team routines could be considered as preventative strategies for mental and sleep health problems. Support (if any) Mitacs #IT23469

Small intestine and colon tissue-resident memory CD8 +T cells exhibit molecular heterogeneity and differential dependence on Eomesodermin

Abstract Tissue-resident memory CD8 +T (T RM) cells are a subset of memory T cells positioned at barrier sites, critical for host defense against infection. Recent studies have begun to reveal substantial heterogeneity in CD8 +T RMpopulations across different tissues. The gut consists of four anatomically distinct compartments, the epithelia and lamina propria in the small intestine (siIEL, siLPL) and colon (cIEL, cLPL). While CD8 +T RMin the small intestine epithelia are well-described, findings are often generalized to the whole gut, thus obscuring potential differences in CD8 +T RMin the much-less studied small intestine lamina propria and colon compartments. We thereby characterized CD8 +T RMin the four intestinal layers using single cell RNA-sequencing and flow cytometry. We show that CD8 +T RMacross the intestinal compartments exhibit differences in the expression of surface molecules, cytokines, granzymes, and transcription factors. For example, small intestine intraepithelial CD8 +T RMproduce the highest levels of granzymes, whereas colon intraepithelial CD8 +T RMhave the greatest capacity for cytokine production. We also uncover an unexpected, tissue-specific role for the T-box transcription factor Eomesodermin (Eomes), known to suppress T RMformation in the skin, liver, and kidney. Our data demonstrate that Eomes is dispensable for CD8 +T RMformation in both the small intestine and colon. By contrast, following CD8 +T RMformation, Eomes plays a critical role in promoting the maintenance of established CD8 +T RMin the small intestine, but not in the colon. Together, our study provides new insights into intestinal CD8 +T RMheterogeneity and differential transcriptional regulation of intestinal T RMformation vs. maintenance. Presenting author Y.H.L. is supported by the Canadian Institutes of Health Research (CIHR) Doctoral Foreign Study Award. This study is supported by grants from NIH (P01 AI132122).