KGKirk Graff

Kirk Graff

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Kirk Graff has a total of 61 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Stability of the functional connectome across childhood: a five-year dense-sampling case study with multi-subject validation

    by Kirk Graff, Yiwen Zhang, Denny Schaedig, Nadine Melhem, Susan B. Perlman

  • Open Access

    Towards precision EEG connectomics: Evaluating the benefits of dense sampling

    by Kirk Graff, Shefali Rai, Shelly Yin, Kate J. Godfrey, Daria Merrikh, Ryann Tansey, Tamara Vanderwal, Andrea B. Protzner, Signe Bray

    EEG connectomics research offers the potential to better understand human neurodevelopment, brain disorders, and brain–behavior associations. Several functional connectivity (FC) measures are widely used but few studies have directly compared metrics of reliability across measures and how data quantity influences these properties. Here, we collected a densely sampled dataset from 25 parent–child pairs, with 80 minutes of passive viewing EEG data collected per participant over 4 sessions, and calculated connectomes using 9 popular phase- and envelope-based measures (coherence, COH; phase locking value, PLV; corrected imaginary phase locking value, CIPLV; imaginary coherence, IMCOH; phase-lag index, PLI; weighted phase-lag index, WPLI; envelope correlation with pairwise orthogonalization, ECPWO; and envelope correlation with symmetric orthogonalization, ECSO), including one effective connectivity measure (phase slope index; PSI). We used connectome individualization, derived from fingerprinting-style analyses, as a multivariate reliability metric, as used in fMRI-FC studies. We used simulations with individual head geometry but no “true” connectivity to assess whether identifiability was influenced by volume conduction. We found that COH and PLV were vulnerable to volume conduction influence on identifiability; CIPLV, ECSO, and ECPWO were semi-vulnerable; and IMCOH, WPLI, PLI, and PSI were minimally vulnerable. Next, we considered individualization and reliability of age-group effects with increasing time of data collection. IMCOH had the overall best performance among minimally vulnerable measures, considering individualization and reliability of group effects. We further found that reliability of IMCOH along with other volume conduction-corrected EEG-FC measures continued to improve with up to 25–30 minutes of data. Together, our findings can support study design decisions in EEG connectomics research.

  • Open Access

    Trends in the Incidence of Human Papillomavirus-Related Cancers in Canada and the Netherlands

    by Emily Niu, Ioana A. Nicolau, Ellen Olthof, Inge M. C. M. de Kok, Kirk Graff, Jennifer Gillis, Corinne M. Doll, Kristin A. Black, Angel Chu, Folkert J. van Kemenade, et al.

    Introduction: Human papillomavirus (HPV) is a well-established cause of cervical cancer and is increasingly recognized as an important risk factor for non-cervical cancers, including oropharyngeal, anal, vulvar, vaginal, and penile cancers. This study assessed temporal trends in the incidence of HPV-related cervical and non-cervical cancers in Canada and the Netherlands by sex assigned at birth and cancer type using population-based registry data. Methods: This study is a registry-based observational study, in which we analyzed retrospective data from the Canadian Cancer Registry (1992-2022) and the Netherlands Cancer Registry (2000-2019) for HPV-related cancers. Age-standardized incidence rates, annual percentage changes (APC), and average annual percentage changes (AAPC), with corresponding 95% confidence intervals (95% CI), were estimated using Joinpoint regression. Results: In Canada, cervical cancer incidence rate generally declined since 1992 but has increased by 1.1% (0.5, 2.7) annually between 2013 and 2022. In the Netherlands, cervical cancer rates remained stable until 2015, followed by a 5.1% (1.9, 11.5) annual increase between 2015 and 2019. Non-cervical HPV-related cancers generally increased in both countries. In Canada, from 1992 to 2022, incidence rates increased for oropharyngeal (AAPC: 2.0% [1.6, 2.4]), anal (1.7% [1.3, 2.2]), and vulvovaginal (1.9% [1.6, 2.3]) cancers. In the Netherlands, from 2000 to 2019, incidence rates increased for anal (5.3% [4.2, 6.8]), vulvar (3.0% [2.2, 4.0]), penile cancers (2.8% [1.7, 4.0]), and oropharyngeal cancer (0.7% [0.1, 1.2]). Conclusion: In general, we observed similar trends in both countries: that the incidence of HPV-related cervical and non-cervical cancers in Canada and the Netherlands is rising. These findings highlight the need to expand HPV vaccination programs, optimize HPV-related cancer screening programs, and enhance public health initiatives for the prevention and early detection of non-cervical HPV-related cancers.


  • Open Access

    Individualized network topography in pre-adolescent children and adults using naturalistic precision fMRI

    by Shefali Rai, Kate J. Godfrey, Kirk Graff, Ryann Tansey, Daria Merrikh, Shelly Yin, Matthew Feigelis, Damion V. Demeter, Tamara Vanderwal, Deanna J. Greene, et al.

    Group-averaged network definitions, commonly used in developmental functional connectivity research, limit our understanding of how network topography may change with age and can lead to inaccurate estimates of age effects on intra- and inter-network functional connectivity. Here, we collected a precision fMRI dataset from 24 parent-child pairs (children 6 to 8 years, 13 females; adults 33 to 47 years, 12 females) during passive video viewing and derived individual template-matched functional network maps from 60 minutes of motion-censored data per participant. Overall, large-scale network architecture was broadly shared between children and adults, with no age-effects observed for network-level surface area, few age differences in network boundaries and generally lower assignment confidence in children. Considering inter-individual similarity in network topography, association networks showed stronger within-family similarity and no overall effect of age on similarity of networks between pairs of individuals. We further asked how individualized network definitions might impact estimates of age effects on dense functional connectivity. While all approaches pointed to greater within-network functional connectivity in adults, we found that individualized approaches had larger age effects and lower sensitivity to head motion. Together, our results suggest that, relative to adults, pre-adolescent children show reduced network assignment confidence and weaker within-network connectivity, but limited differences in network borders and size, and underscore the value of individualized mapping for increasing sensitivity to age effects.


  • Open Access

    Increasing Cervical Cancer Rates Among Women Age 35-54 Years in Canada: Age-Specific Cervical Cancer Incidence Trends in Canada, 1992-2022

    by Ioana A. Nicolau, Matthew T. Warkentin, Kirk Graff, Corinne Doll, Heather Bryant, Darren R. Brenner

    PURPOSE: The vast majority of cervical cancer is preventable through human papillomavirus vaccination and screening with cytology or DNA testing. After decades of progress, recent cervical cancer trends in Western populations show a plateau or modest increase in incidence rates. Further investigation is required to understand the drivers of these emerging trends. In this study, we examined age-specific cervical cancer incidence rates in Canada from 1992 to 2022. METHODS: Data were obtained from the Canadian Cancer Registry maintained by Statistics Canada, which included cancer cases, population counts, and incidence rates of cervical cancer by age and province for the period 1992 to 2022. Joinpoint regression analysis was used to estimate temporal incidence trends across age groups. RESULTS: Cervical cancer incidence rates in Canada decreased among women age 25-34 years and those 65 years and older since 1992. Incidence rates among women age 35-44 years and 45-54 years have increased by 1.1% (95% CI, 0.5 to 2.5) and 1.6% (95% CI, –0.1 to 8.6) per year since 2001 and 2012, respectively. In 2022, the highest incidence rate of cervical cancer was among women age 35-44 years (18.1 per 100,000 women), which is comparable with rates in 1992. CONCLUSION: Cervical cancer incidence rates have been increasing in recent years among women age 35-54 years. This cohort may be falling into a cancer prevention gap. Targeted public health interventions are warranted to address the rising incidence of cervical cancer among this cohort of Canadian women.


  • Open Access

    Minimal Variation in Functional Connectivity in Relation to Daily Affect

    by Kate J. Godfrey, Shefali Rai, Kirk Graff, Shelly Yin, Daria Merrikh, Ryann Tansey, Tamara Vanderwal, Ashley D. Harris, Signe Bray

    Reported associations between functional connectivity and affective disorder symptoms are minimally reproducible, which can partially be attributed to difficulty capturing highly variable clinical symptoms in cross-sectional study designs. “Dense sampling” protocols, where participants are sampled across multiple sessions, can overcome this limitation by studying associations between functional connectivity and variable clinical states. Here, we characterized effect sizes for the association between functional connectivity and time-varying positive and negative daily affect in a nonclinical cohort. Data were analyzed from 24 adults who attended four research visits, where participants self-reported daily affect using the PANAS-X questionnaire and completed 39 min of functional magnetic resonance imaging across three passive viewing conditions. We modeled positive and negative daily affect in relation to network-level functional connectivity, with hypotheses regarding within-network connectivity of the default mode, salience/cingulo-opercular, frontoparietal, dorsal attention, and visual networks and between-network connectivity of affective subcortical regions (amygdala and nucleus accumbens) with both default mode and salience/cingulo-opercular networks. Effect sizes for associations between affect and network-level functional connectivity were small and nonsignificant across analyses. We additionally report that functional connectivity variance is largely attributable to individual identity with small relative variance (<3%) accounted for by within-subject daily affect variation. These results support previous reports that functional connectivity is dominated by stable subject-specific connectivity patterns, while additionally suggesting relatively minimal influence of day-to-day affect. Researchers planning studies examining functional connectivity in relation to daily affect, or other varying stable states, should therefore anticipate small effect sizes and carefully consider power in study planning.

  • Longitudinal Functional Connectome in Pediatric Concussion: An Advancing Concussion Assessment in Pediatrics Study

    by Adrian I. Onicas, Stephanie Deighton, Keith Owen Yeates, Signe Bray, Kirk Graff, Nishard Abdeen, Miriam Helen Beauchamp, Christian Beaulieu, Bruce Hal Bjornson, William Craig, et al.

    Advanced magnetic resonance imaging (MRI) techniques indicate that concussion (i.e., mild traumatic brain injury) disrupts brain structure and function in children. However, the functional connectivity of brain regions within global and local networks (i.e., functional connectome) is poorly understood in pediatric concussion. This prospective, longitudinal study addressed this gap using data from the largest neuroimaging study of pediatric concussion to date to study the functional connectome longitudinally after concussion as compared with mild orthopedic injury (OI). Children and adolescents ( n = 967) 8–16.99 years with concussion or mild OI were recruited from pediatric emergency departments within 48 h post-injury. Pre-injury and 1-month post-injury symptom ratings were used to classify concussion with or without persistent symptoms based on reliable change. Subjects completed a post-acute (2–33 days) and chronic (3 or 6 months via random assignment) MRI scan. Graph theory metrics were derived from 918 resting-state functional MRI scans in 585 children (386 concussion/199 OI). Linear mixed-effects modeling was performed to assess group differences over time, correcting for multiple comparisons. Relative to OI, the global clustering coefficient was reduced at 3 months post-injury in older children with concussion and in females with concussion and persistent symptoms. Time post-injury and sex moderated group differences in local (regional) network metrics of several brain regions, including degree centrality, efficiency, and clustering coefficient of the angular gyrus, calcarine fissure, cuneus, and inferior occipital, lingual, middle occipital, post-central, and superior occipital gyrus. Relative to OI, degree centrality and nodal efficiency were reduced post-acutely, and nodal efficiency and clustering coefficient were reduced chronically after concussion (i.e., at 3 and 6 months post-injury in females; at 6 months post-injury in males). Functional network alterations were more robust and widespread chronically as opposed to post-acutely after concussion, and varied by sex, age, and symptom recovery at 1-month post-injury. Local network segregation reductions emerged globally (across the whole brain network) in older children and in females with poor recovery chronically after concussion. Reduced functioning between neighboring regions could negatively disrupt specialized information processing. Local network metric alterations were demonstrated in several posterior regions that are involved in vision and attention after concussion relative to OI. This indicates that functioning of superior parietal and occipital regions could be particularly susceptibile to the effects of concussion. Moreover, those regional alterations were especially apparent at later time periods post-injury, emerging after post-concussive symptoms resolved in most and persisted up to 6 months post-injury, and differed by biological sex. This indicates that neurobiological changes continue to occur up to 6 months after pediatric concussion, although changes emerge earlier in females than in males. Changes could reflect neural compensation mechanisms.


  • Open Access

    Functional connectomes become more longitudinally self-stable, but not more distinct from others, across early childhood

    by Kirk Graff, Ryann Tansey, Shefali Rai, Amanda Ip, Christiane Rohr, Dennis Dimond, Deborah Dewey, Signe Bray


  • Open Access

    Inattentive and hyperactive traits differentially associate with interindividual functional synchrony during video viewing in young children without ADHD

    by Ryann Tansey, Kirk Graff, Christiane S Rohr, Dennis Dimond, Amanda Ip, Deborah Dewey, Signe Bray

    Inattention and hyperactivity present on a spectrum and may influence the way children perceive and interact with the world. We investigated whether normative variation in inattentive and hyperactive traits was associated with differences in brain function, while children watched clips from an age-appropriate television program. Functional magnetic resonance imaging (fMRI) data and parent reports of inattention and hyperactivity traits were collected from 81 children 4–7 years of age with no parent-reported diagnoses. Data were analyzed using intersubject correlations (ISCs) in mixed effects models to determine if inattentive and hyperactive traits were associated with idiosyncrasy of fMRI response to the video. We hypothesized that pairs of children with higher average inattention and hyperactivity scores would show less interindividual brain synchrony to one another than pairs with lower average scores on these traits. Video watching engaged widespread visual, auditory, default mode and dorsal prefrontal regions. Inattention and hyperactivity were separably associated with ISC in many of these regions. Our findings suggest that the spectrum of inattention and hyperactivity traits in children without ADHD are differentially associated with neural processing of naturalistic video stimuli, which may have implications for understanding how children with different levels of these traits process audiovisual information in unconstrained conditions.


  • Open Access

    Benchmarking common preprocessing strategies in early childhood functional connectivity and intersubject correlation fMRI

    by Kirk Graff, Ryann Tansey, Amanda Ip, Christiane Rohr, Dennis Dimond, Deborah Dewey, Signe Bray