RESUMO
INTRODUCTION: Cleft lip and/or palate (CL/P) affects 1 in 700 live births globally. Children born with CL/P and their families face various challenges throughout the child's development. Extant research is often limited by small numbers and single-centre data. The Cleft Collective, a national cohort study in the UK, aims to build a resource, available to collaborators across the globe, to understand causes, best treatments and long-term outcomes for those born with CL/P, ultimately seeking to enhance their quality of life through improved understanding and care. METHODS AND ANALYSIS: A longitudinal prospective cohort study of children born with CL/P and their families. Recruitment occurs across the UK and started in November 2013. Recruitment will continue until September 2027 with an estimated final sample of 4822 children born with CL/P (1157 cleft lip including/excluding the alveolus; 2112 cleft palate only; 1042 unilateral cleft lip and palate and 511 bilateral cleft lip and palate). Biological samples are collected from all recruited members of the family. Parental and child questionnaires are collected at key time points throughout the child's development. Surgical data are collected at the time of surgical repair of the child's cleft. Consent is obtained to link to external data sources. Nested substudies can be hosted within the cohort. Regular engagement with participants takes place through birthday cards for the children, social media posts and newsletters. Patient and Public Involvement is conducted through the Cleft Lip And Palate Association and Cleft Collective Patient Consultation Group who provide insightful and essential guidance to the Cleft Collective throughout planning and conducting research. ETHICS AND DISSEMINATION: The Cleft Collective was ethically approved by the National Research Ethics Service committee South West-Central Bristol (REC13/SW/0064). Parental informed consent is required for participation. Findings from the Cleft Collective are disseminated through peer-reviewed publications, conference presentations, newsletters and social media.
Assuntos
Fenda Labial , Fissura Palatina , Humanos , Fenda Labial/cirurgia , Fenda Labial/epidemiologia , Fissura Palatina/cirurgia , Fissura Palatina/epidemiologia , Estudos Prospectivos , Estudos Longitudinais , Reino Unido , Criança , Lactente , Qualidade de Vida , Pré-Escolar , Feminino , Masculino , Projetos de Pesquisa , Inquéritos e Questionários , Pais/psicologiaRESUMO
RATIONALE AND OBJECTIVES: Integrating learning spacing in medicine has shown promise in enhancing knowledge retention and diagnostic proficiency. While studies demonstrate the effectiveness of spaced learning in various fields, limited research exists on its application in radiological training. This study aims to investigate the impact of intervals in spaced training on radiologists' and trainees' diagnostic performance via mammogram test sets. METHODS: 54 radiologists and 101 radiology trainees completed 207 and 458 first-time readings of 9 mammogram test sets between 2019 and 2023. Each test set comprised of 60 mammograms (20 cancer and 40 normal), sourced retrospectively from BreastScreen Australia. Each radiologist evaluated mammograms using the BIRADS lexicon. Readers' performance was compared with truth data and evaluated in terms of specificity, case sensitivity, lesion sensitivity, ROC AUC and JAFROC FOM. The progress of readers' performances in following test sets after the first one was analyzed using Wilcoxon Signed Rank test. The association of participants' performances and the intervals among test sets' completions was investigated using Pearson's test. RESULTS: Significant positive correlations were found between intervals and radiologists' improvement in specificity and JAFROC FOM (P < 0.05). The separation of 4 to 10 days showed the most improvement among radiologists across all metrics, while intervals exceeding 90 days related to highest increase in case sensitivity (5.15%), lesion sensitivity (6.55%), ROC AUC (3.05%) and JAFROC FOM (6.3%). Trainees completing test sets in one day showed positive correlations with their ROC AUC (R=0.45; P = 0.008) and JAFROC FOM (R=0.43; P = 0.02), while those taking a longer time to complete showed negative impacts on case sensitivity (P = 0.009) and ROC AUC (P = 0.02). Remarkable progress in trainees was found in case sensitivity (6.15%), lesion sensitivity (11.6%), ROC AUC (3.5%) and JAFROC FOM (4.35%) with test set intervals of 31-90 days. CONCLUSIONS: Radiologists demonstrated superior performance when the training test sets were spaced over longer intervals, whereas trainees exhibited proficiency with shorter time separations. By optimizing the spacing of reviewing and practicing radiological concepts, mammogram readers can bolster memory retention and diagnostic decision-making skills.
RESUMO
Objectives This study explored the familiarity, perceptions and confidence of Australian radiology clinicians involved in reading screening mammograms, regarding artificial intelligence (AI) applications in breast cancer detection. Methods Sixty-five radiologists, breast physicians and radiology trainees participated in an online survey that consisted of 23 multiple choice questions asking about their experience and familiarity with AI products. Furthermore, the survey asked about their confidence in using AI outputs and their preference for AI modes applied in a breast screening context. Participants' responses to questions were compared using Pearson's χ 2 test. Bonferroni-adjusted significance tests were used for pairwise comparisons. Results Fifty-five percent of respondents had experience with AI in their workplaces, with automatic density measurement powered by machine learning being the most familiar AI product (69.4%). The top AI outputs with the highest ranks of perceived confidence were 'Displaying suspicious areas on mammograms with the percentage of cancer possibility' (67.8%) and 'Automatic mammogram classification (normal, benign, cancer, uncertain)' (64.6%). Radiology and breast physicians preferred using AI as second-reader mode (75.4% saying 'somewhat happy' to 'extremely happy') over triage (47.7%), pre-screening and first-reader modes (both with 26.2%) (P < 0.001). Conclusion The majority of screen readers expressed increased confidence in utilising AI for highlighting suspicious areas on mammograms and for automatically classifying mammograms. They considered AI as an optimal second-reader mode being the most ideal use in a screening program. The findings provide valuable insights into the familiarities and expectations of radiologists and breast clinicians for the AI products that can enhance the effectiveness of the breast cancer screening programs, benefitting both healthcare professionals and patients alike.
Assuntos
Inteligência Artificial , Neoplasias da Mama , Detecção Precoce de Câncer , Mamografia , Adulto , Feminino , Humanos , Pessoa de Meia-Idade , Austrália , Neoplasias da Mama/diagnóstico , Neoplasias da Mama/psicologia , Detecção Precoce de Câncer/métodos , Detecção Precoce de Câncer/psicologia , Mamografia/métodos , Radiologistas/psicologia , Inquéritos e QuestionáriosRESUMO
Although the value of adding AI as a surrogate second reader in various scenarios has been investigated, it is unknown whether implementing an AI tool within double reading practice would capture additional subtle cancers missed by both radiologists who independently assessed the mammograms. This paper assesses the effectiveness of two state-of-the-art Artificial Intelligence (AI) models in detecting retrospectively-identified missed cancers within a screening program employing double reading practices. The study also explores the agreement between AI and radiologists in locating the lesions, considering various levels of concordance among the radiologists in locating the lesions. The Globally-aware Multiple Instance Classifier (GMIC) and Global-Local Activation Maps (GLAM) models were fine-tuned for our dataset. We evaluated the sensitivity of both models on missed cancers retrospectively identified by a panel of three radiologists who reviewed prior examinations of 729 cancer cases detected in a screening program with double reading practice. Two of these experts annotated the lesions, and based on their concordance levels, cases were categorized as 'almost perfect,' 'substantial,' 'moderate,' and 'poor.' We employed Similarity or Histogram Intersection (SIM) and Kullback-Leibler Divergence (KLD) metrics to compare saliency maps of malignant cases from the AI model with annotations from radiologists in each category. In total, 24.82% of cancers were labeled as "missed." The performance of GMIC and GLAM on the missed cancer cases was 82.98% and 79.79%, respectively, while for the true screen-detected cancers, the performances were 89.54% and 87.25%, respectively (p-values for the difference in sensitivity < 0.05). As anticipated, SIM and KLD from saliency maps were best in 'almost perfect,' followed by 'substantial,' 'moderate,' and 'poor.' Both GMIC and GLAM (p-values < 0.05) exhibited greater sensitivity at higher concordance. Even in a screening program with independent double reading, adding AI could potentially identify missed cancers. However, the challenging-to-locate lesions for radiologists impose a similar challenge for AI.
Assuntos
Inteligência Artificial , Neoplasias da Mama , Detecção Precoce de Câncer , Mamografia , Humanos , Mamografia/métodos , Feminino , Neoplasias da Mama/diagnóstico por imagem , Neoplasias da Mama/diagnóstico , Estudos Retrospectivos , Detecção Precoce de Câncer/métodos , Pessoa de Meia-Idade , Idoso , Interpretação de Imagem Radiográfica Assistida por Computador/métodos , Sensibilidade e EspecificidadeRESUMO
BACKGROUND: Colorectal cancer (CRC) is the third-most-common cancer worldwide and its rates are increasing. Elevated body mass index (BMI) is an established risk factor for CRC, although the molecular mechanisms behind this association remain unclear. Using the Mendelian randomization (MR) framework, we aimed to investigate the mediating effects of putative biomarkers and other CRC risk factors in the association between BMI and CRC. METHODS: We selected as mediators biomarkers of established cancer-related mechanisms and other CRC risk factors for which a plausible association with obesity exists, such as inflammatory biomarkers, glucose homeostasis traits, lipids, adipokines, insulin-like growth factor 1 (IGF1), sex hormones, 25-hydroxy-vitamin D, smoking, physical activity (PA) and alcohol consumption. We used inverse-variance weighted MR in the main univariable analyses and performed sensitivity analyses (weighted-median, MR-Egger, Contamination Mixture). We used multivariable MR for the mediation analyses. RESULTS: Genetically predicted BMI was positively associated with CRC risk [odds ratio per SD (5 kg/m2) = 1.17, 95% CI: 1.08-1.24, P-value = 1.4 × 10-5] and robustly associated with nearly all potential mediators. Genetically predicted IGF1, fasting insulin, low-density lipoprotein cholesterol, smoking, PA and alcohol were associated with CRC risk. Evidence for attenuation was found for IGF1 [explained 7% (95% CI: 2-13%) of the association], smoking (31%, 4-57%) and PA (7%, 2-11%). There was little evidence for pleiotropy, although smoking was bidirectionally associated with BMI and instruments were weak for PA. CONCLUSIONS: The effect of BMI on CRC risk is possibly partly mediated through plasma IGF1, whereas the attenuation of the BMI-CRC association by smoking and PA may reflect confounding and shared underlying mechanisms rather than mediation.
Assuntos
Índice de Massa Corporal , Neoplasias Colorretais , Análise da Randomização Mendeliana , Obesidade , Humanos , Neoplasias Colorretais/genética , Neoplasias Colorretais/epidemiologia , Fatores de Risco , Obesidade/genética , Obesidade/epidemiologia , Fator de Crescimento Insulin-Like I/metabolismo , Consumo de Bebidas Alcoólicas/epidemiologiaRESUMO
Based on the World Cancer Research Fund Global Cancer Update Programme, we performed systematic reviews and meta-analyses to investigate the association of post-diagnosis adiposity, physical activity, sedentary behaviour, and dietary factors with colorectal cancer prognosis. We searched PubMed and Embase until 28th February, 2022. An independent expert committee and expert panel graded the quality of evidence. A total of 167 unique publications were reviewed, and all but five were observational studies. The quality of the evidence was graded conservatively due to the high risk of several biases. There was evidence of non-linearity in the associations between body mass index and colorectal cancer prognosis. The associations appeared reverse J-shaped, and the quality of this evidence was graded as limited (likelihood of causality: limited-no conclusion). The evidence on recreational physical activity and lower risk of all-cause mortality (relative risk [RR] highest vs. lowest: 0.69, 95% confidence interval [CI]: 0.62-0.77) and recurrence/disease-free survival (RR: 0.80, 95% CI: 0.70-0.92) was graded as limited-suggestive. There was limited-suggestive evidence for the associations between healthy dietary and/or lifestyle patterns (including diets that comprised plant-based foods), intake of whole grains and coffee with lower risk of all-cause mortality, and between unhealthy dietary patterns and intake of sugary drinks with higher risk of all-cause mortality. The evidence for other exposures on colorectal cancer outcomes was sparse and graded as limited-no conclusion. Analyses were conducted excluding cancer patients with metastases without substantial changes in the findings. Well-designed intervention and cohort studies are needed to support the development of lifestyle recommendations for colorectal cancer patients.
Assuntos
Adiposidade , Neoplasias Colorretais , Dieta , Exercício Físico , Comportamento Sedentário , Humanos , Prognóstico , Suplementos Nutricionais , Fatores de RiscoRESUMO
Low physical activity and high sedentary behaviour have been clearly linked with colorectal cancer development, yet data on their potential role in colorectal cancer survival is limited. Better characterisation of these relationships is needed for the development of post-diagnosis physical activity and sedentary behaviour guidance for colorectal cancer survivors. We searched PubMed and Embase through 28 February 2022 for studies assessing post-diagnosis physical activity, and/or sedentary behaviour in relation to all-cause and cause-specific mortality and recurrence after colorectal cancer diagnosis. Total and recreational physical activity were assessed overall and by frequency, duration, intensity, and volume using categorical, linear, and non-linear dose-response random-effects meta-analyses. The Global Cancer Update Programme (CUP Global) independent Expert Committee on Cancer Survivorship and Expert Panel interpreted and graded the likelihood of causality. We identified 16 observational studies on 82,220 non-overlapping patients from six countries. Physical activity was consistently inversely associated with colorectal cancer morbidity and mortality outcomes, with 13%-60% estimated reductions in risk. Sedentary behaviour was positively associated with all-cause mortality. The evidence had methodological limitations including potential confounding, selection bias and reverse causation, coupled with a limited number of studies for most associations. The CUP Global Expert panel concluded limited-suggestive evidence for recreational physical activity with all-cause mortality and cancer recurrence. Total physical activity and its specific domains and dimensions, and sedentary behaviour were all graded as limited-no conclusion for all outcomes. Future research should focus on randomised trials, while observational studies should obtain objective and repeated physical activity measures and better adjustment for confounders.
Assuntos
Neoplasias Colorretais , Exercício Físico , Comportamento Sedentário , Humanos , Neoplasias Colorretais/mortalidade , Neoplasias Colorretais/diagnóstico , Prognóstico , Estudos Observacionais como AssuntoRESUMO
The role of diet in colorectal cancer prognosis is not well understood and specific lifestyle recommendations are lacking. We searched for randomised controlled trials (RCTs) and longitudinal observational studies on post-diagnosis dietary factors, supplement use and colorectal cancer survival outcomes in PubMed and Embase from inception until 28th February 2022. Random-effects dose-response meta-analyses were conducted when at least three studies had sufficient information. The evidence was interpreted and graded by the CUP Global independent Expert Committee on Cancer Survivorship and Expert Panel. Five RCTs and 35 observational studies were included (30,242 cases, over 8700 all-cause and 2100 colorectal cancer deaths, 3700 progression, recurrence, or disease-free events). Meta-analyses, including 3-10 observational studies each, were conducted for: whole grains, nuts/peanuts, red and processed meat, dairy products, sugary drinks, artificially sweetened beverages, coffee, alcohol, dietary glycaemic load/index, insulin load/index, marine omega-3 polyunsaturated fatty acids, supplemental calcium, circulating 25-hydroxyvitamin D (25[OH]D) and all-cause mortality; for alcohol, supplemental calcium, circulating 25(OH)D and colorectal cancer-specific mortality; and for circulating 25(OH)D and recurrence/disease-free survival. The overall evidence was graded as 'limited'. The inverse associations between healthy dietary and/or lifestyle patterns (including diets that comprised plant-based foods), whole grains, total, caffeinated, or decaffeinated coffee and all-cause mortality and the positive associations between unhealthy dietary patterns, sugary drinks and all-cause mortality provided 'limited-suggestive' evidence. All other exposure-outcome associations provided 'limited-no conclusion' evidence. Additional, well-conducted cohort studies and carefully designed RCTs are needed to develop specific lifestyle recommendations for colorectal cancer survivors.
Assuntos
Neoplasias Colorretais , Suplementos Nutricionais , Humanos , Neoplasias Colorretais/mortalidade , Neoplasias Colorretais/epidemiologia , Prognóstico , Dieta , Vitamina D/administração & dosagem , Vitamina D/análogos & derivados , Ensaios Clínicos Controlados Aleatórios como Assunto , Estudos Observacionais como AssuntoRESUMO
The adiposity influence on colorectal cancer prognosis remains poorly characterised. We performed a systematic review and meta-analysis on post-diagnosis adiposity measures (body mass index [BMI], waist circumference, waist-to-hip ratio, weight) or their changes and colorectal cancer outcomes. PubMed and Embase were searched through 28 February 2022. Random-effects meta-analyses were conducted when at least three studies had sufficient information. The quality of evidence was interpreted and graded by the Global Cancer Update Programme (CUP Global) independent Expert Committee on Cancer Survivorship and Expert Panel. We reviewed 124 observational studies (85 publications). Meta-analyses were possible for BMI and all-cause mortality, colorectal cancer-specific mortality, and cancer recurrence/disease-free survival. Non-linear meta-analysis indicated a reverse J-shaped association between BMI and colorectal cancer outcomes (nadir at BMI 28 kg/m2). The highest risk, relative to the nadir, was observed at both ends of the BMI distribution (18 and 38 kg/m2), namely 60% and 23% higher risk for all-cause mortality; 95% and 26% for colorectal cancer-specific mortality; and 37% and 24% for cancer recurrence/disease-free survival, respectively. The higher risk with low BMI was attenuated in secondary analyses of RCTs (compared to cohort studies), among studies with longer follow-up, and in women suggesting potential methodological limitations and/or altered physiological state. Descriptively synthesised studies on other adiposity-outcome associations of interest were limited in number and methodological quality. All the associations were graded as limited (likelihood of causality: no conclusion) due to potential methodological limitations (reverse causation, confounding, selection bias). Additional well-designed observational studies and interventional trials are needed to provide further clarification.
Assuntos
Adiposidade , Índice de Massa Corporal , Neoplasias Colorretais , Humanos , Neoplasias Colorretais/mortalidade , Neoplasias Colorretais/diagnóstico , Prognóstico , Circunferência da Cintura , Relação Cintura-Quadril , Feminino , Obesidade/complicaçõesRESUMO
There is evidence that tissue concentrations of mercury (Hg) and selenium (Se) are predicted by numerous dietary, sociodemographic, environmental, and genetic factors. This study aimed to estimate the relative importance of predictors of Hg and Se concentrations in blood samples taken from pregnant women. The Avon Longitudinal Study of Parents and Children (ALSPAC) in the UK measured whole blood Hg and Se concentrations in 3,972 pregnant women. We identified 30 potential predictors of Hg and 24 of Se, which were evaluated using cross-validated random forests to identify the optimal models for predictive power. The relative importance of individual variables was estimated by averaging the added-R2 per predictor. Linkage disequilibrium score regression was used to estimate the variance explained by genotype. A multivariable model of 14 predictors explained 22.4% of Hg variance (95% CI: 13.0 to 37.1), including 6.9% from blood Se and 3.2% from white fish consumption. There were 11 predictors which explained 15.3% of Se variance (CI: 8.9 to 25.9), including 6.4% from blood Hg, 1.3% from blood lead, and 1.3% from oily fish. Measured genetic variation explained 30% of Hg variance (CI: 8.4 to 51.5) and 37.5% of Se (CI: 10.4 to 64.5). A high proportion of Hg and Se variance could be explained from dietary, sociodemographic, metabolic, and genetic factors. Seafood consumption was less predictive of Hg than may be expected and other factors should be considered when determining risk of exposure. There was tentative evidence that genotype is a major contributor to Hg and Se variation, possibly by modifying the efficacy of internal metabolism.
RESUMO
BACKGROUND: Tumour-promoting inflammation is a "hallmark" of cancer and conventional epidemiological studies have reported links between various inflammatory markers and cancer risk. The causal nature of these relationships and, thus, the suitability of these markers as intervention targets for cancer prevention is unclear. METHODS: We meta-analysed 6 genome-wide association studies of circulating inflammatory markers comprising 59,969 participants of European ancestry. We then used combined cis-Mendelian randomization and colocalisation analysis to evaluate the causal role of 66 circulating inflammatory markers in risk of 30 adult cancers in 338,294 cancer cases and up to 1,238,345 controls. Genetic instruments for inflammatory markers were constructed using genome-wide significant (P < 5.0 × 10-8) cis-acting SNPs (i.e., in or ±250 kb from the gene encoding the relevant protein) in weak linkage disequilibrium (LD, r2 < 0.10). Effect estimates were generated using inverse-variance weighted random-effects models and standard errors were inflated to account for weak LD between variants with reference to the 1000 Genomes Phase 3 CEU panel. A false discovery rate (FDR)-corrected P-value ("q-value") <0.05 was used as a threshold to define "strong evidence" to support associations and 0.05 ≤ q-value < 0.20 to define "suggestive evidence". A colocalisation posterior probability (PPH4) >70% was employed to indicate support for shared causal variants across inflammatory markers and cancer outcomes. Findings were replicated in the FinnGen study and then pooled using meta-analysis. FINDINGS: We found strong evidence to support an association of genetically-proxied circulating pro-adrenomedullin concentrations with increased breast cancer risk (OR: 1.19, 95% CI: 1.10-1.29, q-value = 0.033, PPH4 = 84.3%) and suggestive evidence to support associations of interleukin-23 receptor concentrations with increased pancreatic cancer risk (OR: 1.42, 95% CI: 1.20-1.69, q-value = 0.055, PPH4 = 73.9%), prothrombin concentrations with decreased basal cell carcinoma risk (OR: 0.66, 95% CI: 0.53-0.81, q-value = 0.067, PPH4 = 81.8%), and interleukin-1 receptor-like 1 concentrations with decreased triple-negative breast cancer risk (OR: 0.92, 95% CI: 0.88-0.97, q-value = 0.15, PPH4 = 85.6%). These findings were replicated in pooled analyses with the FinnGen study. Though suggestive evidence was found to support an association of macrophage migration inhibitory factor concentrations with increased bladder cancer risk (OR: 2.46, 95% CI: 1.48-4.10, q-value = 0.072, PPH4 = 76.1%), this finding was not replicated when pooled with the FinnGen study. For 22 of 30 cancer outcomes examined, there was little evidence (q-value ≥0.20) that any of the 66 circulating inflammatory markers examined were associated with cancer risk. INTERPRETATION: Our comprehensive joint Mendelian randomization and colocalisation analysis of the role of circulating inflammatory markers in cancer risk identified potential roles for 4 circulating inflammatory markers in risk of 4 site-specific cancers. Contrary to reports from some prior conventional epidemiological studies, we found little evidence of association of circulating inflammatory markers with the majority of site-specific cancers evaluated. FUNDING: Cancer Research UK (C68933/A28534, C18281/A29019, PPRCPJT∖100005), World Cancer Research Fund (IIG_FULL_2020_022), National Institute for Health Research (NIHR202411, BRC-1215-20011), Medical Research Council (MC_UU_00011/1, MC_UU_00011/3, MC_UU_00011/6, and MC_UU_00011/4), Academy of Finland Project 326291, European Union's Horizon 2020 grant agreement no. 848158 (EarlyCause), French National Cancer Institute (INCa SHSESP20, 2020-076), Versus Arthritis (21173, 21754, 21755), National Institutes of Health (U19 CA203654), National Cancer Institute (U19CA203654).
Assuntos
Estudo de Associação Genômica Ampla , Neoplasias , Adulto , Humanos , Análise da Randomização Mendeliana , Risco , Neoplasias/epidemiologia , Neoplasias/genética , Inflamação/genética , Polimorfismo de Nucleotídeo ÚnicoRESUMO
OBJECTIVE: Radiologists can detect the gist of abnormal based on their rapid initial impression on a mammogram (ie, global gist signal [GGS]). This study explores (1) whether global radiomic (ie, computer-extracted) features can predict the GGS; and if so, (ii) what features are the most important drivers of the signals. METHODS: The GGS of cases in two extreme conditions was considered: when observers detect a very strong gist (high-gist) and when the gist of abnormal was not/poorly perceived (low-gist). Gist signals/scores from 13 observers reading 4191 craniocaudal mammograms were collected. As gist is a noisy signal, the gist scores from all observers were averaged and assigned to each image. The high-gist and low-gist categories contained all images in the fourth and first quartiles, respectively. One hundred thirty handcrafted global radiomic features (GRFs) per mammogram were extracted and utilized to construct eight separate machine learning random forest classifiers (All, Normal, Cancer, Prior-1, Prior-2, Missed, Prior-Visible, and Prior-Invisible) for characterizing high-gist from low-gist images. The models were trained and validated using the 10-fold cross-validation approach. The models' performances were evaluated by the area under receiver operating characteristic curve (AUC). Important features for each model were identified through a scree test. RESULTS: The Prior-Visible model achieved the highest AUC of 0.84 followed by the Prior-Invisible (0.83), Normal (0.82), Prior-1 (0.81), All (0.79), Prior-2 (0.77), Missed (0.75), and Cancer model (0.69). Cluster shade, standard deviation, skewness, kurtosis, and range were identified to be the most important features. CONCLUSIONS: Our findings suggest that GRFs can accurately classify high- from low-gist images. ADVANCES IN KNOWLEDGE: Global mammographic radiomic features can accurately predict high- from low-gist images with five features identified to be valuable in describing high-gist images. These are critical in providing better understanding of the mammographic image characteristics that drive the strength of the GGSs which could be exploited to advance breast cancer (BC) screening and risk prediction, enabling early detection and treatment of BC thereby further reducing BC-related deaths.
Assuntos
Neoplasias da Mama , Radiômica , Humanos , Feminino , Mamografia , Computadores , RadiologistasRESUMO
This paper investigates the adaptability of four state-of-the-art artificial intelligence (AI) models to the Australian mammographic context through transfer learning, explores the impact of image enhancement on model performance and analyses the relationship between AI outputs and histopathological features for clinical relevance and accuracy assessment. A total of 1712 screening mammograms (n = 856 cancer cases and n = 856 matched normal cases) were used in this study. The 856 cases with cancer lesions were annotated by two expert radiologists and the level of concordance between their annotations was used to establish two sets: a 'high-concordances subset' with 99% agreement of cancer location and an 'entire dataset' with all cases included. The area under the receiver operating characteristic curve (AUC) was used to evaluate the performance of Globally aware Multiple Instance Classifier (GMIC), Global-Local Activation Maps (GLAM), I&H and End2End AI models, both in the pretrained and transfer learning modes, with and without applying the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm. The four AI models with and without transfer learning in the high-concordance subset outperformed those in the entire dataset. Applying the CLAHE algorithm to mammograms improved the performance of the AI models. In the high-concordance subset with the transfer learning and CLAHE algorithm applied, the AUC of the GMIC model was highest (0.912), followed by the GLAM model (0.909), I&H (0.893) and End2End (0.875). There were significant differences (p < 0.05) in the performances of the four AI models between the high-concordance subset and the entire dataset. The AI models demonstrated significant differences in malignancy probability concerning different tumour size categories in mammograms. The performance of AI models was affected by several factors such as concordance classification, image enhancement and transfer learning. Mammograms with a strong concordance with radiologists' annotations, applying image enhancement and transfer learning could enhance the accuracy of AI models.
RESUMO
OBJECTIVES: To determine the UK prevalence of behavioral problems in 5-year-old children born with isolated or syndromic cleft lip and/or palate (CL/P) compared to the general population and identify potentially associated factors. DESIGN: Observational study using questionnaire data from the Cleft Collective 5-Year-Old Cohort study and three general population samples. MAIN OUTCOME MEASURE: The Strengths and Difficulties Questionnaire (SDQ). PARTICIPANTS: Mothers of children (age: 4.9-6.8 years) born with CL/P (n = 325). UK general population cohorts for SDQ scores were: Millennium Cohort Study (MCS) (n = 12 511), Office of National Statistics (ONS) normative school-age SDQ data (n = 5855), and Avon Longitudinal Study of Parents and Children (ALSPAC) (n = 9386). RESULTS: By maternal report, 14.2% of children born with CL/P were above clinical cut-off for behavioral problems, which was more likely than in general population samples: 7.5% of MCS (OR = 2.05 [1.49-2.82], P < 0.001), 9.8% of ONS (OR = 1.52 [1.10-2.09], P = 0.008), and 6.6% of ALSPAC (OR = 2.34 [1.70-3.24], P < 0.001). Children in the Cleft Collective had higher odds for hyperactivity, emotional and peer problems, and less prosocial behaviors. Maternal stress, lower maternal health-related quality of life and family functioning, receiving government income support, and maternal smoking showed evidence of association (OR range: 4.41-10.13) with behavioral problems, along with maternal relationship status, younger age, and lower education (OR range: 2.34-3.73). CONCLUSIONS: Findings suggest elevated levels of behavioral problems in children born with CL/P compared to the general population with several associated maternal factors similar to the general population.
Assuntos
Fenda Labial , Fissura Palatina , Comportamento Problema , Criança , Pré-Escolar , Humanos , Fenda Labial/epidemiologia , Fenda Labial/psicologia , Fissura Palatina/epidemiologia , Fissura Palatina/psicologia , Estudos de Coortes , Estudos Longitudinais , Prevalência , Qualidade de VidaRESUMO
BACKGROUND: Sedentary behaviours have been associated with increased risks of some common cancers in epidemiological studies; however, it is unclear if these associations are causal. METHODS: We used univariable and multivariable two-sample Mendelian randomisation (MR) to examine potential causal relationships between sedentary behaviours and risks of breast, colorectal and prostate cancer. Genetic variants associated with self-reported leisure television watching and computer use were identified from a recent genome-wide association study (GWAS). Data related to cancer risk were obtained from cancer GWAS consortia. A series of sensitivity analyses were applied to examine the robustness of the results to the presence of confounding. RESULTS: A 1-standard deviation (SD: 1.5 h/day) increment in hours of television watching increased risk of breast cancer (OR per 1-SD: 1.15, 95% confidence interval [CI]: 1.05-1.26) and colorectal cancer (OR per 1-SD: 1.32, 95% CI: 1.16-1.49) while there was little evidence of an association for prostate cancer risk (OR per 1-SD: 0.94, 95% CI: 0.84-1.06). After adjusting for years of education, the effect estimates for television watching were attenuated (breast cancer, OR per 1-SD: 1.08, 95% CI: 0.92-1.27; colorectal cancer, OR per 1-SD: 1.08, 95% CI: 0.90-1.31). Post hoc analyses showed that years of education might have a possible confounding and mediating role in the association between television watching with breast and colorectal cancer. Consistent results were observed for each cancer site according to sex (colorectal cancer), anatomical subsites and cancer subtypes. There was little evidence of associations between genetically predicted computer use and cancer risk. CONCLUSIONS: Our univariable analysis identified some positive associations between hours of television watching and risks of breast and colorectal cancer. However, further adjustment for additional lifestyle factors especially years of education attenuated these results. Future studies using objective measures of exposure can provide new insights into the possible role of sedentary behaviour in cancer development.
Assuntos
Neoplasias da Mama , Neoplasias Colorretais , Análise da Randomização Mendeliana , Neoplasias da Próstata , Comportamento Sedentário , Televisão , Humanos , Masculino , Neoplasias da Próstata/genética , Neoplasias da Próstata/epidemiologia , Neoplasias da Próstata/etiologia , Televisão/estatística & dados numéricos , Neoplasias da Mama/genética , Neoplasias da Mama/epidemiologia , Neoplasias da Mama/etiologia , Feminino , Neoplasias Colorretais/genética , Neoplasias Colorretais/epidemiologia , Neoplasias Colorretais/etiologia , Computadores/estatística & dados numéricos , Estudo de Associação Genômica Ampla , Atividades de Lazer , Fatores de RiscoRESUMO
Mercury has high industrial utility and is present in many products, and environmental contamination and occupational exposure are widespread. There are numerous biological systems involved in the absorption, metabolism, and excretion of Hg, and it is possible that some systems may be impacted by genetic variation. If so, genotype may affect tissue concentrations of Hg and subsequent toxic effects. Genome-wide association testing was performed on blood Hg samples from pregnant women of the Avon Longitudinal Study of Parents and Children (n = 2893) and children of the Human Early Life Exposome (n = 1042). Directly-genotyped single-nucleotide polymorphisms (SNPs) were imputed to the Haplotype Reference Consortium r1.1 panel of whole genotypes and modelled againstlog-transformed Hg. Heritability was estimated using linkage disequilibrium score regression. The heritability of Hg was estimated as 24.0% (95% CI: 16.9% to 46.4%) in pregnant women, but could not be determined in children. There were 16 SNPs associated with Hg in pregnant women above a suggestive p-value threshold (p < 1 × 10-5), and 21 for children. However, no SNP passed this threshold in both studies, and none were genome-wide significant (p < 5 × 10-8). SNP-Hg associations were highly discordant between women and children, and this may reflect differences in metabolism, a gene-age interaction, or dose-response effects. Several suggestive variants had plausible links to Hg metabolism, such as rs146099921 in metal transporter SLC39A14, and two variants (rs28618224, rs7154700) in potassium voltage-gated channel genes. The findings would benefit from external validation, as suggestive results may contain both true associations and false positives.