scholarly journals Reducing the number of unnecessary biopsies of US-BI-RADS 4a lesions through a deep learning method for residents-in-training: a cross-sectional study

BMJ Open ◽  
2020 ◽  
Vol 10 (6) ◽  
pp. e035757
Author(s):  
Chenyang Zhao ◽  
Mengsu Xiao ◽  
He Liu ◽  
Ming Wang ◽  
Hongyan Wang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Breast Imaging ◽  
Diagnostic Performance ◽  
Intraclass Correlation ◽  
Cross Sectional Study ◽  
Computer Assisted ◽  
Sectional Study ◽  
Benign Lesions ◽  
Cross Sectional ◽  
Experienced Radiologist

ObjectiveThe aim of the study is to explore the potential value of S-Detect for residents-in-training, a computer-assisted diagnosis system based on deep learning (DL) algorithm.MethodsThe study was designed as a cross-sectional study. Routine breast ultrasound examinations were conducted by an experienced radiologist. The ultrasonic images of the lesions were retrospectively assessed by five residents-in-training according to the Breast Imaging Report and Data System (BI-RADS) lexicon, and a dichotomic classification of the lesions was provided by S-Detect. The diagnostic performances of S-Detect and the five residents were measured and compared using the pathological results as the gold standard. The category 4a lesions assessed by the residents were downgraded to possibly benign as classified by S-Detect. The diagnostic performance of the integrated results was compared with the original results of the residents.ParticipantsA total of 195 focal breast lesions were consecutively enrolled, including 82 malignant lesions and 113 benign lesions.ResultsS-Detect presented higher specificity (77.88%) and area under the curve (AUC) (0.82) than the residents (specificity: 19.47%–48.67%, AUC: 0.62–0.74). A total of 24, 31, 38, 32 and 42 identified as BI-RADS 4a lesions by residents 1, 2, 3, 4 and 5 were downgraded to possibly benign lesions by S-Detect, respectively. Among these downgraded lesions, 24, 28, 35, 30 and 40 lesions were proven to be pathologically benign, respectively. After combining the residents' results with the results of the software in category 4a lesions, the specificity and AUC of the five residents significantly improved (specificity: 46.02%–76.11%, AUC: 0.71–0.85, p<0.001). The intraclass correlation coefficient of the five residents also increased after integration (from 0.480 to 0.643).ConclusionsWith the help of the DL software, the specificity, overall diagnostic performance and interobserver agreement of the residents greatly improved. The software can be used as adjunctive tool for residents-in-training, downgrading 4a lesions to possibly benign and reducing unnecessary biopsies.

scholarly journals Understanding neighbourhood retail food environmental mechanisms influencing BMI in the Caribbean: a multilevel analysis from the Jamaica Health and Lifestyle Survey: a cross-sectional study

BMJ Open ◽  
2020 ◽  
Vol 10 (8) ◽  
pp. e033839
Author(s):  
Colette Andrea Cunningham-Myrie ◽  
Novie O Younger ◽  
Katherine P Theall ◽  
Lisa-Gaye Greene ◽  
Parris Lyew-Ayee ◽  
...  
Keyword(s):  
Social Class ◽  
Fast Food ◽  
Intraclass Correlation ◽  
Buffer Zone ◽  
Population Based ◽  
Cross Sectional Study ◽  
Sectional Study ◽  
Cross Sectional ◽  
Middle Income ◽  
Retail Food

ObjectiveTo derive estimates of the associations between measures of the retail food environments and mean body mass index (BMI) in Jamaica, a middle-income country with increasing prevalence of obesity.DesignCross-sectional study.SettingData from the Jamaica Health and Lifestyle Survey 2008 (JHLS II), a nationally representative population-based survey that recruited persons at their homes over a 4-month period from all 14 parishes and 113 neighbourhoods defined as enumeration districts.ParticipantsA subsample of 2529 participants aged 18–74 years from the JHLS II who completed interviewer-administered surveys, provided anthropometric measurements and whose addresses were geocoded.Primary outcome measureMean BMI, calculated as weight divided by height squared (kg/m2).ResultsThere was significant clustering across neighbourhoods for mean BMI (intraclass correlation coefficients=4.16%). Fully adjusted models revealed higher mean BMI among women, with further distance away from supermarkets (β=0.12; 95% CI 8.20×10−3, 0.24; p=0.036) and the absence of supermarkets within a 1 km buffer zone (β=1.36; 95% CI 0.20 to 2.52; p=0.022). A 10 km increase in the distance from a supermarket was associated with a 1.7 kg/m2 higher mean BMI (95% CI 0.03 to 0.32; p=0.020) in the middle class. No associations were detected with fast-food outlets or interaction by urbanicity.ConclusionsHigher mean BMI in Jamaicans may be partially explained by the presence of supermarkets and markets and differ by sex and social class. National efforts to curtail obesity in middle-income countries should consider interventions focused at the neighbourhood level that target the location and density of supermarkets and markets and consider sex and social class-specific factors that may be influencing the associations.

scholarly journals Are surface and deep learning approaches associated with study patterns and choices among medical students? A cross-sectional study

2018 ◽  
Vol 136 (5) ◽  
pp. 414-420 ◽  
Author(s):  
Álvaro Henrique de Almeida Delgado ◽  
João Paulo Rodrigues Almeida ◽  
Larissa Souza Borowski Mendes ◽  
Isabella Noceli de Oliveira ◽  
Oscarina da Silva Ezequiel ◽  
...  
Keyword(s):  
Deep Learning ◽  
Medical Students ◽  
Cross Sectional Study ◽  
Learning Approaches ◽  
Sectional Study ◽  
Cross Sectional

scholarly journals Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: A cross-sectional study

PLoS Medicine ◽  
2018 ◽  
Vol 15 (11) ◽  
pp. e1002683 ◽  
Author(s):  
John R. Zech ◽  
Marcus A. Badgeley ◽  
Manway Liu ◽  
Anthony B. Costa ◽  
Joseph J. Titano ◽  
...  
Keyword(s):  
Deep Learning ◽  
Learning Model ◽  
Cross Sectional Study ◽  
Chest Radiographs ◽  
Sectional Study ◽  
Cross Sectional ◽  
Generalization Performance ◽  
Deep Learning Model

scholarly journals Breast imaging hindered during covid-19 pandemic, in Brazil

2021 ◽  
Vol 55 ◽  
pp. 8
Author(s):  
Jordana de Faria Bessa
Keyword(s):  
Breast Imaging ◽  
Detection Rate ◽  
Cross Sectional Study ◽  
Federal State ◽  
Private Institutions ◽  
Sectional Study ◽  
Public Health Services ◽  
Cross Sectional ◽  
Entire Country ◽  
Open Access Database

OBJECTIVE: To report the decrease in breast imaging after covid-19 pandemic, obtaining the number of mammograms performed in 2019 and 2020. Additionally, to investigate if there was an increase in the proportion of women undergoing mammography for diagnostic purposes, with palpable lesions. METHOD: This is a cross-sectional study, based on the number of mammograms performed by the Brazilian public health services, provided by DATASUS, an open access database. Mammograms from private institutions were not included. This study compares the number of mammograms performed in 2019 and 2020, in women aged 50–69 years, stratified by month, in each federal state, and the presence of palpable lumps (physician-reported). RESULTS: In total, 1,948,471 mammograms were performed in 2019 and 1,126,688 in 2020, for the population studied. These values represent a 42% decline. Monthly, a significant decreased is observed after April 2020. The results varied slightly according to federal state; yet the entire country was affected. Rondônia was the most affected state, with 67% decline. The proportion of women presenting palpable lumps increased from 7.06% on average in 2019 to 7.94% in 2020 (OR = 1.135, 95%CI 1.125–1.145, p = 0,001). DISCUSSION: The number of mammograms performed in 2020 declined considerably. Out of the women who presented for mammogram, the proportion of palpable lumps was significantly higher in 2020. Considering the detection rate of digital mammography, the loss of 800,000 exams means 4,000 undiagnosed breast cancer cases, by the end of 2020.

scholarly journals Validity of four clinical prediction scores for pulmonary embolism in a sub-Saharan African setting: a protocol for a Cameroonian multicentre cross-sectional study

BMJ Open ◽  
2019 ◽  
Vol 9 (10) ◽  
pp. e031322
Author(s):  
Agnès Esiéné ◽  
Paul Owono Etoundi ◽  
Joel Noutakdie Tochie ◽  
Junette Arlette Mbengono Metogo ◽  
Jacqueline Ze Minkande
Keyword(s):  
Pulmonary Embolism ◽  
Emergency Departments ◽  
Diagnostic Performance ◽  
Cross Sectional Study ◽  
Sub Saharan Africa ◽  
Receiver Operating Curve ◽  
Clinical Prediction ◽  
Sectional Study ◽  
Cross Sectional ◽  
Sub Saharan

IntroductionPulmonary embolism poses one of the most challenging diagnoses in medicine. Resolving these diagnostic difficulties is more crucial in emergency departments where fast and accurate decisions are needed for a life-saving purpose. Here, clinical pretest evaluation is an important step in the diagnostic algorithm of pulmonary embolism. Although clinical probability scores are widely used in emergency departments of sub-Saharan Africa, no study has cited their diagnostic performance in this resource-constrained environment. This study will seek to assess the performance of four routinely used clinical prediction models in Cameroonians presenting with suspicion of pulmonary embolism at the emergency department.Methods and analysisIt will be a cross-sectional study comparing the sensitivity, specificity, positive and negative predictive values and accuracy of the Wells, Simplified Wells, Revised Geneva and the Simplified Revised Geneva Scores to CT pulmonary angiography as gold standard in all consecutive consenting patients aged above 15 years admitted for clinical suspicion of pulmonary embolism to the emergency departments of seven major referral hospitals of Cameroon between 1 July 2019 and 31 December 2020. The area under the receiver operating curve, calibration plots, Hosmer and Lemeshow statistics, observed/expected event rates, net benefit and decision curve will be measured of each the clinical prediction test to ascertain the clinical score with the best diagnostic performance.Ethics and disseminationClearance has been obtained from the Institutional Review Board of the Faculty of medicine and biomedical sciences of the University of Yaounde I, Cameroon and the directorates of all participating hospitals to conduct this study. Also, informed consent will be sought from each patient or their legal next of kin and parents for minors, before enrolment into this study. The final study will be published in a peer-review journal and the findings presented to health authorities and healthcare providers.

scholarly journals Interdevice agreement in respiratory resistance values by oscillometry in asthmatic children

2019 ◽  
Vol 5 (1) ◽  
pp. 00138-2018 ◽  
Author(s):  
Francine M. Ducharme ◽  
Imane Jroundi ◽  
Guillaume Jean ◽  
Guillaume Lavoie Boutin ◽  
Christiane Lawson ◽  
...  
Keyword(s):  
Intraclass Correlation ◽  
Random Order ◽  
Cross Sectional Study ◽  
Normative Values ◽  
Impulse Oscillometry ◽  
Respiratory Resistance ◽  
Sectional Study ◽  
Cross Sectional ◽  
Asthmatic Children ◽  
Bland Altman Method

BackgroundWith several commercially available devices measuring respiratory impedance by oscillometry, the agreement between values obtained on different instruments or frequencies remains unclear. Our aim was to examine the agreement between resistance and reactance parameters on two oscillometry instruments using different waveforms.MethodsWe conducted a prospective cross-sectional study in asthmatic children aged 3–17 years. Reproducible oscillometry measurements were obtained in random order, by blinded operators, at three modes: 5–10–15–20–25 Hz (5–25 Hz) multifrequency mode on the MasterScreen impulse oscillometry system, and both 5–25 Hz multifrequency mode and 7 Hz monofrequency on the tremoFlo C-100 airwave sinusoidal system. Resistance, reactance and within-breath parameters were examined using the intraclass correlation coefficient (ICC), paired t-test, linear regression and Bland–Altman method.ResultsOf 50 participants, 44 and 38 completed between-device and within-frequency measurements, respectively. Between-device measurements at 5–25 Hz showed high (ICC 0.88–0.91) and good (ICC 0.69–0.87) agreement in resistance and reactance, respectively, but with an absolute within-patient difference (≥0.05 kPa·L−1·s−1) and proportional bias (≥30% per kPa·L−1·s−1) in all parameters and oscillatory frequencies, apart from resistance at 5 Hz. A significant proportional bias was documented in most within-breath parameters at 5 versus 7 Hz on tremoFlo.ConclusionObserved differences in resistance and reactance suggest the need for instrument- and frequency-specific paediatric normative values.

scholarly journals Validity of self-reported age at menarche in computer-assisted interview among Chinese schoolgirls: a cross-sectional study

BMJ Open ◽  
2018 ◽  
Vol 8 (1) ◽  
pp. e016799 ◽  
Author(s):  
Yanyan Mao ◽  
Qiguo Lian ◽  
Xiayun Zuo ◽  
Yan Zhang ◽  
Shan Luo ◽  
...  
Keyword(s):  
Cross Sectional Study ◽  
Age At Menarche ◽  
Computer Assisted ◽  
Sectional Study ◽  
Cross Sectional

scholarly journals Improving the diagnostic performance of ultrasound in classifying breast lesions: the potential value of S-detect for residents-in-training

2019 ◽  
Author(s):  
Chenyang Zhao ◽  
Mengsu Xiao ◽  
Yuxin Jiang ◽  
He Liu ◽  
Ming Wang ◽  
...  
Keyword(s):  
Diagnostic Performance ◽  
Color Doppler ◽  
Intraclass Correlation ◽  
Area Under The Curve ◽  
Computer Assisted ◽  
Breast Lesions ◽  
Training Methods ◽  
Potential Value ◽  
Experienced Radiologist ◽  
Assisted Diagnosis

Abstract Background To explore the potential value of S-Detect™, a high-end computer-assisted diagnosis (CAD) software system for residents-in-training. Methods Routine breast ultrasound (US) examinations were conducted and assessed by an experienced radiologist. Archived images of the lesions (including grayscale, color Doppler flow and elastography images) were retrospectively assessed by each of five in-training residents who were blinded to the histopathological findings and any other US diagnosis. The diagnostic performances of S-Detect™ and the five residents were measured and compared. Afterwards, category 4a lesions assessed by the residents were downgraded when classified as possibly benign by S-Detect™. The diagnostic performance of the integrated results was compared with the original results of the residents. Results A total of 195 focal breast lesions were consecutively enrolled, including 82 malignant lesions and 113 benign lesions. S-Detect™ presented higher specificity and area under the curve (AUC)than the residents. After combination with S-Detect™ in category 4a lesions, the specificity and AUC of the five residents were significantly improved. The intraclass correlation coefficient (ICC) of the five residents also increased after integration. Conclusions With the help of the CAD software, the specificity, overall diagnostic performances and interobserver agreements of the residents greatly improved. S-Detect™ can be utilized as an assistant tool for residents-in-training in classifying breast lesions.

Sign in / Sign up

Export Citation Format

Share Document