Search PubMed⌕ Search
Home/Archive/PubMed

PubMed · 7016765

Fitting a linear relationship with confidence intervals by correlation.

J Winter·A A Afifi·D Sarti
View original source ↗Export metadata ↓

Abstract

The study of the empirical linear relationship between two measured variables should use correlation analysis which treats the variables symmetrically, rather than linear regression which inappropriately assumes a dependent and an independent variable. An APL computer program calculates the best fitting bivariate Gaussian distribution and plots the data with the superimposed straight line relationship and the confidence interval.

Explore related subjects

Body ConstitutionCongenital AbnormalitiesDiagnosis, Computer-AssistedFemaleHeadHumansPregnancyPrenatal DiagnosisReference ValuesUltrasonographyComparative StudyJournal Article

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

J Winter, A A Afifi, D Sarti. 1981. Fitting a linear relationship with confidence intervals by correlation.. https://doi.org/10.1016/0020-7101(81)90026-x

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2006

Influence of patient thickness and operation modes on occupational and patient radiation doses in interventional cardiology.

Patient and staff dose values in an interventional cardiology laboratory for different operational modes and several patient thicknesses (from 16 to 28 cm, simulated using polymethylmethacrylate) are presented. When increasing patient thicknesses and depending on fluoroscopy and cine modes, occupational doses can increase >30 times the baseline level. Scatter dose rates at the cardiologist's position with no radiation protective tools ranged from 1 to 14 mSv h(-1) for fluoroscopy, and from 10 to 47 mSv h(-1) during cine acquisition. Patient entrance surface air kerma rates increased by nearly 3 and staff dose rates by up to 2.6 when fluoroscopy was moved from the low to the high mode, for a typical 20 cm thickness. The respective increase factors were 6 and 4.2 when patient thickness rose from 16 to 28 cm, and by 10 and 8.3, when comparing cine acquisition with the low fluoroscopy mode. The knowledge of typical dose rates for each X-ray system in use in catheterisation laboratories is essential in order to optimise protection of patients and staff.

Body Constitution↗
PubMed2006

Simulator-assisted setting of scan protocols for X-ray CT: development and clinical usefulness of the Scan Plan Simulator.

To assist in the selection of complicated computed tomography (CT) scan protocols and to obtain stable image SD values, prototype "Scan Plan Simulator" software with the following functions was developed and evaluated. 1) The image SD value that will be obtained in actual scanning is estimated by entering the patient's body size and scan protocol, after which a simulated image is displayed so that the estimated image SD value can be checked. The exposure dose can also be estimated in the Scan Plan Simulator. 2) The appropriate tube current is automatically set by entering the required image SD value. We evaluated the accuracy of the Scan Plan Simulator by comparing the simulation results with the image SD values and exposure dose obtained in actual scanning and assessed the usefulness of this method using Monte Carlo simulation based on body thickness data obtained in clinical examinations. The results showed that the Scan Plan Simulator not only stabilizes image SD by minimizing the effects of body size but also permits the exposure dose to be reduced by optimizing tube current.

Body Constitution↗