scholarly journals Stability of Radiomic Features across Different Region of Interest Sizes—A CT and MR Phantom Study

Tomography ◽  
2021 ◽  
Vol 7 (2) ◽  
pp. 238-252
Author(s):  
Laura J. Jensen ◽  
Damon Kim ◽  
Thomas Elgeti ◽  
Ingo G. Steffen ◽  
Bernd Hamm ◽  
...  
Keyword(s):  
Excellent Agreement ◽  
Correlation Coefficients ◽  
Region Of Interest ◽  
Systematic Difference ◽  
Gradient Echo ◽  
Mr Images ◽  
Concordance Correlation ◽  
Disease Entities ◽  
Different Diameters ◽  
Turbo Inversion Recovery Magnitude

We aimed to evaluate radiomic features’ stability across different region of interest (ROI) sizes in CT and MR images. We chose a phantom with a homogenous internal structure so no differences for a feature extracted from ROIs of different sizes would be expected. For this, we scanned a plastic cup filled with sodium chloride solution ten times in CT and per MR sequence (T1-weighted-gradient-echo and T2-weighted-turbo-inversion-recovery-magnitude). We placed sphere-shaped ROIs of different diameters (4, 8, and 16 mm, and 4, 8, and 16 pixels) into the phantom’s center. Features were extracted using PyRadiomics. We assessed feature stability across ROI sizes with overall concordance correlation coefficients (OCCCs). Differences were tested for significance with the Mann–Whitney U-test. Of 93 features, 87 T1w-derived, 87 TIRM-derived, and 70 CT-derived features were significantly different between ROI sizes. Among MR-derived features, OCCCs showed excellent (>0.90) agreement for mean, median, and root mean squared for ROI sizes between 4 and 16 mm and pixels. We further observed excellent agreement for 10th and 90th percentile in T1w and 10th percentile in T2w TIRM images. There was no excellent agreement among the OCCCs of CT-derived features. In summary, many features indicated significant differences and only few showed excellent agreement across varying ROI sizes, although we examined a homogenous phantom. Since we considered a small phantom in an experimental setting, further studies to investigate this size effect would be necessary for a generalization. Nevertheless, we believe knowledge about this effect is crucial in interpreting radiomics studies, as features that supposedly discriminate disease entities may only indicate a systematic difference in ROI size.

scholarly journals Stability of Liver Radiomics across Different 3D ROI Sizes—An MRI In Vivo Study

Tomography ◽  
2021 ◽  
Vol 7 (4) ◽  
pp. 866-876
Author(s):  
Laura J. Jensen ◽  
Damon Kim ◽  
Thomas Elgeti ◽  
Ingo G. Steffen ◽  
Bernd Hamm ◽  
...  
Keyword(s):  
Three Dimensional ◽  
Size Variation ◽  
Correlation Coefficients ◽  
Liver Parenchyma ◽  
Systematic Difference ◽  
3 Tesla ◽  
Concordance Correlation ◽  
The Stability ◽  
Different Diameters

We aimed to evaluate the stability of radiomic features in the liver of healthy individuals across different three-dimensional regions of interest (3D ROI) sizes in T1-weighted (T1w) and T2-weighted (T2w) images from different MR scanners. We retrospectively included 66 examinations of patients without known diseases or pathological imaging findings acquired on three MRI scanners (3 Tesla I: 25 patients, 3 Tesla II: 19 patients, 1.5 Tesla: 22 patients). 3D ROIs of different diameters (10, 20, 30 mm) were drawn on T1w GRE and T2w TSE images into the liver parenchyma (segment V–VIII). We extracted 93 radiomic features from the different ROIs and tested features for significant differences with the Mann–Whitney-U (MWU)-test. The MWU-test revealed significant differences for most second- and higher-order features, indicating a systematic difference dependent on the ROI size. The features mean, median, root mean squared (RMS), 10th percentile, and 90th percentile were not significantly different. We also assessed feature robustness to ROI size variation with overall concordance correlation coefficients (OCCCs). OCCCs across the different ROI-sizes for mean, median, and RMS were excellent (>0.90) in both sequences on all three scanners. These features, therefore, seem robust to ROI-size variation and suitable for radiomic studies of liver MRI.

An independent software system for the analysis of dynamic MR images

1997 ◽  
Vol 38 (1) ◽  
pp. 165-172 ◽  
Author(s):  
G. Torheim ◽  
M. Lombardi ◽  
P. A. Rinck
Keyword(s):  
Visual Analysis ◽  
Area Under The Curve ◽  
Correlation Coefficients ◽  
Region Of Interest ◽  
Mr Images ◽  
Relative Signal Intensity ◽  
Time To Peak ◽  
File Formats ◽  
Dynamic Mr ◽  
Cine Loop

Purpose: A computer system for the manual, semi-automatic, and automatic analysis of dynamic MR images was to be developed on UNIX and personal computer platforms. The system was to offer an integrated and standardized way of performing both image processing and analysis that was independent of the MR unit used. Material and Methods: The system consists of modules that are easily adaptable to special needs. Data from MR units or other diagnostic imaging equipment in techniques such as CT, ultrasonography, or nuclear medicine can be processed through the ACR-NEMA/DICOM standard file formats. A full set of functions is available, among them cine-loop visual analysis, and generation of time-intensity curves., Parameters such as cross-correlation coefficients, area under the curve, peak/maximum intensity, wash-in and wash-out slopes, time to peak, and relative signal intensity/contrast enhancement can be calculated. Other parameters can be extracted by fitting functions like the gamma-variate function. Region-of-interest data and parametric values can easily be exported. Results and Conclusion: The system has been successfully tested in animal and patient examinations.

scholarly journals Stability of MRI radiomic features according to various imaging parameters in fast scanned T2-FLAIR for acute ischemic stroke patients

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Leehi Joo ◽  
Seung Chai Jung ◽  
Hyunna Lee ◽  
Seo Young Park ◽  
Minjae Kim ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Acute Ischemic Stroke ◽  
Correlation Coefficients ◽  
Region Of Interest ◽  
Planar Imaging ◽  
Concordance Correlation ◽  
Repetition Time ◽  
Group A ◽  
Train Length ◽  
Echo Planar

AbstractFrom May 2015 to June 2016, data on 296 patients undergoing 1.5-Tesla MRI for symptoms of acute ischemic stroke were retrospectively collected. Conventional, echo-planar imaging (EPI) and echo train length (ETL)-T2-FLAIR were simultaneously obtained in 118 patients (first group), and conventional, ETL-, and repetition time (TR)-T2-FLAIR were simultaneously obtained in 178 patients (second group). A total of 595 radiomics features were extracted from one region-of-interest (ROI) reflecting the acute and chronic ischemic hyperintensity, and concordance correlation coefficients (CCC) of the radiomics features were calculated between the fast scanned and conventional T2-FLAIR for paired patients (1st group and 2nd group). Stabilities of the radiomics features were compared with the proportions of features with a CCC higher than 0.85, which were considered to be stable in the fast scanned T2-FLAIR. EPI-T2-FLAIR showed higher proportions of stable features than ETL-T2-FLAIR, and TR-T2-FLAIR also showed higher proportions of stable features than ETL-T2-FLAIR, both in acute and chronic ischemic hyperintensities of whole- and intersection masks (p < .002). Radiomics features in fast scanned T2-FLAIR showed variable stabilities according to the sequences compared with conventional T2-FLAIR. Therefore, radiomics features may be used cautiously in applications for feature analysis as their stability and robustness can be variable.

scholarly journals Stability of Radiomic Features in Fast Scanned FLAIR for Acute Ischemic Stroke Patients: A Comparison With Conventional FLAIR 

2021 ◽  
Author(s):  
Leehi Joo ◽  
Seung Chai Jung ◽  
Hyunna Lee ◽  
Seo Young Park ◽  
Minjae Kim ◽  
...  
Keyword(s):  
Ischemic Stroke ◽  
Acute Ischemic Stroke ◽  
Correlation Coefficients ◽  
Region Of Interest ◽  
Planar Imaging ◽  
Concordance Correlation ◽  
Repetition Time ◽  
Group A ◽  
Train Length ◽  
Echo Planar

Abstract From May 2015 to June 2016, 296 patients undergoing 1.5-Tesla MRI for symptoms of acute ischemic stroke were retrospectively enrolled. Conventional, echo-planar imaging (EPI) and echo train length (ETL)-FLAIR were simultaneously obtained in 118 patients (1st group), and conventional, ETL-, and repetition time (TR)-FLAIR were simultaneously obtained in 178 patients (2nd group). A total of 595 radiomics features were extracted from one region-of-interest (ROI) reflecting the acute and chronic ischemic hyperintensity, and concordance correlation coefficients (CCC) of the radiomics features were calculated between the fast scanned and conventional FLAIR for paired patients (1st group and 2nd group). Stabilities of the radiomics features were compared with the proportions of features with a CCC higher than 0.85, which were considered to be stable in the fast scanned FLAIR. EPI-FLAIR showed higher proportions of stable features than ETL-FLAIR, and TR-FLAIR also showed higher proportions of stable features than ETL-FLAIR, both in acute and chronic ischemic hyperintensities of whole- and intersection masks (p <.002). Radiomics features in fast scanned FLAIR showed variable stabilities according to the sequences compared with conventional FLAIR. Therefore, radiomics features may be used cautiously in applications for feature analysis as their stability and robustness can be variable.

scholarly journals AB0798 IS THERE A REPRODUCIBILITY OF THE HISTOGRAM AND GRAY SCALE IN ECOEXTRUCTURAL DAMAGES IN RHEUMATOTOLOGICAL DISEASES?

2021 ◽  
Vol 80 (Suppl 1) ◽  
pp. 1423.2-1424
Author(s):  
J. A. Mendonça ◽  
I. Siste de Almeida Aoki ◽  
C. C. Cavuto ◽  
V. A. Leandro-Merhi ◽  
J. L. B. D. Aquino
Keyword(s):  
Correlation Coefficient ◽  
Intraclass Correlation Coefficient ◽  
Rheumatic Diseases ◽  
Femoral Condyle ◽  
Intraclass Correlation ◽  
Correlation Coefficients ◽  
Concordance Correlation Coefficient ◽  
Image Method ◽  
Concordance Correlation ◽  
Gray Scale

Background:The gray scale (GS) in high resolution ultrasound is already well validated for use in rheumatological diseases, but the color map or the histogram, can be considered a new proposal, to better define and complement the echotextural damages detection1.Objectives:To calculate the lesions area measures reproducibility index in arthropathies, between 3 blind evaluators and correlate these measures using the GS and the histogram.Methods:Observational and retrospective study approved by the ethics committee of the Pontifical Catholic University of Campinas, with the opinion number: 1.526.307. A total of 29 patients have been assessed (31% males and 69% females) on period 2014 to 2019 in Rheumatology service. A MyLab 50 -Esaote equipment was used with frequency transducer that ranged between 6.0 and 18.0 MHz, 10 different area measures were performed from each recorded images previously, by the GS and the histogram. Statistical analysis: Spearman’s correlation coefficients, Lin’s concordance coefficient (CCC) and the intraclass correlation coefficient (ICC) and their respective 95% confidence intervals, with the SPSS software package for Windows v. 17.0 (SPSS Inc., Chicago, IL, USA).Results:Average age 43.5 ± 21.5 years of age; with disease duration that varied between ≤ 1 month (48.3%) and ≥36 months (24.1%); with the following diseases: juvenile idiopathic arthritis (17.24%); osteoarthritis (13.79%); psoriatic arthritis (13.79%); undifferentiated spondyloarthritis (3.44%); gout (20.68%); rheumatoid arthritis (27.58%) and reactive arthritis (3.44%). A total of 840 measures of exudative (27.58%), proliferative (27.58%) and snowstorm appearance (6.89%) synovitis were performed; femoral-condyle cartilage (3.44%); synovial cyst (3.44%); paratendinitis (6.89%); calcification (3.44%); nail enthesitis (3.44%); tenosynovitis (6.89%) and tophi (10.34%) (Figure 1). The concordance correlation coefficient showed values closer to 1; p <0.001, the intraclass correlation coefficients with excellent reproducibility (ICC ≥ 0.75); p <0.001, always in relation to the three evaluators (Table 1) and the Spearman correlation between the GS and the histogram ranged from rs = 0.665 to rs = 1,000; p <0.001.Conclusion:The histogram can be considered an image method to better identify echotextural damages.References:[1]Mendonça J, Provenza J, Guissa V, et al AB1059 2D Histogram Ultrasound and 3D Ultrasound Correlation in Rheumatic Diseases Annals of the Rheumatic Diseases 2015; 74:1253-1254.Table 1.Concordance Correlation Coefficient (CCC) and Intraclass Correlation Coefficient (ICC):EvaluatorsGS - CCC(IC 95%)p-valorHistogram (IC 95%)p-valor1 e 20,998(0,994-0,999)<0,0010,999(0,995-1,000)<0,0011 e 30,998(0,995-0,999)<0,0010,999(0,995-1,000)<0,0012 e 30,992(0,980-0,997)<0,010,996(0,979-0,999)<0,01Standard by US 2DICCp-valorGS0,997(0,992-0,999)<0,001Histogram0,998(0,992-0,999)<0,001Legends: Gray Scale (GS).Figure 1.Patient with gout: A and B: Tophi area measures (star) in right metatarsos and efusion (arrow) by GS (45 mm2) and histogram (39 mm2), respectively.Disclosure of Interests:José Alexandre Mendonça Speakers bureau: Novartis, Janssen, Bristol, UCB, Isabella Siste de Almeida Aoki: None declared, Caique Chagas Cavuto: None declared, Vânia Aparecida Leandro-Merhi: None declared, José Luis Braga de Aquino: None declared

Reliability of Shear Wave Ultrasound Elastography for Neck Lesions Identified in Routine Clinical Practice

2012 ◽  
Vol 33 (05) ◽  
pp. 463-468 ◽  
Author(s):  
K. Bhatia ◽  
C. S. L. Tong ◽  
C. C. M. Cho ◽  
E. H. Y. Yuen ◽  
J. Lee ◽  
...  
Keyword(s):  
Lymph Nodes ◽  
Shear Wave ◽  
Correlation Coefficients ◽  
Routine Clinical Practice ◽  
Ultrasound Elastography ◽  
Concordance Correlation ◽  
Substantial Agreement ◽  
P Values ◽  
Static Image ◽  
Acquisition Technique

Abstract Purpose: To evaluate the reliability of shear wave ultrasound elastography (SWE) in the neck. Materials and Methods: 176 neck lesions (40 thy roid, 56 lymph nodes, 46 salivary, 34 miscella neous) identified in a routine US clinic underwent SWE by one or two blinded radiologists. For this study, SWE required the operator to acquire three 10 second dynamic colour-coded SWE cineloops per lesion, select one static image per cineloop, and place circular regions-of-interest within the entire lesion and stiffest part to generate 3 SWE measurements per static image. For logistical rea sons, one radiologist evaluated all 176 lesions and the other evaluated 58 lesions. Both radiologists also reviewed 27 archived cineloops independ ently to assess SWE excluding practical technique. Reliability was assessed using intraclass correla tion coefficients (ICCs) concordance correlation coefficients (CCCs) and coefficients of repeatabil ity (CORs). Results: Test-retest ICCs for the radiologist evalu ating 176 lesions were 0.78–0.85 (fair-excellent agreement), CCCs were 0.85–0.88 (substantial agreement), and CORs were 14.9–36.1 kPa. For both radiologists evaluating 58 lesions, intra-rater and inter-rater ICCs were 0.65–0.78 and 0.72–0.77 respectively. For SWE excluding practical technique, inter-rater ICCs were 0.97–0.98 (ex cellent agreement). ICCs differed according to tis sue, being higher in thyroid lesions than lymph nodes (p < 0.001), and higher in benign than ma lignant lesions (p values < 0.001). Conclusion: Intra- and inter-rater reliability of SWE is fair to excellent according to ICCs. SWE re liability is influenced appreciably by acquisition technique. Nevertheless, CORs for SWE are not negligible. To determine whether these results are acceptable clinically, further research is re quired to establish SWE stiffness values of normal and pathological tissues in the neck.

The superior pericardial sinus: normal appearance on gradient-echo MR images.

1993 ◽  
Vol 160 (4) ◽  
pp. 749-751 ◽  
Author(s):  
C M Black ◽  
L K Hedges ◽  
M C Javitt
Keyword(s):  
Gradient Echo ◽  
Mr Images ◽  
Normal Appearance

Is the location of the signal intensity weighted centroid a reliable measurement of fluid displacement within the disc?

2018 ◽  
Vol 63 (4) ◽  
pp. 453-460 ◽  
Author(s):  
Vahid Abdollah ◽  
Eric C. Parent ◽  
Michele C. Battié
Keyword(s):  
Signal Intensity ◽  
Water Distribution ◽  
Intraclass Correlation ◽  
Correlation Coefficients ◽  
Region Of Interest ◽  
Rater Reliability ◽  
Fluid Displacement ◽  
Intraclass Correlation Coefficients ◽  
The Mean ◽  
Standard Error Of Measurement

Abstract Degenerated discs have shorter T2-relaxation time and lower MR signal. The location of the signal-intensity-weighted-centroid reflects the water distribution within a region-of-interest (ROI). This study compared the reliability of the location of the signal-intensity-weighted-centroid to mean signal intensity and area measurements. L4-L5 and L5-S1 discs were measured on 43 mid-sagittal T2-weighted 3T MRI images in adults with back pain. One rater analysed images twice and another once, blinded to measurements. Discs were semi-automatically segmented into a whole disc, nucleus, anterior and posterior annulus. The coordinates of the signal-intensity-weighted-centroid for all regions demonstrated excellent intraclass-correlation-coefficients for intra- (0.99–1.00) and inter-rater reliability (0.97–1.00). The standard error of measurement for the Y-coordinates of the signal-intensity-weighted-centroid for all ROIs were 0 at both levels and 0 to 2.7 mm for X-coordinates. The mean signal intensity and area for the whole disc and nucleus presented excellent intra-rater reliability with intraclass-correlation-coefficients from 0.93 to 1.00, and 0.92 to 1.00 for inter-rater reliability. The mean signal intensity and area had lower reliability for annulus ROIs, with intra-rater intraclass-correlation-coefficient from 0.5 to 0.76 and inter-rater from 0.33 to 0.58. The location of the signal-intensity-weighted-centroid is a reliable biomarker for investigating the effects of disc interventions.

Sign in / Sign up

Export Citation Format

Share Document