Precise whole liver automatic segmentation and quantification of PDFF and R2* on MR images

Jiménez-Pastor, Ana; López‐González, Rafael; Martí-Aguado, David; França, Manuela; Bachmann, Rodrigo San Martin; Mazzucco, Juan; Martí‐Bonmatí, Luis

doi:10.1007/s00330-021-07838-5

Cited by 15 publications

(10 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Aiming at presenting the overall extent of fat metaplasia in the sacroiliac joints, this study adopted a sampling strategy instead of a whole-organ segmentation strategy. Common approaches to measuring quantitative imaging data include sampling strategy ( 28 ) and whole-organ segmentation strategy, with the latter one often applied in the assessment of organs in regular shapes, such as liver and pancreas ( 21 , 29 ). In the case of the SIJs, the subchondral bone area come in an irregular shape, rendering the process of manually delineating the subchondral bone “painstaking”.…”

Section: Discussionmentioning

confidence: 99%

“…Compared with other sequences such as T1-in-and-out-of-phase (IOP) MRI, CSE-MRI is a more sophisticated fat quantification approach which corrects for a number of confounding factors including T1 bias, noise bias and eddy currents ( 17 – 19 ). An iterative least-squares decomposition algorithm is used to solve for proton density fat fraction (PDFF) maps enabling the quantitative assessment of fatty lesions, which have been widely applied in the imaging of non-alcoholic fatty liver disease ( 20 , 21 ) as well as bone marrow fat in hematological diseases ( 22 ).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Quantification of Fat Metaplasia in the Sacroiliac Joints of Patients With Axial Spondyloarthritis by Chemical Shift-Encoded MRI: A Diagnostic Trial

et al. 2022

View full text Add to dashboard Cite

ObjectiveTo study the diagnostic performance of chemical shift-encoded MRI (CSE-MRI) in the diagnosis of axial spondyloarthritis (axSpA).MethodsCSE-MRI images were acquired for consecutive patients complaining of back pain as well as healthy volunteers. Proton density fat fraction (PDFF) values were measured independently by two readers. Diagnostic performance of CSE-MRI was analyzed by sensitivity analysis and ROC curve analysis. Logistic regression analysis was employed to investigate the risk factors of extensive fat deposition in the SIJs.ResultsA total of 52 r-axSpA patients, 37 nr-axSpA patients, 24 non-SpA patients and 34 healthy volunteers were included. Mean PDFF values in the SIJs of patients with r-axSpA and nr-axSpA (72.7% and 64.5%) were significantly higher than non-SpA patients and healthy volunteers (56.0% and 57.6%) (p<0.001). By defining extensive fat deposition in the SIJs as ≥8 ROIs with PDFF values over 70%, its sensitivity and specificity in diagnosing axSpA reached 72.47% and 86.21%%. By joining bone marrow edema (BME) with ≥8 ROIs (PDFF>70%), 22 (24.71%) and 23 (25.84%) more axSpA patients were classified as SIJ MRI (+) by reader 1 and 2, but specificities decreased by 15.52% and 10.34%. Multivariate logistic regression analysis confirmed longer disease duration as the independent risk factor of extensive fat deposition in SIJs (OR=1.15, 95%CI[1.03, 1.32]), while bDMARDs medication was a protective factor (OR=0.15, 95%CI[0.04, 0.51]).ConclusionCSE-MRI is a reliable tool to quantitively assess the fat metaplasia in the SIJs of axSpA patients. Extensive fat deposition in the SIJs could add incremental diagnostic value to BME, but at the cost of decreased specificities.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Quantification of Fat Metaplasia in the Sacroiliac Joints of Patients With Axial Spondyloarthritis by Chemical Shift-Encoded MRI: A Diagnostic Trial

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Compared with CT, there are few DL studies on magnetic resonance imaging (MRI). Most current research has focused on the diagnosis of liver, pancreatic, and rectal diseases, such as liver cancer [ 45 ], liver fibrosis [ 46 ], liver fat segmentation [ 47 ], pancreatic tumors [ 48 ], rectal cancer [ 49 ], etc. Abdominal organ segmentation and fat segmentation are the advantages of MRI.…”

Section: Application Of DL In Digestive System Imagingmentioning

confidence: 99%

A systematic review on application of deep learning in digestive system image processing

2021

View full text Add to dashboard Cite

“…It is also essential where correction for motion during examinations is required and is of particular importance in the delivery of radiation and other focal therapy. Repeatability is potentially improved by automation [7,8], while semi-automated methods increase accuracy and avoid the variability of manual methods [9]. Automation may employ methods such as adaptive thresholding and region growing, image gradient-based active contours and level sets, unsupervised statistical and clustering classification, supervised machine learning classification and deep learning methods based on convolutional neural networks [10][11][12].…”

Section: Introductionmentioning

confidence: 99%

Standardised lesion segmentation for imaging biomarker quantitation: a consensus recommendation from ESR and EORTC

et al. 2022

Self Cite

View full text Add to dashboard Cite

Background Lesion/tissue segmentation on digital medical images enables biomarker extraction, image-guided therapy delivery, treatment response measurement, and training/validation for developing artificial intelligence algorithms and workflows. To ensure data reproducibility, criteria for standardised segmentation are critical but currently unavailable. Methods A modified Delphi process initiated by the European Imaging Biomarker Alliance (EIBALL) of the European Society of Radiology (ESR) and the European Organisation for Research and Treatment of Cancer (EORTC) Imaging Group was undertaken. Three multidisciplinary task forces addressed modality and image acquisition, segmentation methodology itself, and standards and logistics. Devised survey questions were fed via a facilitator to expert participants. The 58 respondents to Round 1 were invited to participate in Rounds 2–4. Subsequent rounds were informed by responses of previous rounds. Results/conclusions Items with ≥ 75% consensus are considered a recommendation. These include system performance certification, thresholds for image signal-to-noise, contrast-to-noise and tumour-to-background ratios, spatial resolution, and artefact levels. Direct, iterative, and machine or deep learning reconstruction methods, use of a mixture of CE marked and verified research tools were agreed and use of specified reference standards and validation processes considered essential. Operator training and refreshment were considered mandatory for clinical trials and clinical research. Items with a 60–74% agreement require reporting (site-specific accreditation for clinical research, minimal pixel number within lesion segmented, use of post-reconstruction algorithms, operator training refreshment for clinical practice). Items with ≤ 60% agreement are outside current recommendations for segmentation (frequency of system performance tests, use of only CE-marked tools, board certification of operators, frequency of operator refresher training). Recommendations by anatomical area are also specified.

show abstract

Precise whole liver automatic segmentation and quantification of PDFF and R2* on MR images

Cited by 15 publications

References 31 publications

Quantification of Fat Metaplasia in the Sacroiliac Joints of Patients With Axial Spondyloarthritis by Chemical Shift-Encoded MRI: A Diagnostic Trial

Quantification of Fat Metaplasia in the Sacroiliac Joints of Patients With Axial Spondyloarthritis by Chemical Shift-Encoded MRI: A Diagnostic Trial

A systematic review on application of deep learning in digestive system image processing

Standardised lesion segmentation for imaging biomarker quantitation: a consensus recommendation from ESR and EORTC

Contact Info

Product

Resources

About