Quality control of fat fraction quantification in magnetic resonance imaging: A two-center phantom study
BACKGROUND: Assessment of quantitative parameters using magnetic resonance imaging (MRI) is a relevant trend. Fat fraction (FF) calculation provides new opportunities for accurate diagnosis and will replace invasive methods such as biopsy in the future. Quantification will enable reliable dynamic mo...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Eco-Vector
2023-06-01
|
| Series: | Digital Diagnostics |
| Subjects: | |
| Online Access: | https://jdigitaldiagnostics.com/DD/article/viewFile/430357/125009 |
| _version_ | 1827877980522676224 |
|---|---|
| author | Olga Yu. Panina Varvara A. Ignatyeva Alyona A. Monakhova |
| author_facet | Olga Yu. Panina Varvara A. Ignatyeva Alyona A. Monakhova |
| author_sort | Olga Yu. Panina |
| collection | DOAJ |
| description | BACKGROUND: Assessment of quantitative parameters using magnetic resonance imaging (MRI) is a relevant trend. Fat fraction (FF) calculation provides new opportunities for accurate diagnosis and will replace invasive methods such as biopsy in the future. Quantification will enable reliable dynamic monitoring and assessment of drug therapy. However, radiologists and clinical specialists must be confident in the accuracy and reliability of the quantitative measures.
AIM: To assess the accuracy of quantitative FF measurement using phantom simulation in the range of 0% to 60%.
METHODS: Emulsions of the oil-in-water type were chosen to simulate the objects of the study. Concentrations of vegetable oil-based emulsions were presented in the range of 0% to 60%. Tubes containing the emulsions were placed in a cylindrical phantom. Scans were performed on an Optima, MR450w 1.5 Tesla (GE) tomograph in Lava Flex mode and on an Ingenia, 1.5 Tesla (Philips) tomograph in DIXON mode. FF was determined by formulas using images in In-phase and Out-phase based on signal characteristics (FF=[InOut]/2∙In∙100) and images weighted by Water and Fat data (FF=Fat/[Fat+water]∙100.
RESULTS: The accuracy of the fat percentage measurement with the DIXON technique was identical to that of the Lava Flex. The data of the measured fat concentration were systematically overestimated in relation to the target values by an average of 57.6% with an average absolute difference of 17.2%. In addition, an irregular underestimation in the range of 20% to 40% was detected.
CONCLUSIONS: Phantom simulation using direct oil-in-water emulsions allowed to control the performance of the Dixon sequences in quantifying the FF. For correct FF quantification, calculation from Water and Fat image data using the formula FF=Fat/(Fat+water)∙100 is preferable. Calculations based on In-phase and Out-phase images provide ambiguous results. The FF calculation in the Lava Flex and DIXON modes must be performed with a correction factor. The use of the phantom allows proper quality control and calibration of the MRI scanner and makes quantitative fat measurement widely available. |
| first_indexed | 2024-03-12T17:44:07Z |
| format | Article |
| id | doaj.art-e165065562a9425190b1c173811d9191 |
| institution | Directory Open Access Journal |
| issn | 2712-8490 2712-8962 |
| language | English |
| last_indexed | 2024-03-12T17:44:07Z |
| publishDate | 2023-06-01 |
| publisher | Eco-Vector |
| record_format | Article |
| series | Digital Diagnostics |
| spelling | doaj.art-e165065562a9425190b1c173811d91912023-08-03T20:08:46ZengEco-VectorDigital Diagnostics2712-84902712-89622023-06-0141S969810.17816/DD43035776514Quality control of fat fraction quantification in magnetic resonance imaging: A two-center phantom studyOlga Yu. Panina0https://orcid.org/0000-0002-8684-775XVarvara A. Ignatyeva1https://orcid.org/0009-0009-6229-0342Alyona A. Monakhova2https://orcid.org/0009-0009-0271-2953Research and Practical Clinical Center for Diagnostics and Telemedicine TechnologiesA.I. Yevdokimov Moscow State University of Medicine and DentistryA.I. Yevdokimov Moscow State University of Medicine and DentistryBACKGROUND: Assessment of quantitative parameters using magnetic resonance imaging (MRI) is a relevant trend. Fat fraction (FF) calculation provides new opportunities for accurate diagnosis and will replace invasive methods such as biopsy in the future. Quantification will enable reliable dynamic monitoring and assessment of drug therapy. However, radiologists and clinical specialists must be confident in the accuracy and reliability of the quantitative measures. AIM: To assess the accuracy of quantitative FF measurement using phantom simulation in the range of 0% to 60%. METHODS: Emulsions of the oil-in-water type were chosen to simulate the objects of the study. Concentrations of vegetable oil-based emulsions were presented in the range of 0% to 60%. Tubes containing the emulsions were placed in a cylindrical phantom. Scans were performed on an Optima, MR450w 1.5 Tesla (GE) tomograph in Lava Flex mode and on an Ingenia, 1.5 Tesla (Philips) tomograph in DIXON mode. FF was determined by formulas using images in In-phase and Out-phase based on signal characteristics (FF=[InOut]/2∙In∙100) and images weighted by Water and Fat data (FF=Fat/[Fat+water]∙100. RESULTS: The accuracy of the fat percentage measurement with the DIXON technique was identical to that of the Lava Flex. The data of the measured fat concentration were systematically overestimated in relation to the target values by an average of 57.6% with an average absolute difference of 17.2%. In addition, an irregular underestimation in the range of 20% to 40% was detected. CONCLUSIONS: Phantom simulation using direct oil-in-water emulsions allowed to control the performance of the Dixon sequences in quantifying the FF. For correct FF quantification, calculation from Water and Fat image data using the formula FF=Fat/(Fat+water)∙100 is preferable. Calculations based on In-phase and Out-phase images provide ambiguous results. The FF calculation in the Lava Flex and DIXON modes must be performed with a correction factor. The use of the phantom allows proper quality control and calibration of the MRI scanner and makes quantitative fat measurement widely available.https://jdigitaldiagnostics.com/DD/article/viewFile/430357/125009dixon techniquefat fractionphantom |
| spellingShingle | Olga Yu. Panina Varvara A. Ignatyeva Alyona A. Monakhova Quality control of fat fraction quantification in magnetic resonance imaging: A two-center phantom study Digital Diagnostics dixon technique fat fraction phantom |
| title | Quality control of fat fraction quantification in magnetic resonance imaging: A two-center phantom study |
| title_full | Quality control of fat fraction quantification in magnetic resonance imaging: A two-center phantom study |
| title_fullStr | Quality control of fat fraction quantification in magnetic resonance imaging: A two-center phantom study |
| title_full_unstemmed | Quality control of fat fraction quantification in magnetic resonance imaging: A two-center phantom study |
| title_short | Quality control of fat fraction quantification in magnetic resonance imaging: A two-center phantom study |
| title_sort | quality control of fat fraction quantification in magnetic resonance imaging a two center phantom study |
| topic | dixon technique fat fraction phantom |
| url | https://jdigitaldiagnostics.com/DD/article/viewFile/430357/125009 |
| work_keys_str_mv | AT olgayupanina qualitycontroloffatfractionquantificationinmagneticresonanceimagingatwocenterphantomstudy AT varvaraaignatyeva qualitycontroloffatfractionquantificationinmagneticresonanceimagingatwocenterphantomstudy AT alyonaamonakhova qualitycontroloffatfractionquantificationinmagneticresonanceimagingatwocenterphantomstudy |