Neural networks with personalized training for improved MOLLI T1 mapping
Abstract Background The aim of this study was to develop a method for personalized training of Deep Neural Networks by means of an MRI simulator to improve MOLLI native T1 estimates relative to conventional fitting methods. Methods The proposed Personalized Training Neural Network (PTNN) for T1 mapp...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2025-07-01
|
| Series: | BMC Medical Imaging |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12880-025-01769-z |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849334174702370816 |
|---|---|
| author | Olympia Gkatsoni Christos G. Xanthis Sebastian Johansson Einar Heiberg Håkan Arheden Anthony H. Aletras |
| author_facet | Olympia Gkatsoni Christos G. Xanthis Sebastian Johansson Einar Heiberg Håkan Arheden Anthony H. Aletras |
| author_sort | Olympia Gkatsoni |
| collection | DOAJ |
| description | Abstract Background The aim of this study was to develop a method for personalized training of Deep Neural Networks by means of an MRI simulator to improve MOLLI native T1 estimates relative to conventional fitting methods. Methods The proposed Personalized Training Neural Network (PTNN) for T1 mapping was based on a neural network which was trained with simulated MOLLI signals generated for each individual scan, taking into account both the pulse sequence parameters and the heart rate triggers of the specific healthy volunteer. Experimental data from eleven phantoms and ten healthy volunteers were included in the study. Results In phantom studies, agreement between T1 reference values and those obtained with the PTNN yielded a statistically significant smaller bias than conventional fitting estimates (-26.69 ± 29.5ms vs. -65.0 ± 33.25ms, p < 0.001). For in vivo studies, T1 estimates derived from the PTNN yielded higher T1 values (1152.4 ± 25.8ms myocardium, 1640.7 ± 30.6ms blood) than conventional fitting (1050.8 ± 24.7ms myocardium, 1597.2 ± 39.9ms blood). For PTNN, shortening the acquisition time by eliminating the pause between inversion pulses yielded higher myocardial T1 values (1162.2 ± 19.7ms with pause vs. 1127.1 ± 19.7ms, p = 0.01 myocardium), (1624.7 ± 33.9ms with pause vs. 1645.4 ± 18.7ms, p = 0.16 blood). For conventional fitting statistically significant differences were found. Conclusions Compared to T1 maps derived by conventional fitting, PTNN is a post-processing method that yielded T1 maps with higher values and better accuracy in phantoms for a physiological range of T1 and T2 values. In normal volunteers PTNN yielded higher T1 values even with a shorter acquisition scheme of eight heartbeats scan time, without deploying new pulse sequences. |
| format | Article |
| id | doaj-art-5c4d510da8c64d719cb0937a3e2f4419 |
| institution | Kabale University |
| issn | 1471-2342 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Medical Imaging |
| spelling | doaj-art-5c4d510da8c64d719cb0937a3e2f44192025-08-20T03:45:39ZengBMCBMC Medical Imaging1471-23422025-07-0125111610.1186/s12880-025-01769-zNeural networks with personalized training for improved MOLLI T1 mappingOlympia Gkatsoni0Christos G. Xanthis1Sebastian Johansson2Einar Heiberg3Håkan Arheden4Anthony H. Aletras5Laboratory of Computing, Medical Informatics and Biomedical – Imaging Technologies, School of Medicine, Aristotle University of ThessalonikiClinical Physiology, Department of Clinical Sciences Lund, Lund University, Skåne University HospitalClinical Physiology, Department of Clinical Sciences Lund, Lund University, Skåne University HospitalClinical Physiology, Department of Clinical Sciences Lund, Lund University, Skåne University HospitalClinical Physiology, Department of Clinical Sciences Lund, Lund University, Skåne University HospitalLaboratory of Computing, Medical Informatics and Biomedical – Imaging Technologies, School of Medicine, Aristotle University of ThessalonikiAbstract Background The aim of this study was to develop a method for personalized training of Deep Neural Networks by means of an MRI simulator to improve MOLLI native T1 estimates relative to conventional fitting methods. Methods The proposed Personalized Training Neural Network (PTNN) for T1 mapping was based on a neural network which was trained with simulated MOLLI signals generated for each individual scan, taking into account both the pulse sequence parameters and the heart rate triggers of the specific healthy volunteer. Experimental data from eleven phantoms and ten healthy volunteers were included in the study. Results In phantom studies, agreement between T1 reference values and those obtained with the PTNN yielded a statistically significant smaller bias than conventional fitting estimates (-26.69 ± 29.5ms vs. -65.0 ± 33.25ms, p < 0.001). For in vivo studies, T1 estimates derived from the PTNN yielded higher T1 values (1152.4 ± 25.8ms myocardium, 1640.7 ± 30.6ms blood) than conventional fitting (1050.8 ± 24.7ms myocardium, 1597.2 ± 39.9ms blood). For PTNN, shortening the acquisition time by eliminating the pause between inversion pulses yielded higher myocardial T1 values (1162.2 ± 19.7ms with pause vs. 1127.1 ± 19.7ms, p = 0.01 myocardium), (1624.7 ± 33.9ms with pause vs. 1645.4 ± 18.7ms, p = 0.16 blood). For conventional fitting statistically significant differences were found. Conclusions Compared to T1 maps derived by conventional fitting, PTNN is a post-processing method that yielded T1 maps with higher values and better accuracy in phantoms for a physiological range of T1 and T2 values. In normal volunteers PTNN yielded higher T1 values even with a shorter acquisition scheme of eight heartbeats scan time, without deploying new pulse sequences.https://doi.org/10.1186/s12880-025-01769-zCardiac MRIDeep learningT1 mappingMRI simulator |
| spellingShingle | Olympia Gkatsoni Christos G. Xanthis Sebastian Johansson Einar Heiberg Håkan Arheden Anthony H. Aletras Neural networks with personalized training for improved MOLLI T1 mapping BMC Medical Imaging Cardiac MRI Deep learning T1 mapping MRI simulator |
| title | Neural networks with personalized training for improved MOLLI T1 mapping |
| title_full | Neural networks with personalized training for improved MOLLI T1 mapping |
| title_fullStr | Neural networks with personalized training for improved MOLLI T1 mapping |
| title_full_unstemmed | Neural networks with personalized training for improved MOLLI T1 mapping |
| title_short | Neural networks with personalized training for improved MOLLI T1 mapping |
| title_sort | neural networks with personalized training for improved molli t1 mapping |
| topic | Cardiac MRI Deep learning T1 mapping MRI simulator |
| url | https://doi.org/10.1186/s12880-025-01769-z |
| work_keys_str_mv | AT olympiagkatsoni neuralnetworkswithpersonalizedtrainingforimprovedmollit1mapping AT christosgxanthis neuralnetworkswithpersonalizedtrainingforimprovedmollit1mapping AT sebastianjohansson neuralnetworkswithpersonalizedtrainingforimprovedmollit1mapping AT einarheiberg neuralnetworkswithpersonalizedtrainingforimprovedmollit1mapping AT hakanarheden neuralnetworkswithpersonalizedtrainingforimprovedmollit1mapping AT anthonyhaletras neuralnetworkswithpersonalizedtrainingforimprovedmollit1mapping |