Zhiqing Yin, Huay Din, Jessie E P Sun, Christina J MacAskill, Sree Harsha Tirumani, Pew-Thian Yap, Mark Griswold, Chris A Flask, Yong Chen
{"title":"使用基于深度学习的组织量化加速二维肾脏磁共振指纹识别。","authors":"Zhiqing Yin, Huay Din, Jessie E P Sun, Christina J MacAskill, Sree Harsha Tirumani, Pew-Thian Yap, Mark Griswold, Chris A Flask, Yong Chen","doi":"10.1002/jmri.70150","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Magnetic Resonance Fingerprinting (MRF) is a technique that can provide rapid quantification of multiple tissue properties. Deep learning may potentially contribute to an accelerated acquisition of MRF.</p><p><strong>Purpose: </strong>(1) To develop a deep learning method to accelerate the acquisition for kidney MRF; (2) to evaluate its performance in healthy subjects and patients with renal masses.</p><p><strong>Study type: </strong>Retrospective and based on internal reference data.</p><p><strong>Subjects: </strong>Development set was 36 healthy subjects and 20 patients with renal masses. The testing set: 4 healthy subjects and 16 patients.</p><p><strong>Field strength/sequence: </strong>3T, Steady-State Free Precession (FISP)-based MRF.</p><p><strong>Assessment: </strong>Quantification accuracy was evaluated in healthy kidneys and renal masses using quantitative metrics including normalized root-mean-square error (NRMSE) calculated based on reference maps generated using the standard template matching approach with all acquired MRF time frames.</p><p><strong>Statistical tests: </strong>Paired Student's t-test. p < 0.05 was considered statistically significant.</p><p><strong>Results: </strong>Accurate quantification in both T<sub>1</sub> (NRMSE = 0.025 ± 0.003) and T<sub>2</sub> (NRMSE = 0.053 ± 0.010) maps was obtained for healthy kidney tissues with a three-fold acceleration (576 time frames, 5 s of scan time), outperforming the template matching approach (T<sub>1</sub>, NRMSE = 0.057 ± 0.015; T<sub>2</sub>, NRMSE = 0.143 ± 0.080). For renal masses with T<sub>1</sub> and T<sub>2</sub> values in close range of healthy kidney tissues, similar performance was achieved with a three-fold acceleration. For renal masses presenting distinct T<sub>1</sub> or T<sub>2</sub> values, more MRF time frames were required to provide accurate tissue quantification. No significant difference was noticed in tissue/tumor quantification between neural networks trained using only healthy subjects versus a mixed dataset with healthy subjects and patients (p > 0.05).</p><p><strong>Conclusion: </strong>A deep learning-based method was developed to accelerate acquisition without compromising the accuracy of relaxation time mapping using kidney MRF. These results demonstrate reliable tissue quantification with at least a two-fold acceleration for both healthy kidneys and renal masses with various subtypes and histopathological grades.</p><p><strong>Evidence level: </strong>4.</p><p><strong>Technical efficacy: </strong>Stage 1.</p>","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5000,"publicationDate":"2025-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Accelerating 2D Kidney Magnetic Resonance Fingerprinting Using Deep Learning Based Tissue Quantification.\",\"authors\":\"Zhiqing Yin, Huay Din, Jessie E P Sun, Christina J MacAskill, Sree Harsha Tirumani, Pew-Thian Yap, Mark Griswold, Chris A Flask, Yong Chen\",\"doi\":\"10.1002/jmri.70150\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Background: </strong>Magnetic Resonance Fingerprinting (MRF) is a technique that can provide rapid quantification of multiple tissue properties. Deep learning may potentially contribute to an accelerated acquisition of MRF.</p><p><strong>Purpose: </strong>(1) To develop a deep learning method to accelerate the acquisition for kidney MRF; (2) to evaluate its performance in healthy subjects and patients with renal masses.</p><p><strong>Study type: </strong>Retrospective and based on internal reference data.</p><p><strong>Subjects: </strong>Development set was 36 healthy subjects and 20 patients with renal masses. The testing set: 4 healthy subjects and 16 patients.</p><p><strong>Field strength/sequence: </strong>3T, Steady-State Free Precession (FISP)-based MRF.</p><p><strong>Assessment: </strong>Quantification accuracy was evaluated in healthy kidneys and renal masses using quantitative metrics including normalized root-mean-square error (NRMSE) calculated based on reference maps generated using the standard template matching approach with all acquired MRF time frames.</p><p><strong>Statistical tests: </strong>Paired Student's t-test. p < 0.05 was considered statistically significant.</p><p><strong>Results: </strong>Accurate quantification in both T<sub>1</sub> (NRMSE = 0.025 ± 0.003) and T<sub>2</sub> (NRMSE = 0.053 ± 0.010) maps was obtained for healthy kidney tissues with a three-fold acceleration (576 time frames, 5 s of scan time), outperforming the template matching approach (T<sub>1</sub>, NRMSE = 0.057 ± 0.015; T<sub>2</sub>, NRMSE = 0.143 ± 0.080). For renal masses with T<sub>1</sub> and T<sub>2</sub> values in close range of healthy kidney tissues, similar performance was achieved with a three-fold acceleration. For renal masses presenting distinct T<sub>1</sub> or T<sub>2</sub> values, more MRF time frames were required to provide accurate tissue quantification. No significant difference was noticed in tissue/tumor quantification between neural networks trained using only healthy subjects versus a mixed dataset with healthy subjects and patients (p > 0.05).</p><p><strong>Conclusion: </strong>A deep learning-based method was developed to accelerate acquisition without compromising the accuracy of relaxation time mapping using kidney MRF. These results demonstrate reliable tissue quantification with at least a two-fold acceleration for both healthy kidneys and renal masses with various subtypes and histopathological grades.</p><p><strong>Evidence level: </strong>4.</p><p><strong>Technical efficacy: </strong>Stage 1.</p>\",\"PeriodicalId\":16140,\"journal\":{\"name\":\"Journal of Magnetic Resonance Imaging\",\"volume\":\" \",\"pages\":\"\"},\"PeriodicalIF\":3.5000,\"publicationDate\":\"2025-10-14\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Magnetic Resonance Imaging\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1002/jmri.70150\",\"RegionNum\":2,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Magnetic Resonance Imaging","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1002/jmri.70150","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING","Score":null,"Total":0}
Accelerating 2D Kidney Magnetic Resonance Fingerprinting Using Deep Learning Based Tissue Quantification.
Background: Magnetic Resonance Fingerprinting (MRF) is a technique that can provide rapid quantification of multiple tissue properties. Deep learning may potentially contribute to an accelerated acquisition of MRF.
Purpose: (1) To develop a deep learning method to accelerate the acquisition for kidney MRF; (2) to evaluate its performance in healthy subjects and patients with renal masses.
Study type: Retrospective and based on internal reference data.
Subjects: Development set was 36 healthy subjects and 20 patients with renal masses. The testing set: 4 healthy subjects and 16 patients.
Field strength/sequence: 3T, Steady-State Free Precession (FISP)-based MRF.
Assessment: Quantification accuracy was evaluated in healthy kidneys and renal masses using quantitative metrics including normalized root-mean-square error (NRMSE) calculated based on reference maps generated using the standard template matching approach with all acquired MRF time frames.
Statistical tests: Paired Student's t-test. p < 0.05 was considered statistically significant.
Results: Accurate quantification in both T1 (NRMSE = 0.025 ± 0.003) and T2 (NRMSE = 0.053 ± 0.010) maps was obtained for healthy kidney tissues with a three-fold acceleration (576 time frames, 5 s of scan time), outperforming the template matching approach (T1, NRMSE = 0.057 ± 0.015; T2, NRMSE = 0.143 ± 0.080). For renal masses with T1 and T2 values in close range of healthy kidney tissues, similar performance was achieved with a three-fold acceleration. For renal masses presenting distinct T1 or T2 values, more MRF time frames were required to provide accurate tissue quantification. No significant difference was noticed in tissue/tumor quantification between neural networks trained using only healthy subjects versus a mixed dataset with healthy subjects and patients (p > 0.05).
Conclusion: A deep learning-based method was developed to accelerate acquisition without compromising the accuracy of relaxation time mapping using kidney MRF. These results demonstrate reliable tissue quantification with at least a two-fold acceleration for both healthy kidneys and renal masses with various subtypes and histopathological grades.
期刊介绍：
The Journal of Magnetic Resonance Imaging (JMRI) is an international journal devoted to the timely publication of basic and clinical research, educational and review articles, and other information related to the diagnostic applications of magnetic resonance.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
