Fei Peng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Fei Peng has a total of 58 co-authors. Here are the top co-authors by number of shared publications:
- Heng Zhao
Department of Radiology, The First Affiliated Hospital of the University of South China, Hengyang, China
4 shared publications Limin Liu
Department of Ultrasound, The Second Affiliated Hospital, Hengyang Medical School, University of South China, Jiefang Road No. 35, Hengyang, 421001, Hunan, China
4 shared publications- Ying-Kun Guo
Department of Radiology, Key Laboratory of Obstetric & Gynecologic and Pediatric Diseases and Birth Defects of Ministry of Education, West China Second University Hospital, Sichuan University, Chengdu, China
3 shared publications Chunyu Wu
Department of Radiology, The First Affiliated Hospital, Hengyang Medical School, University of South China
2 shared publicationsFang Wang
Department of Radiology, The First Affiliated Hospital, Hengyang Medical School, University of South China, Chuanshan Road No. 69, Hengyang, 421001, Hunan, China
2 shared publications- Huayan Xu
Department of Radiology, West China Second University Hospital, Sichuan University, Chengdu, China
2 shared publications
Latest Publications
- Open Access
- 16 June 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
M²B-Net: a lightweight multi-scale multi-attention boundary-aware network for liver tumor segmentation from CT images
Liver tumor segmentation from CT images remains challenging due to large variations in lesion scale, blurred boundaries, low tissue contrast, and the high computational cost of existing deep learning models. This study aims to develop a lightweight yet accurate segmentation network suitable for clinical deployment. We propose a Multi-scale Multi-attention Boundary-aware Network (M²B-Net) based on a U-shaped encoder-decoder architecture, integrating four modules: Multi-Dimensional Spatial-Location Attention (MDSLA) for feature enhancement across encoder scales, Manhattan Self-Attention (MaSA) for global dependency modeling, Multi-Scale Feature Refinement Module (MSFRM) for cross-scale feature alignment, and Boundary-Convolution Attention Module (BCAM) for edge detail enhancement. A weighted composite loss function combining cross-entropy and Dice loss is used. Experiments were conducted on the public LiTS dataset. M²B-Net achieved a Dice coefficient of 0.77 ± 0.13, volumetric overlap error of 0.35 ± 0.15, average symmetric surface distance of 2.99 ± 2.03 mm, and maximum symmetric surface distance of 5.20 ± 3.07 mm. The model contains 21.7 million parameters and 28.5 GFLOPs, with a training time of 4.1 h and testing time of 37 s per case. Ablation and comparative experiments confirmed the contribution of each module and showed superior performance over SegNet, TD-Net, SBC-Net, and RIS-UNet. M²B-Net effectively addresses multi-scale adaptation, weak boundary capture, and global context modeling in liver tumor segmentation while maintaining a lightweight architecture, demonstrating strong potential for rapid and accurate clinical deployment in resource-constrained settings.
- 16 June 2026
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessM²B-Net: a lightweight multi-scale multi-attention boundary-aware network for liver tumor segmentation from CT images
Liver tumor segmentation from CT images remains challenging due to large variations in lesion scale, blurred boundaries, low tissue contrast, and the high computational cost of existing deep learning models. This study aims to develop a lightweight yet accurate segmentation network suitable for clinical deployment. We propose a Multi-scale Multi-attention Boundary-aware Network (M²B-Net) based on a U-shaped encoder-decoder architecture, integrating four modules: Multi-Dimensional Spatial-Location Attention (MDSLA) for feature enhancement across encoder scales, Manhattan Self-Attention (MaSA) for global dependency modeling, Multi-Scale Feature Refinement Module (MSFRM) for cross-scale feature alignment, and Boundary-Convolution Attention Module (BCAM) for edge detail enhancement. A weighted composite loss function combining cross-entropy and Dice loss is used. Experiments were conducted on the public LiTS dataset. M²B-Net achieved a Dice coefficient of 0.77 ± 0.13, volumetric overlap error of 0.35 ± 0.15, average symmetric surface distance of 2.99 ± 2.03 mm, and maximum symmetric surface distance of 5.20 ± 3.07 mm. The model contains 21.7 million parameters and 28.5 GFLOPs, with a training time of 4.1 h and testing time of 37 s per case. Ablation and comparative experiments confirmed the contribution of each module and showed superior performance over SegNet, TD-Net, SBC-Net, and RIS-UNet. M²B-Net effectively addresses multi-scale adaptation, weak boundary capture, and global context modeling in liver tumor segmentation while maintaining a lightweight architecture, demonstrating strong potential for rapid and accurate clinical deployment in resource-constrained settings.
- Open Access
- 26 March 2026
- Published by Frontiers Media SA in Frontiers in Endocrinology
- Vol. 17, 1803735
Association between abdominal muscle mass measured by dual-source computed tomography and coronary artery calcification in patients with Type 2 diabetes mellitus
Objective: To describe the correlation between abdominal muscle mass and CAC in T2DM patients using DSCT, and to determine the preferred muscle imaging indicators in diagnosing and predicting positive CAC patients. Materials and methods: 108 T2DM patients were included (57.0 ± 10.9 years old). We acquired both CAC score and abdominal-chest DSCT data. DSCT measurements (intermuscular adipose tissue [IMAT], total abdominal muscle area [TAMA], normal attenuation muscle area [NAMA], low attenuation muscle area [LAMA], fat fraction [FF], LAMA/body mass index [BMI], NAMA/BMI, and NAMA/TAMA index) of 7 muscles (psoas major [PM], quadratus lumborum [QL], erector spinae [ES], rectus abdominis [RA], transversus abdominis [TA], oblique abdominals [OA], abdominal core muscles [ACM]) on the level of the third lumbar vertebra were conducted. T2DM patients were divided into four subgroups based on CAC score: negative controls (NCs) (0 score), mild (300 score). The following statistical analyses were conducted: intergroup differences were compared using the Mann-Whitney test, diagnostic performance was evaluated via receiver operating characteristic (ROC) curve analysis, associations were assessed with Spearman correlation, and predictors were identified through logistic regression. A P -value P P P < 0.01), and NAMA/BMI in the ACM emerged as an independent risk factor for positive CAC (odds ratio=0.63, P = 0.02). Conclusion: Abdominal muscle mass measured by DSCT was significantly associated with CAC score in T2DM patients. Overall, the NAMA/BMI showed optimal diagnostic value for CAC across severity levels. Moreover, NAMA/BMI in the ACM may serve as a predictive biomarker for positive CAC.
- 26 March 2026
- Published by Frontiers Media SA in Frontiers in Endocrinology
- Vol. 17, 1803735
Open AccessAssociation between abdominal muscle mass measured by dual-source computed tomography and coronary artery calcification in patients with Type 2 diabetes mellitus
Objective: To describe the correlation between abdominal muscle mass and CAC in T2DM patients using DSCT, and to determine the preferred muscle imaging indicators in diagnosing and predicting positive CAC patients. Materials and methods: 108 T2DM patients were included (57.0 ± 10.9 years old). We acquired both CAC score and abdominal-chest DSCT data. DSCT measurements (intermuscular adipose tissue [IMAT], total abdominal muscle area [TAMA], normal attenuation muscle area [NAMA], low attenuation muscle area [LAMA], fat fraction [FF], LAMA/body mass index [BMI], NAMA/BMI, and NAMA/TAMA index) of 7 muscles (psoas major [PM], quadratus lumborum [QL], erector spinae [ES], rectus abdominis [RA], transversus abdominis [TA], oblique abdominals [OA], abdominal core muscles [ACM]) on the level of the third lumbar vertebra were conducted. T2DM patients were divided into four subgroups based on CAC score: negative controls (NCs) (0 score), mild (300 score). The following statistical analyses were conducted: intergroup differences were compared using the Mann-Whitney test, diagnostic performance was evaluated via receiver operating characteristic (ROC) curve analysis, associations were assessed with Spearman correlation, and predictors were identified through logistic regression. A P -value P P P < 0.01), and NAMA/BMI in the ACM emerged as an independent risk factor for positive CAC (odds ratio=0.63, P = 0.02). Conclusion: Abdominal muscle mass measured by DSCT was significantly associated with CAC score in T2DM patients. Overall, the NAMA/BMI showed optimal diagnostic value for CAC across severity levels. Moreover, NAMA/BMI in the ACM may serve as a predictive biomarker for positive CAC.
- Open Access
- 24 February 2026
- Published by Frontiers Media SA in Frontiers in Nutrition
- Vol. 13, 1712633
The pectoralis major index: a novel computed tomography biomarker for low thoracic muscle mass and eastern cooperative oncology group performance status in lung cancer patients
Background: Chest computed tomography (CT) seldom covers the third lumbar (L3) vertebral level, the standard landmark for assessing total body muscle mass. As muscle measurements at the fourth thoracic (T4) level show high concordance with those at L3, the T4 level may serve as a viable alternative. We evaluated the discriminatory performance of T4 musculature for lung cancer–related low thoracic muscle mass (LTMM) and its association with Eastern Cooperative Oncology Group (ECOG) performance status. Methods: We retrospectively included 289 inpatients with newly diagnosed lung cancer who underwent chest CT within 3 months. At T4, the pectoralis major, pectoralis minor, and chest–wall muscle group were segmented to derive cross–sectional area (CSA), density, and height–normalized indices (cm 2 /m 2 ). Low thoracic muscle mass (LTMM) was defined by sex–specific 25th–percentile thoracic 4th vertebra level muscle index (T4MI) cutoffs (40.78 cm 2 /m 2 for men, 33.15 cm 2 /m 2 for women). Patients were stratified by sex and clinical stage (I–II early–stage vs. III–IV advanced–stage). ROC analyses compared discriminatory performance; logistic regression tested associations with poor ECOG performance status. Results: Compared with non–low thoracic muscle mass (non–LTMM), low thoracic muscle mass (LTMM) group showed smaller CSA, lower density, and reduced indices. Across muscle groups, indices outperformed CSA and density. The pectoralis major index achieved the highest AUC in the overall cohort (AUC = 0.833) and reached excellent discrimination in the overall male cohort (AUC = 0.922). Notably, stage–stratified analyses showed consistently superior discriminatory performance for the pectoralis major index in the male population (early–stage AUC = 0.917, advanced–stage AUC = 0.912). In addition, the pectoralis major index was independently associated with poor ECOG performance status (odds ratio = 0.932, p = 0.009). Conclusion: T4-level muscle metrics showed good internal discrimination for lung cancer–related low thoracic muscle mass across sex and disease stage, with the pectoralis major index achieving the highest overall performance. Moreover, the pectoralis major index may be a potential imaging biomarker associated with ECOG performance status.
- 24 February 2026
- Published by Frontiers Media SA in Frontiers in Nutrition
- Vol. 13, 1712633
Open AccessThe pectoralis major index: a novel computed tomography biomarker for low thoracic muscle mass and eastern cooperative oncology group performance status in lung cancer patients
Background: Chest computed tomography (CT) seldom covers the third lumbar (L3) vertebral level, the standard landmark for assessing total body muscle mass. As muscle measurements at the fourth thoracic (T4) level show high concordance with those at L3, the T4 level may serve as a viable alternative. We evaluated the discriminatory performance of T4 musculature for lung cancer–related low thoracic muscle mass (LTMM) and its association with Eastern Cooperative Oncology Group (ECOG) performance status. Methods: We retrospectively included 289 inpatients with newly diagnosed lung cancer who underwent chest CT within 3 months. At T4, the pectoralis major, pectoralis minor, and chest–wall muscle group were segmented to derive cross–sectional area (CSA), density, and height–normalized indices (cm 2 /m 2 ). Low thoracic muscle mass (LTMM) was defined by sex–specific 25th–percentile thoracic 4th vertebra level muscle index (T4MI) cutoffs (40.78 cm 2 /m 2 for men, 33.15 cm 2 /m 2 for women). Patients were stratified by sex and clinical stage (I–II early–stage vs. III–IV advanced–stage). ROC analyses compared discriminatory performance; logistic regression tested associations with poor ECOG performance status. Results: Compared with non–low thoracic muscle mass (non–LTMM), low thoracic muscle mass (LTMM) group showed smaller CSA, lower density, and reduced indices. Across muscle groups, indices outperformed CSA and density. The pectoralis major index achieved the highest AUC in the overall cohort (AUC = 0.833) and reached excellent discrimination in the overall male cohort (AUC = 0.922). Notably, stage–stratified analyses showed consistently superior discriminatory performance for the pectoralis major index in the male population (early–stage AUC = 0.917, advanced–stage AUC = 0.912). In addition, the pectoralis major index was independently associated with poor ECOG performance status (odds ratio = 0.932, p = 0.009). Conclusion: T4-level muscle metrics showed good internal discrimination for lung cancer–related low thoracic muscle mass across sex and disease stage, with the pectoralis major index achieving the highest overall performance. Moreover, the pectoralis major index may be a potential imaging biomarker associated with ECOG performance status.
- 4 March 2025
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 62 (2) , 457-467
Multi‐Parametric Quantitative MRI in the Early Differential Diagnosis of Ambulatory Children With Duchenne Muscular Dystrophy and Becker Muscular Dystrophy
Background: Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD), the most common dystrophinopathies, require distinct treatments. Corticosteroids are effective for DMD but less so for BMD. Early diagnosis can improve DMD outcomes and prevent BMD overtreatment. However, differentiating them in younger children with mild symptoms is challenging due to overlapping features.Purpose: To evaluate the performance of multi‐parametric quantitative MRI (qMRI) in early differentiation between BMD and DMD, particularly among younger patients in the mild disease stage.Study Type: Prospective.Subjects: 121 DMD males (mean age 8.5 ± 1.6 years), 28 BMD males (9.7 ± 3.3 years), and 26 male healthy controls (HCs) (9.4 ± 2.7 years).Field Strength/Sequence: 3.0T/3‐point Dixon (fast‐spin‐echo), T1‐mapping (modified‐Look‐Locker‐inversion‐recovery), and T2‐mapping (balance‐steady‐state‐free‐precession).Assessment: qMRI measurements (fat fraction [FF], T1, and T2) in 18 pelvic and thigh muscles were conducted. A linearized NorthStar ambulatory assessment (NSAA) score was used to evaluate the function status, with a mild functional decline stage defined as a score of 76–100.Statistical Tests: Mann–Whitney test, Kruskal–Wallis test, and Receiver operating characteristic curves. A P‐value < 0.05 was considered statistically significant.Results: In all subject groups, DMD exhibits significantly higher FF and T2 and lower T1 compared to BMD. The max differences in mean FF, T1, and T2 was 45.29%, 543 ms, and 31.0 ms, respectively. Overall, the area under curve (AUC) values for FF surpassed those for T2 and T1 in the majority of muscles. In the mild subgroup below 10 years old, DMD had significantly higher FF and T1 than BMD. The combination of FF and T1 in gluteus medius, rectus femoris, vastus lateralis, vastus intermedius, vastus medialis, and adductor magnus achieved higher AUCs (0.816–0.957) than FF (0.773–0.862) or T1 (0.701–0.819) in differentiating the two subgroups.Data Conclusion: Multi‐parametric qMRI demonstrates potential as an effective tool for early differentiation between DMD and BMD.Evidence Level: 2Technical Efficacy: Stage 3
- 4 March 2025
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 62 (2) , 457-467
Multi‐Parametric Quantitative MRI in the Early Differential Diagnosis of Ambulatory Children With Duchenne Muscular Dystrophy and Becker Muscular Dystrophy
Background: Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD), the most common dystrophinopathies, require distinct treatments. Corticosteroids are effective for DMD but less so for BMD. Early diagnosis can improve DMD outcomes and prevent BMD overtreatment. However, differentiating them in younger children with mild symptoms is challenging due to overlapping features.Purpose: To evaluate the performance of multi‐parametric quantitative MRI (qMRI) in early differentiation between BMD and DMD, particularly among younger patients in the mild disease stage.Study Type: Prospective.Subjects: 121 DMD males (mean age 8.5 ± 1.6 years), 28 BMD males (9.7 ± 3.3 years), and 26 male healthy controls (HCs) (9.4 ± 2.7 years).Field Strength/Sequence: 3.0T/3‐point Dixon (fast‐spin‐echo), T1‐mapping (modified‐Look‐Locker‐inversion‐recovery), and T2‐mapping (balance‐steady‐state‐free‐precession).Assessment: qMRI measurements (fat fraction [FF], T1, and T2) in 18 pelvic and thigh muscles were conducted. A linearized NorthStar ambulatory assessment (NSAA) score was used to evaluate the function status, with a mild functional decline stage defined as a score of 76–100.Statistical Tests: Mann–Whitney test, Kruskal–Wallis test, and Receiver operating characteristic curves. A P‐value < 0.05 was considered statistically significant.Results: In all subject groups, DMD exhibits significantly higher FF and T2 and lower T1 compared to BMD. The max differences in mean FF, T1, and T2 was 45.29%, 543 ms, and 31.0 ms, respectively. Overall, the area under curve (AUC) values for FF surpassed those for T2 and T1 in the majority of muscles. In the mild subgroup below 10 years old, DMD had significantly higher FF and T1 than BMD. The combination of FF and T1 in gluteus medius, rectus femoris, vastus lateralis, vastus intermedius, vastus medialis, and adductor magnus achieved higher AUCs (0.816–0.957) than FF (0.773–0.862) or T1 (0.701–0.819) in differentiating the two subgroups.Data Conclusion: Multi‐parametric qMRI demonstrates potential as an effective tool for early differentiation between DMD and BMD.Evidence Level: 2Technical Efficacy: Stage 3
- Open Access
- 27 January 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-10
Dual-energy CT-derived virtual noncalcium imaging to assess bone marrow lesions in patients with knee osteoarthritis
To determine the diagnostic performance of dual-energy CT (DECT) virtual noncalcium (VNCa) technique in the detection of bone marrow lesions (BMLs) in knee osteoarthritis, and further analyze the correlation between the severity of BMLs on VNCa image and the degree of knee pain. 23 consecutive patients with clinically diagnosed knee osteoarthritis were underwent DECT and 3.0T MRI between August 2017 and November 2018. Evaluation of two pain assessment scales (WOMAC and KOOS) were collected. VNCa images and MRI were independently scored by three readers using a four-level scoring system over 15 anatomical subregions in each knee joint. Spearman correlation coefficient was used for total BML scores on DECT and MRI correlation with WOMAC and KOOS. Specificity, Sensitivity, NPV and PPV of reader 1 and reader 2 were 99.4%/99.2%, 89.4%/87.2%, 98.6%/98.3% and 95.5%/93.2%. A cutoff value of − 41.5 HU/− 46.5 HU provided sensitivities of 93.2%/90.9% and specificities of 100.0%/93.9% for diagnosing BMLs with AUC of 0.970/0.996. A stronger correlation was observed between the WOMAC and total BML score compared to the KOOS. DECT possessed excellent diagnostic performance in the detection of BMLs in knee osteoarthritis. And the pain degree increased with the severity of BMLs on VNCa images.
- 27 January 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-10
Open AccessDual-energy CT-derived virtual noncalcium imaging to assess bone marrow lesions in patients with knee osteoarthritis
To determine the diagnostic performance of dual-energy CT (DECT) virtual noncalcium (VNCa) technique in the detection of bone marrow lesions (BMLs) in knee osteoarthritis, and further analyze the correlation between the severity of BMLs on VNCa image and the degree of knee pain. 23 consecutive patients with clinically diagnosed knee osteoarthritis were underwent DECT and 3.0T MRI between August 2017 and November 2018. Evaluation of two pain assessment scales (WOMAC and KOOS) were collected. VNCa images and MRI were independently scored by three readers using a four-level scoring system over 15 anatomical subregions in each knee joint. Spearman correlation coefficient was used for total BML scores on DECT and MRI correlation with WOMAC and KOOS. Specificity, Sensitivity, NPV and PPV of reader 1 and reader 2 were 99.4%/99.2%, 89.4%/87.2%, 98.6%/98.3% and 95.5%/93.2%. A cutoff value of − 41.5 HU/− 46.5 HU provided sensitivities of 93.2%/90.9% and specificities of 100.0%/93.9% for diagnosing BMLs with AUC of 0.970/0.996. A stronger correlation was observed between the WOMAC and total BML score compared to the KOOS. DECT possessed excellent diagnostic performance in the detection of BMLs in knee osteoarthritis. And the pain degree increased with the severity of BMLs on VNCa images.
- 2 December 2024
- Published by Springer Nature in Pediatric Radiology
- Vol. 55 (1) , 136-150
Multi-parameter quantitative magnetic resonance imaging for early detecting skeletal muscle involvement and predicting functional decline in children with Becker muscular dystrophy
Background The extreme clinical heterogeneity of children with Becker muscular dystrophy significantly poses a great challenge to accurately assess their disease status. Objective To detect skeletal muscle involvement in children with Becker muscular dystrophy using multiple-parameter quantitative magnetic resonance imaging (qMRI), and to determine the preferred muscle site and qMRI biomarker. Materials and methods Fat fraction, T1, and T2 measurements were conducted in Becker muscular dystrophy (n=29) and healthy controls (n=23). North Star Ambulatory Assessment (NSAA) was performed in Becker muscular dystrophy. Group differences were compared by using the Mann–Whitney or Kruskal–Wallis test or a linear mixed-effect model. Receiver operating characteristic analysis with area under curve (AUC) was used to compare the diagnostic performance. Logistic regression was used to identify the predictor of functional decline. Results Both fat fraction and T2 were effective in detecting muscle involvement across different functional stages that were categorized by NSAA, with fat fraction in gluteus maximus demonstrating the most superior diagnostic performance (AUC range, 0.85–0.98). The combination of T2 and T1 enables a good diagnosis of no abnormal fat–infiltrated muscles (AUC=0.82). Overall, fat fraction in gluteus maximus exhibited the strongest negative correlation with the NSAA score (r=-0.69, PP=0.02). Conclusion Multi-parametric qMRI demonstrate effective capabilities for early detection of muscle involvement, with gluteus maximus being the preferred muscle site. Fat fraction in gluteus maximus may serve as a potential biomarker for predicting functional decline. Graphical Abstract
- 2 December 2024
- Published by Springer Nature in Pediatric Radiology
- Vol. 55 (1) , 136-150
Multi-parameter quantitative magnetic resonance imaging for early detecting skeletal muscle involvement and predicting functional decline in children with Becker muscular dystrophy
Background The extreme clinical heterogeneity of children with Becker muscular dystrophy significantly poses a great challenge to accurately assess their disease status. Objective To detect skeletal muscle involvement in children with Becker muscular dystrophy using multiple-parameter quantitative magnetic resonance imaging (qMRI), and to determine the preferred muscle site and qMRI biomarker. Materials and methods Fat fraction, T1, and T2 measurements were conducted in Becker muscular dystrophy (n=29) and healthy controls (n=23). North Star Ambulatory Assessment (NSAA) was performed in Becker muscular dystrophy. Group differences were compared by using the Mann–Whitney or Kruskal–Wallis test or a linear mixed-effect model. Receiver operating characteristic analysis with area under curve (AUC) was used to compare the diagnostic performance. Logistic regression was used to identify the predictor of functional decline. Results Both fat fraction and T2 were effective in detecting muscle involvement across different functional stages that were categorized by NSAA, with fat fraction in gluteus maximus demonstrating the most superior diagnostic performance (AUC range, 0.85–0.98). The combination of T2 and T1 enables a good diagnosis of no abnormal fat–infiltrated muscles (AUC=0.82). Overall, fat fraction in gluteus maximus exhibited the strongest negative correlation with the NSAA score (r=-0.69, PP=0.02). Conclusion Multi-parametric qMRI demonstrate effective capabilities for early detection of muscle involvement, with gluteus maximus being the preferred muscle site. Fat fraction in gluteus maximus may serve as a potential biomarker for predicting functional decline. Graphical Abstract
- 14 December 2023
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 60 (4) , 1402-1413
Utilization of Multi‐Parametric Quantitative Magnetic Resonance Imaging in the Early Diagnosis of Duchenne Muscular Dystrophy
Background: It is challenging to diagnose suspected Duchenne muscular dystrophy (DMD) patients in the very early stage of the disease. More evidence is needed to demonstrate the potential of quantitative MRI (qMRI) in precisely identifying patients before substantial physical decline occurs.Purpose: To assess the early diagnostic performance of multi‐parametric qMRI for DMD patients, and the ability to identify DMD patients with mild functional decline.Study Type: Prospective.Subjects: One hundred and forty DMD subjects (9.0 ± 2.2 years old), 24 male healthy controls (HCs) (9.2 ± 2.5 years old).Field Strength/Sequence: 3.0 T/3‐point Dixon, T1‐mapping, and T2‐mapping.Assessment: qMRI measurements (fat fraction [FF], T1, and T2) of 11 thigh muscles (rectus femoris [RF], vastus lateralis [VL], vastus intermedius, vastus medialis, gracilis, sartorius, adductor longus, adductor magnus [AM], semitendinosus, semimembranosus, biceps femoris long head [BFLH]) on the right side were conducted. NorthStar ambulatory assessment (NSAA) score used to evaluate the function of DMD patients and divided them into three subgroups: mild (76–100 score), moderate (51–75 score), and severe (0–50 score) functional decline.Statistical Tests: Independent t‐test, ANOVA analysis, and receiver operating characteristic (ROC) curves. A P‐value <0.05 was considered statistically significant.Results: Compared with HCs, FF and T2 were significantly higher in the group of all DMD patients, while T1 was significantly lower. The combination of T1 and T2 in RF, VL, AM, and BFLH achieved excellent area under curve (AUCs) (0.967–0.992) in differentiating five DMD patients without abnormal fat infiltration from HCs. Overall, T2 reached higher AUCs than FF and T1 in distinguishing DMD with mild functional decline from HCs, whereas FF achieved higher AUCs than T1 and T2 in distinguishing three DMD subgroups with functional decline.Data Conclusion: Multi‐parametric qMRI demonstrate effective diagnostic capabilities for DMD patients in the early stage of the disease, and can identify patients with mild physical decline.Level of Evidence: 2Technical Efficacy: Stage 3
- 14 December 2023
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 60 (4) , 1402-1413
Utilization of Multi‐Parametric Quantitative Magnetic Resonance Imaging in the Early Diagnosis of Duchenne Muscular Dystrophy
Background: It is challenging to diagnose suspected Duchenne muscular dystrophy (DMD) patients in the very early stage of the disease. More evidence is needed to demonstrate the potential of quantitative MRI (qMRI) in precisely identifying patients before substantial physical decline occurs.Purpose: To assess the early diagnostic performance of multi‐parametric qMRI for DMD patients, and the ability to identify DMD patients with mild functional decline.Study Type: Prospective.Subjects: One hundred and forty DMD subjects (9.0 ± 2.2 years old), 24 male healthy controls (HCs) (9.2 ± 2.5 years old).Field Strength/Sequence: 3.0 T/3‐point Dixon, T1‐mapping, and T2‐mapping.Assessment: qMRI measurements (fat fraction [FF], T1, and T2) of 11 thigh muscles (rectus femoris [RF], vastus lateralis [VL], vastus intermedius, vastus medialis, gracilis, sartorius, adductor longus, adductor magnus [AM], semitendinosus, semimembranosus, biceps femoris long head [BFLH]) on the right side were conducted. NorthStar ambulatory assessment (NSAA) score used to evaluate the function of DMD patients and divided them into three subgroups: mild (76–100 score), moderate (51–75 score), and severe (0–50 score) functional decline.Statistical Tests: Independent t‐test, ANOVA analysis, and receiver operating characteristic (ROC) curves. A P‐value <0.05 was considered statistically significant.Results: Compared with HCs, FF and T2 were significantly higher in the group of all DMD patients, while T1 was significantly lower. The combination of T1 and T2 in RF, VL, AM, and BFLH achieved excellent area under curve (AUCs) (0.967–0.992) in differentiating five DMD patients without abnormal fat infiltration from HCs. Overall, T2 reached higher AUCs than FF and T1 in distinguishing DMD with mild functional decline from HCs, whereas FF achieved higher AUCs than T1 and T2 in distinguishing three DMD subgroups with functional decline.Data Conclusion: Multi‐parametric qMRI demonstrate effective diagnostic capabilities for DMD patients in the early stage of the disease, and can identify patients with mild physical decline.Level of Evidence: 2Technical Efficacy: Stage 3
- Open Access
- 18 November 2020
- Published by Pakistan Medical Association in Journal of the Pakistan Medical Association
- Vol. 71 (2) , 1-17
Anatomic study of the carpal tunnel in adults using monoenergetic reconstruction of dual-energy computed tomography
Purpose: To seek optimal keV settings for imaging carpal tunnel in adults by dual-energy computed tomography (DECT) monoenergetic technique; to describe anatomic characteristics of carpal tunnel and to observe correlation between carpal bony and soft tissue structures. Methods: DECT images of 20 wrists (11 left and 9 right wrists; 14 men, mean age 26.93 ± 1.38 years, range 23 to 28, and 6 women, mean age 24.17 ± 0.98 years, range 23 to 26) were evaluated. Monoenergetic images were reconstructed at 42, 62, 82, 102, 122, and 142 keV. Image quality was assessed along a 5-point Likert scale, and the highest-quality images were chosen for quantitative analysis. Two musculoskeletal radiologists performed both analyses independently. Results: The optimal energy spectrum with the best contrast-to-noise ratio (CNR) for monoenergetic images were at 62 keV (19 wrists, 95%) and 61 keV (1 wrist, 5%). There was substantial interobserver agreement between the readers in the 5-point Likert scale analysis of image quality (? = 0.793). Bland-Altman plots also indicated good agreement between observers in quantitative analysis. Intra-category 1 and 2 correlation was mostly discovered at hamate hook level and middle level of pisiform (P < 0.05), while bony and soft tissue structures partly reached correlation (P < 0.05). Conclusion: The optimal energy spectrum for monoenergetic DECT imaging of carpal tunnel structures was 62 keV. DECT monoenergetic imaging could predict changes in soft tissue structures and demonstrate carpal tunnel anatomic structures. Keywords: DECT, monoenergetic, keV, carpal tunnel, Continuous...
- 18 November 2020
- Published by Pakistan Medical Association in Journal of the Pakistan Medical Association
- Vol. 71 (2) , 1-17
Open AccessAnatomic study of the carpal tunnel in adults using monoenergetic reconstruction of dual-energy computed tomography
Purpose: To seek optimal keV settings for imaging carpal tunnel in adults by dual-energy computed tomography (DECT) monoenergetic technique; to describe anatomic characteristics of carpal tunnel and to observe correlation between carpal bony and soft tissue structures. Methods: DECT images of 20 wrists (11 left and 9 right wrists; 14 men, mean age 26.93 ± 1.38 years, range 23 to 28, and 6 women, mean age 24.17 ± 0.98 years, range 23 to 26) were evaluated. Monoenergetic images were reconstructed at 42, 62, 82, 102, 122, and 142 keV. Image quality was assessed along a 5-point Likert scale, and the highest-quality images were chosen for quantitative analysis. Two musculoskeletal radiologists performed both analyses independently. Results: The optimal energy spectrum with the best contrast-to-noise ratio (CNR) for monoenergetic images were at 62 keV (19 wrists, 95%) and 61 keV (1 wrist, 5%). There was substantial interobserver agreement between the readers in the 5-point Likert scale analysis of image quality (? = 0.793). Bland-Altman plots also indicated good agreement between observers in quantitative analysis. Intra-category 1 and 2 correlation was mostly discovered at hamate hook level and middle level of pisiform (P < 0.05), while bony and soft tissue structures partly reached correlation (P < 0.05). Conclusion: The optimal energy spectrum for monoenergetic DECT imaging of carpal tunnel structures was 62 keV. DECT monoenergetic imaging could predict changes in soft tissue structures and demonstrate carpal tunnel anatomic structures. Keywords: DECT, monoenergetic, keV, carpal tunnel, Continuous...