Journal of Magnetic Resonance Imaging最新文献

{"title":"Application of Large Language Models in TN Staging and Treatment Response Evaluation for Patients With Nasopharyngeal Carcinoma: A Comparative Performance Analysis of ChatGPT-4o-Latest and DeepSeek-V3-0324.","authors":"Yifan Yang, Fan Yang, Shengsheng Xiao, Kaiqi Hou, Kexin Chen, Zaiyi Liu, Changhong Liang, Xiaobo Chen, Guangyi Wang","doi":"10.1002/jmri.70140","DOIUrl":"https://doi.org/10.1002/jmri.70140","url":null,"abstract":"<p><strong>Background: </strong>Accurate tumor staging and treatment response evaluation (TRE) are critical for nasopharyngeal carcinoma (NPC) clinical decisions. Conventional methods relying on manual imaging analysis are expertise-dependent, time-consuming, and prone to inter-observer variability and errors.</p><p><strong>Purpose: </strong>To assess the performance of two large language models (LLMs): ChatGPT-4o-latest and DeepSeek-V3-0324 in automating T, N staging and TRE for NPC patients.</p><p><strong>Study type: </strong>Retrospective.</p><p><strong>Population: </strong>Three hundred seven NPC patients from three centers (mean age: 45.5 ± 11.3 years; 216 men, 91 women).</p><p><strong>Field strength/sequence: </strong>All imaging was conducted using 3.0T or 1.5T scanners. The imaging sequence included axial T1-weighted fast spin-echo, T2-weighted fast spin-echo, T2-weighted fat-suppressed spin-echo, and Contrast-Enhanced T1-weighted fast spin-echo.</p><p><strong>Assessment: </strong>Two radiologists established the reference standards for TN staging at baseline and for TRE at two time points: post-induction chemotherapy (TRE-1) and post-concurrent chemoradiotherapy (TRE-2), based on the 9th version of AJCC/UICC guidelines and the RECIST1.1 criteria. LLMs were via few-shot chain-of-thought prompting and tested on 277 patients with 831 reports. Additionally, four radiologists independently assessed 68 cases both with and without the assistance of LLMs and compared the performance and efficiency in both conditions.</p><p><strong>Statistical tests: </strong>McNemar-Bowker test, Wilcoxon signed-rank test. p < 0.05 was considered statistically significant.</p><p><strong>Results: </strong>DeepSeek-V3-0324 significantly outperformed GPT-4o-latest in TRE-1 staging (96.5% vs. 82.9%, p < 0.001). For T staging (95.3% vs. 93.5%, p = 0.24), N staging (93.8% vs. 89.6%, p = 0.265), and TRE-2 (94.9% vs. 93.2%, p = 0.556), the accuracy between DeepSeek-V3-0324 and ChatGPT-4o-latest showed no significant difference. DeepSeek-V3-0324 also showed stronger agreement with expert annotation (κ = 0.85-0.90), compared to ChatGPT-4o-latest (κ = 0.49-0.86). Significant improvements in time efficiency were observed across all radiologists with LLM assistance (p < 0.001).</p><p><strong>Data conclusion: </strong>LLMs, particularly DeepSeek-V3-0324, can automate NPC TN staging and TRE with high accuracy, enhancing clinical efficiency. LLMs integration may improve diagnostic consistency, especially for junior clinicians.</p><p><strong>Level of evidence: 3: </strong></p><p><strong>Technical efficacy: </strong>Stage 4.</p>","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145225442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Editorial for \"The Sense of Smell (SoS) Atlas: Its Creation and First Application to Investigate COVID-19 Related Anosmia With a Comprehensive Quantitative MRI Protocol\".","authors":"Adam Dmytriw, Fulvio Zaccagna, James T Grist","doi":"10.1002/jmri.70129","DOIUrl":"https://doi.org/10.1002/jmri.70129","url":null,"abstract":"","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145225400","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Quantitative Confounder Analysis of Electrocardiogram Signals in Cardiac Magnetic Resonance at 1.5, 3 and 7 T-Assessing Standardized Electrode Positions and Sequence Types-Towards Quality Assurance.","authors":"Richard Hickstein, Stephanie Wiesemann, Darian Viezzer, Denise Kleindienst, Teodora Chitiboi, Bogdan Andrei Gheorghita, Jens Wetzl, Thomas Hadler, Sebastian Dietrich, Sebastian Schmitter, Jeanette Schulz-Menger","doi":"10.1002/jmri.70130","DOIUrl":"https://doi.org/10.1002/jmri.70130","url":null,"abstract":"<p><strong>Background: </strong>The electrocardiogram (ECG) used for gating in cardiac MRI may be compromised by multiple confounders inside the scanner bore.</p><p><strong>Purpose: </strong>To quantify the influence of magnetic field strengths (1.5 T/3 T/7 T), standardized electrode positions, and imaging sequences on ECG signals used for gating.</p><p><strong>Study type: </strong>Prospective.</p><p><strong>Population: </strong>Sixteen healthy volunteers (eight male; mean age 26.25 ± 7.67 years).</p><p><strong>Field strength/sequence: </strong>Balanced steady-state free precession cine (1.5 T/3 T), fast low-angle shot cine (7 T), and 4D flow (1.5 T/3 T/7 T) sequences.</p><p><strong>Assessment: </strong>ECG-signals were recorded during breath-hold and non-breath-hold short axis cine (sax-bh and sax-nbh, respectively) and 4D flow scans at 1.5 T/3 T/7 T. All scans were repeated with 4 standardized electrode positionings (pos1-4) at each field strength. Pos1/2 were vendor-recommended positionings for 1.5 T/3 T/7 T scans, respectively, whereas pos3/4 were alternative positionings recommended in previous studies. Similarity between confounded ECG-signals and unconfounded baseline ECG-signals was assessed by QRS-feature correlation. Cine image quality (IQ) was assessed by 3 readers (with 6, 10, and 22 years experience) on a four-point Likert scale.</p><p><strong>Statistical tests: </strong>Linear mixed models with type III tests of fixed effects (overall) and t tests with adjusted degrees of freedom (pairwise subgroup-comparisons) at significance level p < 0.05.</p><p><strong>Results: </strong>Increasing field strength resulted in significantly decreasing similarity to baseline measurements, with r values (provided with 95% confidence interval) of 1.5 T: 97% (92.6-101.3); 3 T: 91.4% (87.1-95.8); 7 T: 50.4% (46-54.9) and lower IQ: 1.5 T: 2.33 (2.12-2.55); 3 T: 1.96 (1.75-2.17); 7 T: 0.91 (0.7-1.12). Vendor-specified electrode positions pos1: 91.8% (87.2-96.5), pos2: 88.3% (83.7-92.9) showed significantly higher correlation with baseline measurements than alternative positions pos3: 67.5% (62.9-72.1) and pos4: 70.8% (66.2-75.4). The evaluated standardized sequences showed similar amounts of electrocardiogram distortion, with r values of: sax-bh: 77.3% (73-81.7); 4D: 79.3% (75-83.7), p = 0.54; sax-nbh: 82.1% (77.8-86.5), p = 0.31, but the difference between sax-bh and sax-nbh: 4.8% (2.88-6.72) was significant.</p><p><strong>Data conclusion: </strong>Increasing field strength leads to significant ECG signal distortions. Vendor-specified positions 1/2 resulted in less distorted ECG signals than alternative positions 3/4 recommended in previous publications.</p><p><strong>Level of evidence: 2: </strong></p><p><strong>Technical efficacy: </strong>Stage 5.</p>","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145225436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Association of Blood-Brain Barrier Function With Disease Activity and Cognitive Function in Systemic Lupus Erythematosus Patients.","authors":"Xueying Zhao, Zexi Yi, Yimin Cao, Cong Zhang, Zhijuan Li, Duo Gao, Sicong Wang, Zuojun Geng, Haiqing Yang, Shifeng Xiang, Yali Wang, Yue Zhang, Lixia Zhou","doi":"10.1002/jmri.70143","DOIUrl":"https://doi.org/10.1002/jmri.70143","url":null,"abstract":"<p><strong>Background: </strong>Disruption of the blood-brain barrier (BBB) is an important mechanism in the pathogenesis of neuropsychiatric systemic lupus erythematosus (NPSLE), potentially influencing disease progression and cognitive function.</p><p><strong>Purpose: </strong>To investigate voxel-wise BBB permeability differences among NPSLE, non-NPSLE, and healthy controls (HC) using dynamic contrast-enhanced MRI (DCE-MRI), and explore associations with disease activity and cognitive function.</p><p><strong>Study type: </strong>Multicenter, cross-sectional study.</p><p><strong>Subjects: </strong>22 NPSLE patients, 24 non-NPSLE patients, and 27 age- and sex-matched HC.</p><p><strong>Field strength/sequence: </strong>3.0 Tesla, GE: 3D T1-FSPGR, DCE-LAVA-Flex sequences; Philips: 3D T1-FFE, eTHRIVE sequences.</p><p><strong>Assessment: </strong>Post-processing of DCE-MRI images was conducted to calculate the volume transfer constant (K^trans) and plasma volume fraction (V_p) maps, which were transformed to Montreal Neurological Institute space. Brain regions were defined using the Anatomical Automatic Labeling (AAL) atlas. Group differences among NPSLE, non-NPSLE, and HC were assessed. Mean K^trans/V_p values from suprathreshold clusters were evaluated for associations with the Disease Activity Index (SLEDAI) and Montreal Cognitive Assessment (MoCA) scores. Mean K^trans/V_p from corresponding AAL regions was extracted for further validation of the observed correlations.</p><p><strong>Statistical tests: </strong>Voxel-wise ANOVA with Tukey's HSD test; Spearman correlation and linear regression. Threshold: Voxel-wise p < 0.001, cluster size > 20; group comparisons p < 0.05.</p><p><strong>Results: </strong>A significantly elevated K^trans was observed in the right caudate of NPSLE patients compared to non-NPSLE patients, and hippocampal K^trans was significantly higher in NPSLE than both non-NPSLE and HC. Both NPSLE and non-NPSLE patients exhibited widespread significant increases in V_p compared to HC. In the entire SLE cohort, right caudate K^trans was positively associated with SLEDAI (ρ = 0.43, R² = 0.18) and negatively associated with MoCA scores (ρ = -0.35, R² = 0.12) after adjusting for confounding factors.</p><p><strong>Data conclusion: </strong>This study provided voxel-level evidence linking regional BBB leakage to both disease activity and cognitive impairment in SLE.</p><p><strong>Evidence level: </strong>2.</p><p><strong>Technical efficacy: </strong>Stage 2.</p>","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145225413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Sense of Smell (SoS) Atlas: Its Creation and First Application to Investigate COVID-19 Related Anosmia With a Comprehensive Quantitative MRI Protocol.","authors":"Marta Gaviraghi, Eleonora Lupi, Elena Grosso, Andrea Fusari, Mattia Baiguera, Anita Monteverdi, Marco Battiston, Francesco Grussu, Baris Kanber, Ferran Prados Carrasco, Rebecca S Samson, Janine Makaronidis, Marios C Yiannakas, Michael S Zandi, Rachel L Batterham, Egidio D'Angelo, Fulvia Palesi, Claudia A M Gandini Wheeler-Kingshott","doi":"10.1002/jmri.70128","DOIUrl":"https://doi.org/10.1002/jmri.70128","url":null,"abstract":"<p><strong>Background: </strong>The loss of smell (anosmia) has been noted in numerous diseases, including COVID-19. Inflammatory and microstructural alterations are possible underlying mechanisms of anosmia in COVID-19. However, no atlas exists to study olfaction and the associated tissue property changes.</p><p><strong>Purpose: </strong>To develop the sense of smell (SoS) atlas, including gray matter regions and white matter tracts of the olfactory circuit, to investigate the underpinnings of COVID-19 related anosmia.</p><p><strong>Study type: </strong>Retrospective.</p><p><strong>Subjects: </strong>For the SoS atlas, high-resolution tractograms of 10 healthy controls (HC) of the Human Connectome Project (7 females, 22-35 years) were used. The SoS atlas was applied to 8 subjects with persistent anosmia following COVID-19 (COVID-P, 7 females, 52 ± 12 years), 19 subjects that recovered from COVID-19 anosmia (COVID-R, 14 females, 38 ± 13 years), and 17 HC (8 females, 39 ± 12 years).</p><p><strong>Field strength/sequence: </strong>3 T, 3D inversion recovery, 3D fast field echo, and spin-echo echo-planar imaging sequences.</p><p><strong>Assessment: </strong>To create the SoS atlas, regions were identified and tracts were extracted via tractography following biological constraints. MRI metrics sensitive to alterations in neuroinflammation, axonal degeneration, myelin and macromolecular density, and iron were analyzed.</p><p><strong>Statistical tests: </strong>Region-based analysis (p-value < 0.05, false discovery rate (FDR) corrected) and voxel-based analysis (p-value < 0.001 uncorrected, FDR-corrected cluster extent = 5 voxels) were performed on 15 multisequence-MRI metrics between the three groups.</p><p><strong>Results: </strong>The SoS atlas consisted of 35 regions and, after anatomical curation, the initial 506 tracts were refined to 78. Compared to HC, COVID-P presented alterations in neuroinflammation-related (mean: 41% of total alterations) and axonal degeneration-related (31%) MRI metrics, while COVID-R presented alterations of myelin-related metrics (68%). COVID-P alterations mainly affected the hindbrain (56%), while COVID-R the hindbrain (39%).</p><p><strong>Data conclusion: </strong>A novel tool, the SoS atlas, was developed to study the olfactory system and applied in combination with multisequence-MRI metrics to investigate the mechanisms of COVID-19 related anosmia.</p><p><strong>Evidence level: </strong>3.</p><p><strong>Technical efficacy: </strong>Stage 1.</p>","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145225531","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Relationships Between Lateral Ventricle Size, Cerebrospinal Fluid Dynamics, and Aqueductal Resistance in Young Healthy Adults.","authors":"Pan Liu, Jiachen Xie, Kimi Owashi, Yann Attekeble, Jean-Marc Constans, Cyrille Capel, Olivier Balédent","doi":"10.1002/jmri.70139","DOIUrl":"https://doi.org/10.1002/jmri.70139","url":null,"abstract":"<p><strong>Background: </strong>Ventricular enlargement and abnormal cerebrospinal fluid (CSF) circulation are closely associated in communicating hydrocephalus (NPH), yet their causal relationship remains unclear. Studying healthy populations may help clarify these mechanisms. Existing metrics for CSF dynamics and ventricular morphology are limited by physiological variability such as heart rate and brain size, and aqueductal resistance has been little studied in healthy cohorts.</p><p><strong>Purpose: </strong>To quantify aqueductal resistance and two ratio-based indices-the CSF stroke volume ratio of the aqueduct to the cervical region (Ratio-SV) and the lateral ventricle to total brain area ratio (Ratio-Area)-in a healthy population, and to examine their interrelationships.</p><p><strong>Study type: </strong>Prospective.</p><p><strong>Population: </strong>34 healthy young adults (17 female, 17 male; age, 25.2 ± 3.9 years old); 4 NPH patients (2 female, 2 male; age, 50-80 years old).</p><p><strong>Field strength/sequence: </strong>3 T MRI with transverse 3D T1-weighted gradient echo, sagittal 3D Balanced Fast Field Echo (BFFE; gradient echo), and 2D CINE Phase Contrast (CINE-PC; gradient echo) sequences.</p><p><strong>Assessment: </strong>Aqueductal resistance was measured on BFFE, Ratio-Area on T1-weighted images, and Ratio-SV on CINE-PC. Sex differences were also examined.</p><p><strong>Statistical tests: </strong>Wilcoxon test was used for group comparisons, and Spearman's correlation for associations among parameters, with p < 0.05 considered significant.</p><p><strong>Results: </strong>In healthy adults, mean aqueductal resistance was 72 ± 42 mPa s/mm³, Ratio-SV 6.0% ± 2.3% and Ratio-Area 4.5% ± 1.6%. Males exhibited a significantly lower Ratio-Area compared to females (3.9% ± 1.3% vs. 5.2% ± 1.7%). Ratio-SV was unaffected by the cardiac cycle. Aqueductal resistance showed a strong negative correlation with Ratio-SV (r = -0.65) but showed no significant correlation with Ratio-Area (r = -0.27). Ratio-SV and Ratio-Area were also uncorrelated (r = 0.08).</p><p><strong>Data conclusion: </strong>Ratio-based metrics provide useful parameters for evaluating both ventricular morphology and CSF dynamics by reducing the influence of physiological variations. Combined with aqueductal resistance, these baseline data in healthy adults may contribute to a better understanding of the pathophysiology of NPH.</p><p><strong>Evidence level: </strong>2.</p><p><strong>Technical efficacy: </strong>Stage 1.</p>","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145225511","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Editorial for \"Assessing the Impact of Imaging Parameters on MRI Measurement of Kidney T2\".","authors":"Véronique Fortier, Evan McNabb, Ives R Levesque","doi":"10.1002/jmri.70136","DOIUrl":"https://doi.org/10.1002/jmri.70136","url":null,"abstract":"","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145206665","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Editorial for \"Reliability of 4D Flow MRI-Derived Pulmonary Regurgitant Fraction in Repaired Tetralogy of Fallot: Impact of Measurement Location and Pulmonary Artery Geometry\".","authors":"Chenxi Shi, Zezhong Ye","doi":"10.1002/jmri.70137","DOIUrl":"https://doi.org/10.1002/jmri.70137","url":null,"abstract":"","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145206690","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Real-Time Deep-Learning Image Reconstruction and Instrument Tracking in MR-Guided Biopsies.","authors":"Constant R Noordman, Lauren P W Te Molder, Marnix C Maas, Christiaan G Overduin, Jurgen J Fütterer, Henkjan J Huisman","doi":"10.1002/jmri.70138","DOIUrl":"https://doi.org/10.1002/jmri.70138","url":null,"abstract":"<p><strong>Background: </strong>Transrectal in-bore MR-guided biopsy (MRGB) is accurate but time-consuming, limiting clinical throughput. Faster imaging could improve workflow and enable real-time instrument tracking. Existing acceleration methods often use simulated data and lack validation in clinical settings.</p><p><strong>Purpose: </strong>To accelerate MRGB by using deep learning for undersampled image reconstruction and instrument tracking, trained on multi-slice MR DICOM images and evaluated on raw k-space acquisitions.</p><p><strong>Study type: </strong>Prospective feasibility study.</p><p><strong>Population: </strong>Briefly, 1289 male patients (aged 44-87, median age 68) for model training, 8 male patients (aged 59-78, median age 65) for prospective feasibility testing.</p><p><strong>Field strength/sequence: </strong>2D Cartesian balanced steady-state free precession, 3 T.</p><p><strong>Assessment: </strong>Segmentation and reconstruction models were trained on 8464 MRGB confirmation scans containing a biopsy needle guide instrument and evaluated on 10 prospectively acquired dynamic k-space samples. Needle guide tracking accuracy was assessed using instrument tip prediction (ITP) error, computed per frame as the Euclidean distance from reference positions defined via pre- and post-movement scans. Feasibility was measured by the proportion of frames with < 5 mm error. Additional experiments tested model robustness under increasing undersampling rates.</p><p><strong>Statistical tests: </strong>In a segmentation validation experiment, a one-sample t-test tested if the mean ITP error was below 5 mm. Statistical significance was defined as p < 0.05. In the tracking experiments, the mean, standard deviation, and Wilson 95% CI of the ITP success rate were computed per sample, across undersampling levels.</p><p><strong>Results: </strong>ITP was first evaluated independently on 201 fully sampled scans, yielding an ITP error of 1.55 ± 1.01 mm (95% CI: 1.41-1.69). Tracking performance was assessed across increasing undersampling factors, achieving high ITP success rates from 97.5% ± 5.8% (68.8%-99.9%) at 8× up to 92.5% ± 10.3% (62.5%-98.9%) at 16× undersampling. Performance declined at 18×, dropping to 74.6% ± 33.6% (43.8%-91.7%).</p><p><strong>Data conclusion: </strong>Results confirm stable needle guide tip prediction accuracy and support the robustness of the reconstruction model for tracking at high undersampling.</p><p><strong>Evidence level: </strong>2.</p><p><strong>Technical efficacy: </strong>Stage 2.</p>","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145206714","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Application of Machine Learning in the Diagnosis and Prognosis of Mild Traumatic Brain Injury Using Diffusion Tensor Imaging: A Systematic Review.","authors":"Christian John A Saludar, Maryam Tayebi, Eryn Kwon, Joshua McGeown, William Schierding, Alan Wang, Justin Fernandez, Samantha Holdsworth, Vickie Shim","doi":"10.1002/jmri.70122","DOIUrl":"https://doi.org/10.1002/jmri.70122","url":null,"abstract":"<p><strong>Background: </strong>Traumatic Brain Injury (TBI) is a global health concern, with mild TBI (mTBI) being the most common form. Despite its prevalence, accurately diagnosing mTBI remains a significant challenge. While advanced neuroimaging techniques like diffusion tensor imaging (DTI) offer promise for more robust diagnosis, their clinical application is limited by inconsistent and heterogeneous post-injury findings. Recently, machine learning (ML) techniques, utilizing DTI metrics as features, have shown increasing utility in mTBI research. This approach helps identify distinct between-group features, paving the way for more precise and efficient diagnostic and prognostic tools.</p><p><strong>Purpose: </strong>This review aims to analyze studies employing ML techniques to assess changes in DTI metrics after mTBI.</p><p><strong>Study type: </strong>Systematic review.</p><p><strong>Population or subjects or phantom or specimen or animal model: </strong>We conducted a systematic review, adhering to PRISMA guidelines, on the application of ML with DTI for mTBI diagnosis and prognosis on human subjects. This review identified 36 articles.</p><p><strong>Field strength/sequence: </strong>N/A.</p><p><strong>Assessment: </strong>Study quality was assessed using the Modified QualSyst Assessment Tool.</p><p><strong>Statistical tests: </strong>N/A.</p><p><strong>Results: </strong>The review found ML techniques using DTI Metrics either alone or in combination with other modalities (i.e., structural MRI, functional MRI, clinical scores, or demographics) can effectively classify mTBI patients from controls. These approaches have also demonstrated potential in classifying mTBI patients according to the degree of recovery and symptom severity. In addition, these ML models showed strong predictive power toward cognitive scores and brain structural decline, as quantified by brain-predicted age difference.</p><p><strong>Data conclusion: </strong>Larger, externally validated studies are needed to develop robust models for the diagnosis and prognosis of mTBI, using imaging biomarkers (including DTI) in conjunction with non-imaging, on-field, or clinical data. Despite the high predictive performance of ML algorithms, the clinical application remains distant, likely due to the small sample size of studies and lack of external validation, which raises concerns about overfitting.</p><p><strong>Evidence level: </strong>5.</p><p><strong>Technical efficacy: </strong>Stage 1.</p>","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":" ","pages":""},"PeriodicalIF":3.5,"publicationDate":"2025-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145199509","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}