NODDI in clinical research

In the paper

“A very recent report (Guerreri et al., 2020) suggested that the fixed diffusivity assumption, but not the tortuosity model, is the primary cause of NODDI's incompatibility with b-tensor encoding (Lampinen et al., 2017).”

What the sources say

• ?Not checked — no full text of the sourcecites Guerreri et al., 2020
• Disputed

  Section 2.3 describes the fixed neurite diffusivity and the tortuosity model as linked NODDI constraints, not competing explanations. Sections 4.3 and 5 attribute the LTE–STE inconsistency to invalid constraints connecting neurite density with tissue mean diffusivity. The paper therefore does not support excluding tortuosity as a cause. (LLM, conf=0.84)

  Checked by a reasoning model — may be wrong.

  cites Lampinen et al., 2017 sourceExperimental — accuracy not yet measured

In the paper

“There are two main approaches to extracting information from dMRI data: signal representations and biophysical models (Jelescu and Budde, 2017;Jespersen, 2018;Novikov et al., 2018a).”

What the sources say

• Supported

  Section 1 (Models) explicitly identifies two complementary approaches to extracting information from the diffusion signal: signal representations and tissue models. Section 1.4 (Summary) calls these signal representations and biophysical models, while noting that hybrid models combine features of both. (LLM, conf=0.94)

  cites Jelescu and Budde, 2017 source
• ?Not checked — no full text of the sourcecites Jespersen, 2018 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Novikov et al., 2018a

In the paper

“Recent years have seen a marked proliferation of biophysical models with the explicit aim of clinically feasible data acquisition (single-diffusion encoding, b-values up to 2000-3000 s/mm 2 , diffusion time 20-100 ms) (Fieremans et al., 2011;Hansen et al., 2017;Kaden et al., 2016;Zhang et al., 2012).”

What the sources say

• Supported

  The Introduction identifies DKI combined with a tissue model as a clinically feasible alternative to models requiring high b-values or long scans. Methods §2.2.2 reports acquisition at b-values up to 2000 s/mm² and a diffusion time of about 50 ms; Discussion §4.2 emphasizes the model’s shorter acquisition time and clinical usefulness. This paper supports the citation as an example of a biophysical model designed for clinically feasible acquisition, rather than as evidence of proliferation on its own. (LLM, conf=0.87)

  cites Fieremans et al., 2011 source
• Supported

  The Discussion (page unavailable) describes a faster WMTI approach that reduces the number of images required and says it may be useful in clinical settings where scan time is constrained. The human-brain Methods (page unavailable) report single-diffusion encoding, b-values up to 3.0 ms/μm² (3000 s/mm²), and a diffusion time of approximately 50 ms. This paper therefore supports its role as an example of a biophysical approach designed for clinically feasible acquisition; it does not need to establish the broader trend of proliferation on its own. (LLM, conf=0.90)

  cites Hansen et al., 2017 source
• Supported

  This citation serves as an example of a biophysical model designed for clinically feasible acquisition, rather than as evidence for the number of models developed. The Abstract describes a multi-compartment model usable within clinically feasible scan times; Clinical translation specifies a roughly five-minute, two-shell protocol at b-values of 1000 and 2500 s/mm². Human data describes a Stejskal–Tanner acquisition with a 43.1 ms pulse separation, within the cited diffusion-time range. (LLM, conf=0.91)

  cites Kaden et al., 2016 source

In the paper

“In clinical research, diffusion tensor imaging (DTI) (Basser et al., 1994) has long been an important tool that has yielded many insights in a wide range of neurological and psychological disorders (for recent reviews see (Andica et al., 2019a;Goveas et al., 2015;Pasternak et al., 2018).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Andica et al., 2019a
• ?Not checked — no full text of the sourcecites Goveas et al., 2015 source
• Partly supported

  The Introduction reports DTI findings in schizophrenia, bipolar disorder, and major depressive disorder, supporting its value in psychiatric research. It says the review is limited to those conditions, so this source does not establish the claim about a wide range of neurological disorders. (LLM, conf=0.84)

  cites Pasternak et al., 2018 sourceExperimental — accuracy not yet measured

In the paper

“Similar repeatability is also reported in the cervical spinal cord (By et al., 2017;Grussu et al., 2015;Ma et al., 2018), although a greater CoV for ODI and f fw has been observed in GM, reflecting the general challenges of dMRI in the spinal cord (e.g., small structure, magnetic field inhomogeneity, cerebrospinal fluid pulsation and physiological noise).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites By et al., 2017
• Supported

  The Results (§3.2.2, page not available in our copy) report reproducibility in the cervical spinal cord and higher coefficients of variation in gray matter than white matter for both ODI (44% versus 7%) and the isotropic volume fraction, v_iso (140% versus 41%). The paper uses v_iso for the isotropic compartment, corresponding to the claim’s free-water fraction notation. (LLM, conf=0.90)

  cites Grussu et al., 2015 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Ma et al., 2018

In the paper

“f fw is less stable than NDI and ODI, with a scan-rescan CoV of 10 %-40 % (Andica et al., 2019b;Chung et al., 2016;Granberg et al., 2017).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Andica et al., 2019b
• Partly supported

  The Results (Within-Session Reproducibility; page unavailable) say that νiso, the free-water fraction, had the largest within-session CV, supporting its lower stability relative to NDI and ODI. The paper does not give a 10%–40% scan-rescan CV for νiso in the available text. The Discussion instead gives a 3.8%–40% range for between-subject NODDI CV and explicitly excludes νiso from that range. (LLM, conf=0.88)

  cites Chung et al., 2016 sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites Granberg et al., 2017 source

In the paper

“In the GM of patients with young-onset Alzheimer's disease, reductions of NDI and ODI were found in the temporal cortex and precuneus, and a smaller NDI in the cortex was associated with more severe cognitive impairment (Parker et al., 2018).”

What the sources say

• Supported

  The Results report lower NDI and ODI in the inferior and middle temporal gyri and precuneus of young-onset Alzheimer’s patients than in controls. The Discussion reports that lower NDI in those regions was associated with a lower MMSE cognition score. (LLM, conf=0.93)

  cites Parker et al., 2018 source

In the paper

“Several studies (Fukutomi et al., 2018;Genç et al., 2018) found optimizing d i, based on the fitting residual is helpful for re- vealing cortical cytoarchitecture with NODDI, though their method is still dependent on the assumption of equal diffusivities ( = d d i, e, ).”

What the sources say

• Supported

  In Materials and Methods (NODDI and DTI calculation; page unavailable), the authors select an optimized diffusivity using model fit to cortical data. The Discussion (Cortical neurite properties resemble microscopic findings and cortical classifications; page unavailable) links the resulting NODDI measures to cortical cytoarchitecture, while the Discussion (Optimization and validity of NODDI in the gray matter; page unavailable) explicitly retains the assumption of equal intracellular and extracellular axial diffusivity. The paper describes fit quality using maximum likelihood rather than the term “fitting residual,” but the substantive attribution matches. (LLM, conf=0.88)

  cites Fukutomi et al., 2018 source
• Partly supported

  The Methods section (“Analysis of imaging data in the S259 sample”; page unavailable) says the authors adjusted NODDI’s intrinsic free-diffusivity parameter to 1.1 × 10−3 mm²/s for gray matter to analyze cortical structure. It does not say they selected that value by minimizing fitting residuals or state the claimed assumption of equal intra- and extra-neurite diffusivities. (LLM, conf=0.85)

  cites Genç et al., 2018 sourceExperimental — accuracy not yet measured

In the paper

“With NODDI, (Slattery et al., 2017) reported reduced NDI in the parieto-occipital WM in both ε4+ and ε4-patients with young-onset Alzheimer's disease relative to controls, with the reduction more widespread in ε4+ patients.”

What the sources say

• Supported

  The Results section (“White-matter structural differences by APOE ε4 genotype”) states that NDI was reduced in parieto-occipital white-matter projections in both ε4− and ε4+ young-onset Alzheimer’s disease groups relative to controls, and that the reduction was more widespread in ε4+ patients. It also notes that direct comparisons between the patient groups did not reach the prescribed statistical threshold. (LLM, conf=0.95)

  cites Slattery et al., 2017 source

In the paper

“Also, even within usual multi-shell acquisition (linear tensor encoding), the optimal value of d i, seems different between neonates and adults, as well as between white matter (WM) and gray matter (GM) (Guerrero et al., 2019).”

What the sources say

• Supported

  The Discussion reports that both age and tissue type affect the optimal intrinsic parallel diffusivity in the multi-shell diffusion MRI data studied. The Conclusion states that the optimum is near the default value in adult white matter, but lower in adult gray matter and in both white and gray matter of the newborn brain. The supplied passages do not explicitly use the term “linear tensor encoding,” but they support the attributed age- and tissue-dependent finding. (LLM, conf=0.91)

  cites Guerrero et al., 2019 source

In the paper

“Though extended acquisitions have been proposed toward unconstrained estimation of the model parameters (Lampinen et al., 2020 et al., 2018b;Veraart et al., 2018), at present, the increased scan time hampers clinical application of these cutting-edge techniques.”

What the sources say

• Supported

  The citation is attached to the proposal of extended acquisitions for less-constrained parameter estimation. The Results (Sampling protocols and parameter precision; page unavailable) describe an optimized 10-shell protocol, and the Methods (In vivo acquisition and processing; page unavailable) report a 15-minute, multi-TE acquisition. The statement that scan time hampers clinical application follows the citation and is the citing authors’ framing, not a finding attributed to this paper. (LLM, conf=0.88)

  cites Lampinen et al., 2020 source
• Supported

  The Methods describe varying both b-values and TE to estimate diffusion and relaxation parameters simultaneously, and the Results discuss parameter estimation without imposed constraints. The Discussion (Optimal design) says the densely sampled, one-hour protocol has limited value in a clinical setting. (LLM, conf=0.90)

  cites Veraart et al., 2018 source

In the paper

“(Slattery et al., 2017) also demonstrated improved sensitivity of NODDI over DTI, reporting parallel reductions of NDI and ODI in some WM regions that had no observable changes in FA.”

What the sources say

• Supported

  The Results section (Fig. 3B and accompanying text) reports parallel NDI and ODI reductions in a right frontal white-matter region of patients with young-onset Alzheimer’s disease, despite no significant FA change. It explicitly describes NODDI metrics as more sensitive than FA in this setting because the opposing effects cancel in FA. (LLM, conf=0.94)

  cites Slattery et al., 2017 source

In the paper

“The validity of such constraints is a matter of debate, even in the normal brain, with several works (Lampinen et al., 2020(Lampinen et al., , 2017;;Novikov et al., 2018b) pointing out incompatibility of NODDI with extended data acquisitions.”

What the sources say

• Supported

  The Theory section identifies equal axial diffusivities and the tortuosity relation as constraints used by NODDI. The In vivo results report that estimates from the richer acquisition contradicted these commonly enforced relations and that the unconstrained model fit most regions of the healthy brain significantly better than models enforcing equal T2 values or tortuosity. Page numbers are unavailable in our copy. (LLM, conf=0.92)

  cites Lampinen et al., 2020 source
• Supported

  The Introduction says NODDI's model constraints had not been experimentally validated and proposes comparing LTE with STE data. The Discussion reports that the constraints appear invalid in the brain: STE-based neurite-density maps did not resemble LTE-based maps, and the gray-matter fit was relatively poor. This supports the attributed concern even in the healthy brain when the acquisition is extended to STE. (LLM, conf=0.94)

  cites Lampinen et al., 2017 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Novikov et al., 2018b

In the paper

“When applied to clinical studies, the use of different constraints sometimes yields results that agree in their relative trends (Huber et al., 2019;Jelescu et al., 2015;Kamiya et al., 2017), but such agreement is not always guaranteed and is not proof of validity.”

What the sources say

• Partly supported

  The Results (Biological underpinnings of reading-diffusion correlations; page unavailable) report that measures from two models, NODDI and WMTI, both correlated with reading skill in children, supporting an example of agreement in relative trends. The paper does not establish that differing model constraints generally yield agreement across clinical studies. The caution after the citations is not clearly attributed to Huber et al. (LLM, conf=0.82)

  cites Huber et al., 2019 sourceExperimental — accuracy not yet measured
• Supported

  Section 4.1.2 reports relative agreement between WMTI and NODDI in age-related trends, but substantially different quantitative estimates and poorer agreement for some parameters. Sections 5.3 and 2.3 explain that the models’ assumptions can bias those estimates, so agreement in trends does not establish validity. (LLM, conf=0.90)

  cites Jelescu et al., 2015 source
• Supported

  The Introduction says NODDI and WMTI use different constraints and that the authors compared them for consistency. The Discussion reports consistent findings despite different estimated values, while cautioning that the models have intrinsic limitations and the results require careful interpretation. The paper provides an example of agreement in trends, not a guarantee of agreement or proof of validity. (LLM, conf=0.88)

  cites Kamiya et al., 2017 source

In the paper

“In a mouse model of tau pathology, NDI in WM and hippocampus was reduced, and NDI in GM was positively correlated with the level of hyperphosphorylated tau protein (Colgan et al., 2016).”

What the sources say

• Partly supported

  The Results (Hippocampus and Corpus callosum; page unavailable) report lower NDI in the tau-model mice, supporting the reduction in hippocampus and white matter. The Results (Cortex and Hippocampus; page unavailable) report correlations between NDI and tau burden in grey matter, but do not establish the claimed positive direction; they report higher cortical NDI but lower hippocampal NDI in the model mice. (LLM, conf=0.86)

  cites Colgan et al., 2016 sourceExperimental — accuracy not yet measured

In the paper

“In temporal lobe epilepsy (TLE), the most common epilepsy in adults, reduction of NDI was found in the temporal lobes, with more prominent abnormalities in the hemisphere ipsilateral to the epileptogenic focus (Sone et al., 2018;Winston et al., 2020).”

What the sources say

• Supported

  The Introduction (page not provided) calls TLE the most common type of adult epilepsy. The Results and Conclusions (pages not provided) report reduced neurite density mainly in the ipsilateral temporal lobe in both left and right TLE, supporting the claimed hemispheric pattern. (LLM, conf=0.94)

  cites Sone et al., 2018 source
• Supported

  The Results (Intracortical grey matter; page unavailable) report reduced neurite density confined to ipsilateral mesial and basal temporal regions. The Results (Subcortical white matter; page unavailable) report bilateral neurite-density reductions, alongside stronger ipsilateral effects for a related diffusion measure. Together, these support reduced temporal-lobe neurite density with more prominent ipsilateral abnormalities, although the paper does not explicitly quantify an ipsilateral-versus-contralateral difference in neurite density. (LLM, conf=0.87)

  cites Winston et al., 2020 source

In the paper

“(Winston et al., 2014) reported that NDI maps yielded a clearer depiction of FCD than other imaging approaches (i.e., T1-/T2-weighted images and DTI maps).”

What the sources say

• Supported

  The Results state that reduced ICVF, the NODDI measure of neurite density, was more conspicuous than on conventional structural or diffusion images. The Discussion likewise says the abnormality was more readily identifiable than on other clinical or diffusion images, supporting the comparison with T1-/T2-weighted and DTI imaging. (LLM, conf=0.91)

  cites Winston et al., 2014 source

In the paper

“In postmortem human spinal cord, (Grussu et al., 2017) observed decreased ODI in MS lesions and an excellent agreement between ODI and its histological counterpart.”

What the sources say

• Supported

  The Results section reports decreased ODI in a thoracic MS lesion and a trend toward decreased ODI in lumbar lesions. Its statistical analysis reports a strong correlation between ODI and its histological counterpart, CV, in MS cases (r = 0.60; P = 0.001), supporting the described agreement. (LLM, conf=0.88)

  cites Grussu et al., 2017 source

In the paper

“The potential usefulness of NODDI for discriminating Parkinson's disease from the other Parkinsonian disorders has also been suggested (Mitchell et al., 2019), with NODDI showing greater abnormalities in progressive supranuclear palsy and multiple system atrophy than in Parkinson's disease.”

What the sources say

• Supported

  The Differentiation of Parkinsonism section reports that NODDI distinguished PD from atypical Parkinsonism (the studied MSAp and PSP groups), with an AUC of 0.945. The Introduction describes more widespread pathological changes in MSAp and PSP than in PD, and the NODDI results identify abnormalities across multiple regions in those groups. The study examined the Parkinsonian variant of multiple system atrophy, rather than all variants. (LLM, conf=0.86)

  cites Mitchell et al., 2019 source

In the paper

“They also reported that reduced NDI was seen in larger areas than was FA, suggesting a greater sensitivity of NODDI over DTI, and that reduced ODI within the precentral gyrus was correlated with disease duration, presumably representing reduced dendritic density and branching.”

What the sources say

• Supported

  The Discussion reports that, at the same statistical threshold, significantly reduced NDI covered larger areas than significantly reduced FA. It explicitly suggests that NDI may be more sensitive than FA for detecting ALS-related white-matter pathology. (LLM, conf=0.94)

  cites Broad et al., 2019 source

In the paper

“These studies generally agree that NDI in WM is reduced compared with healthy controls in MR-visible lesions, and to a lesser degree in normal-appearing WM, and that NDI reveals more prominent and widespread abnormalities than FA (Fig. 4).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Collorone et al., 2019
• Partly supported

  The Effect size section reports reduced NODDI restricted fraction in MS lesions and normal-appearing white matter, while FA differences did not survive the post-hoc test. This supports the comparison with FA, but the claim is too broad across protocols: the NODDI difference in normal-appearing white matter was not significant at 7 T. The paper does not establish that the abnormalities were more widespread beyond the tissue regions it tested. (LLM, conf=0.87)

  cites De Santis et al., 2019 sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites Granberg et al., 2017 source

In the paper

“Tractography and network analysis both have their own limitations (Jeurissen et al., 2019;Sotiropoulos and Zalesky, 2019), which, in combination with the limitations of NODDI, means that caution is required when interpreting the data.”

What the sources say

• Supported

  Jeurissen's discussion accompanying Figure 14 says fiber-tracking connectomes can contain false positives, false negatives, and biased connection strengths, and advises proceeding with extreme caution. The citation supports the tractography limitation in the citing sentence; that sentence does not attribute the limitations of network analysis or NODDI specifically to Jeurissen. (LLM, conf=0.93)

  cites Jeurissen et al., 2019 source
• Supported

  The citation is attached to the statement that tractography and network analysis have limitations, not to the separate mention of NODDI. In “Mapping edges,” the paper describes tractography errors and urges caution; in “Comparing network properties between groups,” it says interpreting brain graphs is challenging and some network measures rest on unestablished assumptions. (LLM, conf=0.92)

  cites Sotiropoulos and Zalesky, 2019 source

In the paper

“The tuberous sclerosis patients also exhibited smaller NDI compared with controls throughout the WM, although no correlation was observed with severity of mental retardation or epilepsy.”

What the sources say

• Partly supported

  The Results (Quantitative analysis; page unavailable) report reduced Ficv, a measure of neurite density, across widespread white-matter regions—not throughout all white matter. The Results (Correlations of clinical symptoms; page unavailable) found no statistically significant correlations with mental retardation or epilepsy, although the Discussion describes slight associations with mental-retardation severity. (LLM, conf=0.85)

  cites Taoka et al., 2020 sourceExperimental — accuracy not yet measured

In the paper

“NDI in this region was not different from the contralateral side in the subacute phase, and was decreased in the chronic phase, presumably representing the course of Wallerian degeneration (Mastropietro et al., 2019).”

What the sources say

• Supported

  The Discussion (page unavailable) reports no significant ipsilateral–contralateral difference in Fic, the paper’s intraneurite NODDI measure, in the subacute phase. The chronic-phase Results (page unavailable) report significantly lower ipsilateral Fic in the distal CST region (CP), and the Discussion describes the changes as compatible with Wallerian degeneration. (LLM, conf=0.92)

  cites Mastropietro et al., 2019 source

In the paper

“Notably, (Sone et al., 2018) observed reduced NDI in patients with TLE in whom experienced radiologists did not find abnormalities on structural MRI, albeit to a lesser extent than in patients with MRI-positive TLE (Fig. 5).”

What the sources say

• Supported

  The Participants section (page unavailable) says experienced neuroradiologists assessed the MRI scans and found no conventional-MRI abnormalities in the MRI-negative/PET-positive group. The Results and Discussion (page unavailable) report reduced neurite density in that group, with less severe abnormalities than in the MRI-positive TLE group with hippocampal sclerosis. (LLM, conf=0.93)

  cites Sone et al., 2018 source

In the paper

“In addition, increased f fw in the substantia nigra, which is consistent with accumulated evidence (Burciu et al., 2017(Burciu et al., , 2016) ) from free water imaging (Pasternak et al., 2009), was observed by (Mitchell et al., 2019) but not by (Kamagata et al., 2016).”

What the sources say

• Supported

  Section 3.4 (page not provided) reports elevated free water in the posterior substantia nigra in Parkinson’s disease, MSAp, and PSP compared with controls. This supports the finding attributed to Mitchell et al.; the separate claim about Kamagata et al. is not assessed against Mitchell’s paper. (LLM, conf=0.93)

  cites Mitchell et al., 2019 source
• ?Not checked — no full text of the sourcecites Kamagata et al., 2016 source

In the paper

“Interestingly, a study of the long-term effects of sports-related concussion reported increased NDI and decreased ODI in athletes at 9-120 months after their last concussion compared with controls”

What the sources say

• Supported

  The Subject Demographics section reports that athletes were imaged 9 months to 10 years after their last concussion. The Discussion reports increased intracellular volume fraction (VIC), indicating greater neurite density, and decreased ODI in concussed athletes compared with controls. Page numbers are not available for these passages. (LLM, conf=0.92)

  cites Churchill et al., 2017 source

In the paper

“In patients with moyamoya disease, a disease characterized by progressive occlusion of the circle of Willis, NODDI parameters in GM and WM were correlated with hemodynamic activities measured with 15 O-gas positron-emission tomography, as well as neurocognitive dysfunction”

What the sources say

• ?Not checked — no full text of the sourcecites Hara et al., 2019 source
• Partly supported

  The Conclusions (page unavailable) report a correlation between NODDI and neuropsychological performance in adult moyamoya disease, and describe white-matter and cortical NODDI differences. The Statistical Analysis section (page unavailable) describes correlations with cognitive scores, but our copy does not report a correlation between NODDI parameters and hemodynamic activity measured by 15O-gas PET. (LLM, conf=0.86)

  cites Hara et al., 2018 sourceExperimental — accuracy not yet measured

In the paper

“After surgery, the orientation parameter showed partial normalization, whereas the reduced axonal water fraction remained unchanged (Kamiya et al., 2017) (Fig. 7).”

What the sources say

• Supported

  The Results (Human data) report that the orientation parameter τ1 decreased significantly after surgery and that neither imaging model found a significant postoperative change in axonal water fraction (AWF). The Discussion describes the increased orientational coherence as tending to normalize while the decreased axon density remained unchanged. (LLM, conf=0.93)

  cites Kamiya et al., 2017 source

In the paper

“Furthermore, individuals with a history of TBI are at higher risk for dementia, neurodegenerative diseases, psychiatric illness, and even mortality, indicating long-term progression of sub-clinical pathology that manifests only in late adulthood (Smith et al., 2013;Wilson et al., 2017).”

What the sources say

• Partly supported

  The Introduction and “Single TBI” section support increased risk of later dementia and neurodegenerative disease, particularly after moderate to severe TBI. “Repetitive mild TBI” reports neuropsychiatric symptoms in selected cases and higher neurodegenerative mortality in former NFL players, but does not establish a general increased risk of psychiatric illness or mortality for everyone with a TBI history. The “Histological features” section also describes pathology appearing within hours of injury, so the claim that it manifests only in late adulthood is not supported. (LLM, conf=0.87)

  cites Smith et al., 2013 sourceExperimental — accuracy not yet measured
• Partly supported

  The sections on neurodegenerative disease after single TBI and stroke and other neurological disorders after TBI report links with dementia, neurodegenerative disease, and neuropsychiatric illness; the Conclusions and future directions report higher mortality. But the paper does not establish that these outcomes all result from sub-clinical pathology manifesting only in late adulthood: it describes that mechanism as a possibility and also reports young-onset dementia and elevated mortality in younger people. No page numbers are available in our copy. (LLM, conf=0.86)

  cites Wilson et al., 2017 sourceExperimental — accuracy not yet measured

In the paper

“The degree of post-operative improvement varies greatly among individuals, and therefore quantitative metrics capable of predicting treatment response have been sought (Halperin et al., 2015).”

What the sources say

• Supported

  The Abstract reports differing outcomes among shunted patients, including marked or moderate improvement in 14 of 18 patients and minimal worsening in one. The Conclusion and Figure 2 discuss measured CSF flow, outflow resistance, and other tests assessed for their ability to predict shunt response. Together, these support the paraphrase. (LLM, conf=0.88)

  cites Halperin et al., 2015 source

In the paper

“No significant differences were observed for ODI, f fw , myelin water fraction, or intra-or extra-neurite water fraction.”

What the sources say

• Partly supported

  The dMRI Results report no significant differences in ODI or isotropic fluid fraction (FISO), and the multi-exponential T2 Results report none in myelin water fraction or combined intra- and extracellular water fraction (IEWF). The comparison is specifically between unidentified bright objects and normal-appearing white matter in children with NF1. The paper does not report separate intra- or extra-neurite water fractions; calling its combined IEWF measure those fractions is not supported. (LLM, conf=0.88)

  cites Billiet et al., 2014 sourceExperimental — accuracy not yet measured

In the paper

“With NODDI, axonal water fraction and orientation dispersion were both smaller within the CST of patients”

What the sources say

• Supported

  The Results (Human data; page unavailable) report significantly lower NODDI axonal water fraction in the CST of iNPH patients than controls, alongside higher τ1, indicating greater orientational coherence. The Image processing section (page unavailable) defines higher τ1 as more parallel orientation, supporting the claim of lower orientation dispersion. (LLM, conf=0.92)

  cites Kamiya et al., 2017 source

In the paper

“Studies have suggested that, depending on their anatomical location (e.g., thalamus), UBOs may be associated with cognitive deficits (Hyman et al., 2007;Moore et al., 1996).”

What the sources say

• Supported

  The Results state that T2 hyperintensities were not associated with cognitive dysfunction overall, but thalamic lesions were associated with cognitive impairment. The Location of T2H section reports lower cognitive scores in children with thalamic lesions, while lesions in several other locations showed no significant lowering. This supports the claim's location-dependent, hedged association. (LLM, conf=0.94)

  cites Hyman et al., 2007 source
• ?Not checked — no full text of the sourcecites Moore et al., 1996 source

In the paper

“the observed NDI increase has been interpreted as representing the inflammatory process.”

What the sources say

• ?Could not verify

  The Abstract and Discussion report increased glutamate measured as Glu/Cr during IFN-alpha treatment and discuss possible inflammatory mechanisms. Our copy does not report an NDI increase or interpret NDI as representing inflammation (searched: NDI, neurite density index). (LLM, conf=0.76)

  cites Haroon et al., 2014 sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites Capuron et al., 2007 source

In the paper

“In WM, NDI has a positive correlation with age, but the increase is non-linear and more rapid during early life (Dean et al., 2016;Jelescu et al., 2015;Lynch et al., 2020) (Fig. 9B).”

What the sources say

• Supported

  The Methods identify νIC as the NODDI measure of neurite density. The Results (Fig. 4; page not provided) report that mean white-matter νIC increases logarithmically with age in the study’s infants and young children, supporting a nonlinear increase that is faster earlier in life. (LLM, conf=0.92)

  cites Dean et al., 2016 source
• Supported

  The Results (§4.1, page unavailable) report that the NODDI-derived intra-axonal fraction increased in all three white-matter regions during the first three years of life and that an exponential curve best described the increase. The Conclusion (§6, page unavailable) likewise describes a large nonlinear increase with age and notes that the trajectory levels off, supporting a faster increase early in the period studied. (LLM, conf=0.90)

  cites Jelescu et al., 2015 source
• Supported

  The Discussion reports widespread age-related increases in white-matter tract NDI. It describes an exponential growth pattern, with rapid increases in infancy and young childhood followed by a plateau in adolescence. (LLM, conf=0.94)

  cites Lynch et al., 2020 source

In the paper

“Although DTI has consistently shown reduced FA in the compressed spinal cord, the correlation of FA with symptoms and postoperative recovery remains controversial (Jones et al., 2013;Rajasekaran et al., 2017).”

What the sources say

• Partly supported

  The Correlation with Initial Clinical Assessment Scores section reports lower FA with poorer neurologic function in patients with cervical spondylotic myelopathy. The Relation to Surgical Outcome section reports an association between baseline FA and improvement on one postoperative measure, but not on the Nurick or mJOA scales; the Conclusions call the results preliminary. This supports uncertainty about postoperative prediction, but the paper does not establish that correlations with symptoms and recovery are broadly controversial or that reduced FA has been shown consistently across studies. (LLM, conf=0.85)

  cites Jones et al., 2013 sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites Rajasekaran et al., 2017 source

In the paper

“However, (Genç et al., 2018) have reported that higher intelligence in healthy young to middle-aged individuals was associated with lower cortical ODI”

What the sources say

• Supported

  The Results (Associations on a whole-brain level; page not provided) report a negative association between intelligence and cortical ODI in both samples. The same section describes participants aged 18–40 and 22–36, and the Methods describe the experimental sample as healthy. (LLM, conf=0.93)

  cites Genç et al., 2018 source

In the paper

“Reduced ODI in the frontal cortices and increased ODI in the parietal cortices and the striatum have been reported from a study in young-adult binge drinkers (Morris et al., 2018).”

What the sources say

• Supported

  The Discussion reports reduced ODI in frontal cortical grey matter and increased ODI in parietal cortex in young-adult binge drinkers. It also reports higher ODI in the ventral striatum; the Results section confirms this comparison with healthy volunteers. (LLM, conf=0.91)

  cites Morris et al., 2018 source

In the paper

“Another study found that lower NDI in WM was associated with poorer mathematics performance-but not with reading performance-across children born very preterm and full term (Collins et al., 2019).”

What the sources say

• Supported

  The Results (Associations between white matter microstructure and mathematics; page not available) report that higher neurite density was associated with better mathematics performance in the combined VPT and FT sample. The Results (Associations between white matter microstructure and word reading; page not available) report no association between any measured diffusion parameter and word reading performance in that sample. (LLM, conf=0.93)

  cites Collins et al., 2019 source

In the paper

“For example, in patients receiving interferon alpha treatment in whom an inflammatory cytokine response was confirmed by blood test, ODI remained unchanged, although NDI was increased (Dowell et al., 2019).”

What the sources say

• Partly supported

  The Results sections on inflammatory cytokine response and acute NODDI effects report a blood-measured cytokine response and no significant change in ODI. However, they also report no significant increase in NDI after IFN-α; most of the listed NDI means were slightly lower at 4 hours. (LLM, conf=0.90)

  cites Dowell et al., 2019 sourceExperimental — accuracy not yet measured

In the paper

“Compared with controls, infants with congenital heart disease exhibited smaller NDI and greater ODI in WM (Karmacharya et al., 2018), as well as smaller ODI in cortical GM (Kelly et al., 2019)”

What the sources say

• Supported

  The Results section (‘White matter regional abnormalities in CHD’; page unavailable) reports lower NODDI intracellular volume fraction (Vic) in the bilateral SFOF and left PLIC, and higher ODI in the left corpus callosum, in CHD neonates compared with controls. These are region-specific white-matter findings. The cortical-gray-matter claim is attached to Kelly et al. (2019), not to this source. (LLM, conf=0.90)

  cites Karmacharya et al., 2018 source
• Supported

  The Results section ('Cortical ODI Is Reduced in Infants With CHD') reports significantly lower cortical ODI in infants with congenital heart disease than in controls. Kelly et al. is cited for the cortical gray matter finding, not the white matter findings attributed to Karmacharya et al. (LLM, conf=0.94)

  cites Kelly et al., 2019 source

In the paper

“Both lower NDI and greater ODI were correlated with poorer neurodevelopmental outcomes (Kelly et al., 2016;Mürner-Lavanchy et al., 2018;Young et al., 2019).”

What the sources say

• Supported

  The Discussion’s Conclusions state that higher axon dispersion (ODI) and lower axon density (NDI) in very preterm children are associated with some neurodevelopmental delays. The Neurodevelopmental outcomes section also links both measures to poorer motor outcomes. (LLM, conf=0.91)

  cites Kelly et al., 2016 source
• Supported

  The Discussion reports that higher axon density was associated with better IQ and motor performance and fewer behavioural problems, while lower axon dispersion was associated with better motor performance and fewer behavioural problems in very-preterm children. It also reports higher axon density with better semantic performance in this study, although evidence for dispersion–language associations was weaker. Those findings support the broad claim about poorer neurodevelopmental outcomes; the motor and behavioural findings are reported from earlier work. (LLM, conf=0.88)

  cites Mürner-Lavanchy et al., 2018 source
• Partly supported

  The Discussion reports that, among children born very preterm, lower NDI was associated with poorer intelligence and visuomotor skills, but intelligence was associated with NDI rather than ODI. Greater ODI was associated with lower intellectual scores in the full-term group; in that group, higher—not lower—NDI was also associated with lower scores. The claim combines associations observed in different groups as though both applied together. (LLM, conf=0.90)

  cites Young et al., 2019 sourceExperimental — accuracy not yet measured

In the paper

“Focusing on the fixed diffusivity assumption, (Guerrero et al., 2019) have suggested that the default value of 1.7 μm 2 /ms is likely suboptimal for neonates brain and GM, and that the age-dependence of NDI and ODI is affected by the choice of fixed diffusivity value.”

What the sources say

• Supported

  The Results report that the optimal fixed diffusivity is below 1.7 μm²/ms in infant white and gray matter, and in gray matter across age groups. The Results and Discussion also show that changing this value alters age-group differences in neurite density and orientation-related NODDI parameters. The copy provides section labels but no page numbers. (LLM, conf=0.93)

  cites Guerrero et al., 2019 source

In the paper

“Accumulating evidence suggests a major role of chronic neuroinflammation in the neurodegeneration that occurs in aging and in a broad range of diseases including Alzheimer's disease, Parkinson's disease, MS, ALS, stroke, TBI, and psychiatric diseases (Amor et al., 2010;Glass et al., 2010;Vikhreva et al., 2016).”

What the sources say

• Partly supported

  The Abstract and Conclusions describe chronic immune activation as a feature of many neurodegenerative diseases, and the paper discusses the listed neurological conditions, including ALS. The Introduction mentions an aging population but does not establish a role for chronic neuroinflammation in neurodegeneration caused by aging; the passages found also do not address psychiatric diseases. Page numbers are unavailable in our copy. (LLM, conf=0.88)

  cites Amor et al., 2010 sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction and Conclusion support a role for sustained neuroinflammation in the progression of Alzheimer’s, Parkinson’s, ALS, and MS, while noting that inflammation may amplify pathology rather than initiate disease. The Alzheimer’s Disease section identifies AD as age-related, but the cited passages do not establish the broader claim about neurodegeneration in aging, stroke, traumatic brain injury, or psychiatric diseases. Page numbers are unavailable in our copy. (LLM, conf=0.88)

  cites Glass et al., 2010 sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites Vikhreva et al., 2016 source

In the paper

“For example, T2s are likely different among compartments (Lampinen et al., 2020;Veraart et al., 2018) and are affected by pathologies like neuroinflammation (Stanisz et al., 2004), demyelination (Bartzokis, 2004;Lin et al., 2018), and intramyelinic edema (Harkins et al., 2013).”

What the sources say

• Supported

  The Lampinen et al. citation is attached to the claim that T2 values differ among compartments, not to the subsequent examples of pathologies, which have separate citations. The Results (In vivo results; page not provided) report different T2 values for the ‘stick’ and ‘zeppelin’ compartments in both healthy white matter and lesions. (LLM, conf=0.94)

  cites Lampinen et al., 2020 source
• Supported

  Veraart et al. is cited for the first clause, that T2 values differ among compartments. The Discussion reports empirical evidence that intra- and extra-axonal T2 relaxation times differ, and the Conclusion reiterates this finding in the living human brain. The pathology examples have separate citations and are not attributed to Veraart et al. (LLM, conf=0.94)

  cites Veraart et al., 2018 source
• ?Not checked — no full text of the sourcecites Stanisz et al., 2004 source

In the paper

“Although the negative correlation may look counterintuitive, it is in line with a number of studies reporting higher FA within this region in individuals with reading difficulties and the hypothesis that left lateralization of reading-related functions is diminished in struggling readers (Dougherty et al., 2007;Finn et al., 2014).”

What the sources say

• ?Not checked — no full text of the sourcecites Dougherty et al., 2007 source
• Partly supported

  Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  cites Finn et al., 2014 sourceExperimental — accuracy not yet measured

In the paper

“For translation into the clinic, an imaging biomarker needs to pass multiple steps in three parallel tracks: technical validation (precision, accuracy, and realizability), biological/clinical validation (performance in measuring relevant biological properties or predicting clinical outcomes), and cost-effectiveness (Connor et al., 2017).”

What the sources say

• Supported

  The imaging biomarker roadmap section describes parallel technical and biological/clinical validation tracks, with cost-effectiveness as a third track considered as validation evidence accumulates. It specifies precision, accuracy and availability for technical performance, and measurement of relevant biology or prediction of clinical outcomes for biological/clinical performance. The cost-effectiveness section says clinical translation requires good value for money. (LLM, conf=0.91)

  cites Connor et al., 2017 source

In the paper

“Later, (Batalle et al., 2019) reported that cortical development between 25 and 38 weeks is dominated by an increase of ODI, whereas between 38 and 47 weeks it is dominated by an increase of NDI”

What the sources say

• Supported

  The Discussion describes increases in ODI before 38 weeks PMA, followed by increased NDI with no change in ODI after 38 weeks. The Abstract gives the same 25–38 and 38–47 week pattern, while noting that later NDI increases were regionally confined. (LLM, conf=0.91)

  cites Batalle et al., 2019 source

In the paper

“Preliminary applications of NODDI and the g-ratio framework include studies of brain maturation during childhood (Dean et al., 2016;Geeraert et al., 2019), the effects of age and gender in young-to-late adulthood (Berman et al., 2018;Cercignani et al., 2017;Slater et al., 2019), and the effects of diseases like MS (Hagiwara et al., 2019;Stikov et al., 2015;Yu et al., 2019) and cervical spondylotic”

What the sources say

• Supported

  Dean et al. is cited for the childhood-maturation example, not the separately cited adult or disease studies. The Discussion reports a preliminary NODDI-based g-ratio study in children and a developmental trajectory across infancy and early childhood; page numbers are not provided in our copy. (LLM, conf=0.93)

  cites Dean et al., 2016 source
• ?Not checked — no full text of the sourcecites Geeraert et al., 2019 source
• Supported

  Berman et al. is an application of the g-ratio framework to age and sex: Methods §2.1 reports participants aged 7–85, and §2.5 describes testing sex differences across age ranges. The Conclusions state that the study tested g-ratio stability with age and between the sexes. The citing sentence identifies the subject of the study; it does not claim that the study found significant age or sex effects. (LLM, conf=0.92)

  cites Berman et al., 2018 source

In the paper

“A few studies have adopted this type of multi-parameter analysis with NODDI (Andica et al., 2020;Churchill et al., 2019Churchill et al., , 2017)), although most have been limited to combinations with other dMRI measures.”

What the sources say

• Partly supported

  The Methods (Linked independent component analysis) confirm that this study combined NODDI with DTI and VBM measures in a multi-parameter analysis. Because VBM measures white-matter volume, this study was not limited to other dMRI measures; the cited paper also does not establish that most such studies were so limited. (LLM, conf=0.85)

  cites Andica et al., 2020 sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites Churchill et al., 2019 source
• Supported

  The Magnetic Resonance Imaging section describes a protocol estimating both DTI and NODDI parameters. The Multivariate analysis: effects of concussion section says the study used N-way Partial Least Squares to examine covariation across multiple diffusion MRI measures and concussion history. This supports citing the paper as an example of multi-parameter NODDI analysis combined with another dMRI measure; the paper does not establish how common that approach is across other studies. (LLM, conf=0.91)

  cites Churchill et al., 2017 source

In the paper

“Several groups have been even more ambitious and have investigated the use of MRI g-ratio as weights in connectomics studies (Kamagata et al., 2019;Mancini et al., 2018).”

What the sources say

• Supported

  The Introduction says the authors mapped connectomes based on the MR g-ratio in patients with multiple sclerosis. The Methods describe their g-ratio-weighted connectivity matrices, directly supporting the use of MRI g-ratio as connectome weights. (LLM, conf=0.94)

  cites Kamagata et al., 2019 source
• Supported

  Section 2.4.2 (‘G-ratio weighted networks’) says the study modelled connectome edges with the g-ratio itself as their weight. This supports citing Mancini et al. as one group that investigated MRI g-ratio weights in connectomics; it does not need to establish what the other cited group did. (LLM, conf=0.94)

  cites Mancini et al., 2018 source

In the paper

“One rapidly growing field is joint modeling of diffusion and relaxation (Gong et al., 2020;Lampinen et al., 2020;Veraart et al., 2018) (see also Hutter et al. in this special issue).”

What the sources say

• Supported

  The Introduction (page unavailable) describes combined modelling of diffusion and T2 relaxation and traces its development into more recent imaging methods. The Discussion (page unavailable) says an early study inspired many techniques and gives recent examples, supporting the broad characterization of a growing field. (LLM, conf=0.86)

  cites Gong et al., 2020 source
• Supported

  The Introduction (page unavailable) treats diffusion-relaxation experiments as an active research approach, describing their use to estimate compartment-specific T2 values and citing several recent studies. The paper itself develops a joint diffusion-relaxation model, supporting its citation as an example of this growing field. (LLM, conf=0.84)

  cites Lampinen et al., 2020 source
• Supported

  The Introduction describes the paper’s own approach as combining diffusion and relaxation in a biophysical model. It also points to a vast literature on joint diffusion–relaxation methods and recent extensions to imaging, supporting the citing sentence’s broad characterization of an active, growing field. (LLM, conf=0.87)

  cites Veraart et al., 2018 source

In the paper

“As with other advanced imaging techniques, most of the disease studies reported to date have been cross-sectional and have included a sample of limited size from a single center.”

In the paper

“In contrast, biophysical models aim at specificity by parametrizing the dMRI signal as a function of biophysically meaningful parameters (e.g., axon density).”

In the paper

“There are two main reasons for introducing a new imaging technique to the study of a disease: to improve sensitivity to target pathologies and to provide better explanations for observed abnormalities.”

In the paper

“As the models are often based on assumptions that are not yet proven even in the normal brain, users must be aware of the risk of being misled in their interpretations of the data.”

In the paper

“The validity of model assumptions is currently under debate.”

In the paper

“The signal representation approach aligns with the first reason, to improve sensitivity to target pathologies, with the caveat that this approach comes with no obvious biological interpretation.”

In the paper

“neurite orientation dispersion and density imaging (NODDI) has been proposed as a way to relate the dMRI signal to tissue features via biophysically inspired modeling.”

What the sources say

• ?Not checked — no full text of the sourcecites Zhang et al., 2012 source

In the paper

“With NODDI, reduced NDI and ODI have been observed in the substantia nigra and putamen of patients with Parkinson's disease compared with controls, and these reductions were predominant on the side corresponding to symptoms (contralateral) (Kamagata et al., 2016).”

What the sources say

• ?Not checked — no full text of the sourcecites Kamagata et al., 2016 source

In the paper

“Despite the remaining ambiguity about model assumptions, research to investigate the clinical usefulness of NODDI can still be justified.”

In the paper

“NODDI may provide useful biomarkers if the effect of disease is far greater than the bias introduced by the constraints.”

In the paper

“As the one true model has not been identified, results of NODDI, and any other model, are more or less phenomenological fit parameters that probably represent mixes of microstructural features rather than unambiguous quantifications of specific tissue properties.”

In the paper

“The debates on specificity and model validity have caused confusion among clinicians who are not familiar with the art of modeling and who wonder whether it is okay to use NODDI in their research.”

In the paper

“Important to note is that usual multi-shell dMRI data has limited information that allows estimation of only a handful of parameters.”

In the paper

“With regard to the degree of freedom, NODDI has 5 free parameters (NDI, ODI, f fw , and polar and azimuthal angles of the principal direction of the bundle), which is one less than DTI.”

In the paper

“The range of pathologies may be too diverse to be adequately captured with a single model, and models need to be validated in the target disease, not only in the healthy state.”

In the paper

“Remember, apparent diffusion coefficient, virtually the only dMRI metric that has gained a presence in clinical practice, is a mix of a multitude of microstructural effects.”

In the paper

“Besides, the relationships between the model parameters and pathologies (e.g., atrophy and loss of axons, demyelination, microglial activation) are not one-to-one, and in reality different pathologies tend to coexist rather than occur in isolation.”

In the paper

“In a study of mild cognitive impairment, the mild cognitive impairment group had relatively greater ODI and smaller NDI than controls, although DTI was more sensitive to the group differences (Q.Wen et al., 2019b).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Q.Wen et al., 2019b

In the paper

“Another study, which compared healthy controls, subjects with mild cognitive impairment, and patients with Alzheimer's disease, reported that NDI was the most sensitive to group differences, and that NDI, ODI, and f fw were all correlated with cognitive score (Fu et al., 2019).”

What the sources say

• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Fu et al., 2019

In the paper

“We also emphasize dependence on assumptions applies to all of the proposed biophysical models and does not necessarily undermine the value of NODDI.”

In the paper

“The most widespread and successful clinical application of dMRI to date is the detection of acute ischemic infarction (Moseley et al., 1990).”

What the sources say

• ?Not checked — no full text of the sourcecites Moseley et al., 1990 source

In the paper

“In addition, this work found a weak negative correlation between NDI and dendritic markers, which is counter-intuitive to the name 'neurite density index'.”

What the sources say

• ?Not checked — no full text of the sourcecites Crombe et al., 2018 source

In the paper

“Yet, there is an argument that those FCDs presented in (Winston et al., 2014) and (Rostampour et al., 2018) are readily detectable on conventional T1-/T2-weighted images (Chougar et al., 2018).”

What the sources say

• ?Not checked — no full text of the sourcecites Chougar et al., 2018 source

In the paper

“In addition, iNPH is primarily a disorder of cerebrospinal fluid dynamics that leads to cerebral dyshomeostasis, hypoperfusion, and hypometabolism (Bräutigam et al., 2019)”

What the sources say

• ?Not checked — no full text of the sourcecites Bräutigam et al., 2019 source

In the paper

“volume reduction and abnormal activation of the left superior temporal gyrus are reported to be associated with auditory hallucination in schizophrenia (Allen et al., 2007).”

What the sources say

• ?Not checked — no full text of the sourcecites Allen et al., 2007 source

In the paper

“A longitudinal study in athletes with concussion (Churchill et al., 2019) showed reduced NDI in WM in the early symptomatic phase of injury, and this decrease persisted at the time of return-to-play.”

What the sources say

• ?Not checked — no full text of the sourcecites Churchill et al., 2019 source

In the paper

“In addition, increased ODI as well as decreased FA were also observed at both the early symptomatic phase and return-to-play.”

What the sources say

• ?Not checked — no full text of the sourcecites Churchill et al., 2019 source

In the paper

“Overall, despite the issue of the validity of the model assumptions, especially in the acute phase, NODDI has shown the ability to track brain alterations caused by ischemia and it may be a potentially useful means of predicting chronic disabilities.”

In the paper

“In schizophrenia, a severe mental illness characterized by positive (e.g., hallucinations and delusions) and negative (e.g., anhedonia and flat affect) symptoms, axon disruption with concurrent reduced myelin levels and oligodendrocyte abnormalities has been postulated as a key pathology based on both imaging and postmortem studies (Davis et al., 2003;Vikhreva et al., 2016).”

What the sources say

• ?Not checked — no full text of the sourcecites Davis et al., 2003 source
• ?Not checked — no full text of the sourcecites Vikhreva et al., 2016 source

In the paper

“Overall, the reported results and hypothetical interpretations of NODDI in TBI indicate the need to account for variabilities in time after injury, injury biomechanics, and individual pre-injury baseline.”

In the paper

“Some damaged axons show recovery with remyelination, whereas others progress into the irreversible stage characterized by axonal fragmentation and disconnection.”

What the sources say

• ?Not checked — no full text of the sourcecites Armstrong et al., 2016 source

In the paper

“therefore, neurodegeneration is thought to be a consequence rather than a pathogenetic factor, and it remains uncertain as to what degree neurodegeneration explains the symptoms.”

What the sources say

• ?Not checked — no full text of the sourcecites Bräutigam et al., 2019 source

In the paper

“In a longitudinal follow-up, NDI recovered to values comparable with the non-compressed level at 6 months after surgery, with concurrent recovery of neurological score (Iwama et al., 2020).”

What the sources say

• ?Not checked — no full text of the sourcecites Iwama et al., 2020 source

In the paper

“with gliomas and metastases tending to exhibit a higher extra-neurite fraction and a higher f fw , respectively (Fig. 8).”

What the sources say

• ?Not checked — no full text of the sourcecites Caverzasi et al., 2016 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Kadota et al., 2018

In the paper

“NODDI has also revealed a regional variation in WM maturation, which is consistent with the known asynchrony of myelination and fiber development; maturation generally proceeds from posterior to anterior, and primary motor and sensory regions develop earlier than areas of higher order functions (Brody et al., 1987;Yakolev and Lecours, 1967).”

What the sources say

• ?Not checked — no full text of the sourcecites Brody et al., 1987 source
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites Yakolev and Lecours, 1967

In the paper

“Application of NODDI in clinical research has mostly reported promising results that dMRI biophysical models can be potentially useful for improving patient stratification and prediction of neurological functions.”

In the paper

“In addition, the clinical value of NODDI needs to be evaluated in comparison with more accessible measures other than dMRI, with critical consideration given to cost effectiveness (Rao and Levin, 2012).”

What the sources say

• ?Not checked — no full text of the sourcecites Rao and Levin, 2012 source

In the paper

“Although NODDI is a simplified model, it is already complex compared with the current clinical standard, apparent diffusion coefficient which is simply the average of diffusivities measured along three orthogonal axes.”

In the paper

“In the vast majority of studies, disease-related effects were in directions that are biologically plausible, although this does not assure specificity to the tissue pathology.”