ABSTRACT
BACKGROUND: Cognitive dysfunction is the most common non-motor symptom in Parkinson's disease (PD), and timely detection of a slight cognitive decline is crucial for early treatment and prevention of dementia. This study aimed to build a machine learning model based on intra- and/or intervoxel metrics extracted from diffusion tensor imaging (DTI) to automatically classify PD patients without dementia into mild cognitive impairment (PD-MCI) and normal cognition (PD-NC) groups. METHODS: We enrolled PD patients without dementia (52 PD-NC and 68 PD-MCI subtypes) who were assigned to the training and test datasets in an 8:2 ratio. Four intravoxel metrics, including fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD), and two novel intervoxel metrics, local diffusion homogeneity (LDH) using Spearman's rank correlation coefficient (LDHs) and Kendall's coefficient concordance (LDHk), were extracted from the DTI data. Decision tree, random forest, and eXtreme gradient boosting (XGBoost) models based on individual and combined indices were built for classification, and model performance was assessed and compared via the area under the receiver operating characteristic curve (AUC). Finally, feature importance was evaluated using SHapley Additive exPlanation (SHAP) values. RESULTS: The XGBoost model based on a combination of the intra- and intervoxel indices achieved the best classification performance, with an accuracy of 91.67%, sensitivity of 92.86%, and AUC of 0.94 in the test dataset. SHAP analysis showed that the LDH of the brainstem and MD of the right cingulum (hippocampus) were important features. CONCLUSIONS: More comprehensive information on white matter changes can be obtained by combining intra- and intervoxel DTI indices, improving classification accuracy. Furthermore, machine learning methods based on DTI indices can be used as alternatives for the automatic identification of PD-MCI at the individual level.
Subject(s)
Cognitive Dysfunction , Dementia , Parkinson Disease , Humans , Diffusion Tensor Imaging , Parkinson Disease/complications , Parkinson Disease/diagnostic imaging , Cognitive Dysfunction/complications , Cognitive Dysfunction/diagnostic imaging , Decision TreesABSTRACT
BACKGROUND: It is yet unknown if the whole-brain resting-state network is altered in multiple system atrophy with symptoms of depression. This study aimed to investigate if and how depression symptoms in multiple system atrophy are associated with resting-state network dysfunction. METHODS: We assessed the resting-state functional network matric using Degree centrality (DC) coupling with a second ROI-wise functional connectivity (FC) algorithm in a multimodal imaging case-control study that enrolled 32 multiple system atrophy patients with depression symptoms (MSA-D), 30 multiple system atrophy patients without depression symptoms (MSA-ND), and 34 healthy controls (HC). RESULTS: Compared to HC, MSA-D showed more extensive DC hub dysfunction in the left precentral and right middle frontal cortex than MSA-ND. A direct comparison between MSA-D and MSA-ND detected increased DC in the right anterior cingulum cortex, but decreased DC in the left cerebellum lobule IV and lobule V, left middle pole temporal cortex, and right superior frontal cortex. Only right anterior cingulum cortex mean DC values showed a positive correlation with depression severity, and used ACC as seed, a second ROI-wise functional connectivity further revealed MSA-D patients showed decreased connectivity between the ACC and right thalamus and right middle temporal gyrus (MTG). CONCLUSIONS: These findings revealed that dysfunction of rACC, right middle temporal lobe and right thalamus involved in depressed MSA. Our study might help to the understanding of the neuropathological mechanism of depression in MSA.
Subject(s)
Multiple System Atrophy , Brain/diagnostic imaging , Brain Mapping/methods , Case-Control Studies , Depression/diagnostic imaging , Humans , Magnetic Resonance Imaging/methods , Multiple System Atrophy/diagnostic imagingABSTRACT
The aim of this study was to evaluate the structural integrity of the thalamic connectivity of specific fiber tracts in different stages of Alzheimer's disease (AD) using diffusion tensor imaging (DTI). Thirty-five patients with AD and 22 normal control (NC) subjects were recruited. Based on Mini Mental State Examination score, the AD patients were divided into three subgroups for comparison with the NC group: mild (mi-AD, n = 14), moderate (mo-AD, n = 12), and severe (se-AD, n = 9) AD. The fornix (FX), anterior thalamic radiation (ATR), and posterior thalamic radiation (PTR) were selected to represent the thalamic connectivity with other brain regions. The fornix was divided into the column and body of the fornix (FX-1) and the bilateral fornix (crus)/stria terminalis (FX-2/ST) based on the atlas. Through the atlas-based analysis and fiber tracking method, we measured fractional anisotropy (FA), mean diffusivity (MD), and tract volume to reflect the microstructural and macrostructural changes of these fibers during AD progression. There were significant differences in the FA and MD of all fibers, except the right PTR, between the AD and NC subjects. Further subgroup analyses revealed that the mi-AD subgroup had decreased FA only in the FX-1 and increased MD in the FX-1 and bilateral ATR, the mo-AD subgroup showed declined FA and increased MD in the FX-1, bilateral FX-2/ST and ATR; the se-AD subgroup exhibited lower FA and higher MD values in all fibers except the right PTR. We also found reduced tract volume values in the FX and left ATR in the AD patients. Further subgroup analyses revealed that these differences only existed in the se-AD patients. Our DTI analyses indicate that the integrity of thalamic connectivity is progressively disrupted following cognitive decline in AD and that DTI parameters in the column and body of the fornix show promise as potential markers for the early diagnosis of AD and for monitoring disease progression.
Subject(s)
Alzheimer Disease/diagnosis , Alzheimer Disease/metabolism , Diffusion Tensor Imaging , Nerve Net/metabolism , Nerve Net/pathology , Aged , Aged, 80 and over , Diffusion Tensor Imaging/methods , Female , Humans , Male , Middle AgedABSTRACT
Due to heightened level of neuroplasticity, there is a sensitive period (2-4 years after birth) that exists for optimal central auditory development. Using diffusion tensor imaging combined with resting-state functional connectivity (rsFC) analysis, this study directly investigates the structural connectivity alterations of the whole brain white matter (WM) and the functional reorganization of the auditory network in infants with sensorineural hearing loss (SNHL) during the early sensitive period. 46 bilateral profound SNHL infants prior to cochlear implantation (mean age, 17.59 months) and 33 healthy controls (mean age, 18.55 months) were included in the analysis. Compared with controls, SNHL infants showed widespread WM alterations, including bilateral superior longitudinal fasciculus, inferior fronto-occipital fasciculus, inferior longitudinal fasciculus, right corticospinal tract, posterior thalamic radiation and left uncinate fasciculus. Moreover, SNHL infants demonstrated increased rsFC between left/right primary auditory cortex seeds and right insula and superior temporal gyrus. In conclusion, this study suggests that SNHL in the early sensitive period is associated with diffuse WM alterations that mainly affect the auditory and language pathways. Furthermore, increased rsFC in areas mainly associated with auditory and language networks may potentially reflect reorganization and compensatory activation in response to auditory deprivation during the early sensitive period.
Subject(s)
Auditory Cortex/diagnostic imaging , Diffusion Tensor Imaging/methods , Hearing Loss, Sensorineural/diagnostic imaging , Magnetic Resonance Imaging/methods , Nerve Net/diagnostic imaging , Auditory Cortex/physiopathology , Child, Preschool , Female , Hearing Loss, Sensorineural/physiopathology , Humans , Infant , Male , Nerve Net/physiopathology , Neuronal Plasticity/physiologyABSTRACT
Seizures may be the first or sometimes the only manifestation of patients with glioma in clinics. The aim of operation is to eliminate epilepsy far beyond mere resection of tumor mass. The underlying mechanisms of glioma-associated epileptogenesis are poorly understood. Recently the theory of amino-acid like neurotransmitters in chemical synapse is gradually accepted. However, the molecular mechanisms remain to be further investigated on how glutamate release is regulated and how synaptic homeostasis in peripheral neurons is kept or disturbed. So detailed studies are needed to clarify specific molecular target and provide proper evidence for optimal antiepileptic drugs in glioma-associated epileptogenesis.
Subject(s)
Brain Neoplasms/complications , Epilepsy/etiology , Glioma/complications , Neurotransmitter Transport Proteins/metabolism , Synapses/chemistry , Brain Neoplasms/metabolism , Glioma/metabolism , Glutamic Acid/metabolism , Humans , Neurotransmitter Agents/metabolismABSTRACT
PURPOSE: This study assesses the patterns of structural and functional connectivity damage in patients with Parkinson's disease dementia (PDD) compared with cognitively unimpaired Parkinson's disease patients (PD-Cu) and healthy controls (HC). MATERIALS AND METHODS: Resting state functional magnetic resonance imaging (rs-fMRI) and diffusion tensor magnetic resonance imaging (DTI) scans were obtained from 30 PD and 21 sex- and age-matched HC. The between-group difference in posterior cingulate (PCC) functional connectivity (FC) was performed to assess FC dysfunction. Atlas-based spatial statistics of DTI was applied to compare White matter (WM) fibers impairment between groups. RESULTS: (1) Functional connectivity: (1) PD-Cu compared with HC showed a decreased PCC functional connectivity of the right medial temporal lobe (MTL). In addition, PCC-right MTL connectivity strength of PD was significantly correlated with Montreal Cognitive Assessment (MoCA) score. (2) PDD group shows a decreased FC of PCC-right parahippocampa compared with PD-Cu group; while show a widespread decreased PCC FC compared with HC group. (2) Anatomical connectivity: (1) Relative to PD-Cu, significant lower FA values were found in the left hippocampus in PDD. (2) PDD showed higher MD values in a widespread WM regions compared with PD-Cu and HC. (3) Positive correlation was observed between MoCA score and FA value of left inferior longitudinal and hippocampus, and bilateral superior longitudinal fasciculus in PD. CONCLUSIONS: Cognitive decline in PD is associated with FC damage of PCC-right MTL and microstructural damage of left hippocampus. Nevertheless, combining fMRI and DTI method may provide markers able to contribute to the prediction of PDD.
Subject(s)
Cognition Disorders/complications , Magnetic Resonance Imaging , Parkinson Disease/complications , White Matter/pathology , White Matter/physiopathology , Cognition Disorders/physiopathology , Diffusion Magnetic Resonance Imaging , Female , Gyrus Cinguli/pathology , Hippocampus/pathology , Humans , Male , Middle Aged , Nerve Net/physiopathology , Parkinson Disease/physiopathology , White Matter/anatomy & histologyABSTRACT
PURPOSE: To investigate the microstructural integrity of superior cerebellar peduncles (SCP) and middle cerebellar peduncles (MCP) by using DTI tractography method, and further to detect whether the microstructural integrity of these major cerebellar pathways is related to motor function in children with diffuse periventricular leucomalacia (PVL) born preterm. MATERIALS AND METHODS: 46 children with diffuse PVL (30 males and 16 females; age range 3-48 months; mean age 22.4 ± 6.7 months; mean gestational age 30.5 ± 2.2 weeks) and 40 healthy controls (27 males and 13 females; age range 3.5-48 months; mean age 22.1 ± 5.8 months) were enrolled in this study. DTI outcome measurements, fractional anisotropy (FA), for the SCP, MCP and cortical spinal tract (CST) were calculated. The gross motor function classification system (GMFCS) was used for assessing motor functions. RESULTS: Compared to the controls, patients with diffuse PVL had a significantly lower FA in bilateral SCP, MCP and CST. There was a significant negative correlation between GMFCS levels and FA in bilateral SCP, MCP and CST in the patients group. In addition, significant inverse correlation of FA value was found between not only the contralateral but also the ipsilateral CST and SCP/MCP. CONCLUSIONS: These findings suggest that the injury of SCP and MCP may contribute to the motor dysfunction of diffuse PVL. Moreover, the correlations we found between supratentorial and subtentorial injured white matter extend our knowledge about the cerebro-cerebellar white matter interaction in children with diffuse PVL.
Subject(s)
Diffusion Tensor Imaging/methods , Leukomalacia, Periventricular/complications , Leukomalacia, Periventricular/pathology , Middle Cerebellar Peduncle/pathology , Movement Disorders/complications , Movement Disorders/pathology , Child, Preschool , Female , Humans , Infant , Infant, Premature , Male , Neural Pathways/pathology , Reproducibility of Results , Sensitivity and SpecificityABSTRACT
We herein report a rare case of intraspinal ectopic endometrial tissue associated with tethered cord syndrome. The patient underwent MR imaging of the lumbar spine and CT spinal angiography. Asymptomatic dysraphism was also detected, including spinal bifida, low-lying conus medullaris, spinal meningocele and a lumbosacral lipoma. Venous reflux obstruction was also suggested. The patient underwent L2-S1 laminectomy and microdecompression of the lesion. The histological and immunohistochemical features were characteristic of ectopic endometrial tissue. Since the surgery, no neurological signs have been noted, either during or outside the patient's menstrual periods. The two-month follow-up MRI scans showed a regression of the lesion.
Subject(s)
Choristoma/diagnosis , Endometrium , Laminectomy , Lipoma/pathology , Lumbar Vertebrae/pathology , Neural Tube Defects/pathology , Paraparesis/etiology , Spine/pathology , Angiography , Choristoma/complications , Disease Progression , Female , Follow-Up Studies , Humans , Lipoma/complications , Lipoma/physiopathology , Magnetic Resonance Imaging , Middle Aged , Neural Tube Defects/complications , Neural Tube Defects/physiopathology , Paraparesis/physiopathologyABSTRACT
OBJECTIVE: To distinguish lingual thyroglossal duct cyst (LTDC) from laryngomalacia in newborn infants. METHODS: Data of 10 newborn infants with laryngeal stridor and dyspnea, admitted to the department of neonatology in our hospital during December, 2004 to August, 2007, who were finally diagnosed with LTDC though previously diagnosed as congenital laryngeal stridor in other hospitals, were summarized and analyzed. RESULTS: Inspiratory stridor with chest wall retractions was cardinal symptom of newborn with LTDC. A slightly gray and round cyst with smooth surface at the base of the tongue was found with laryngoscopy. On computed tomography examination of larynx, a well-circumscribed lesion with low intensity was detected at the base of the tongue protruding into the air passage. Pathological examination demonstrated that the cyst wall was composed of tabular and columnar epithelium. CONCLUSIONS: LTDC is a common disease in newborns, which is similar to laryngomalacia. For neonates suspected of LTDC, laryngoscopic examination should be performed first, while laryngeal CT scan is an important diagnostic basis. Cyst puncture can ameliorate the symptoms of the patients, while surgical removal is the method of radical cure.
Subject(s)
Thyroglossal Cyst/diagnosis , Early Diagnosis , Female , Humans , Infant, Newborn , Male , Respiratory Sounds , Retrospective StudiesABSTRACT
OBJECTIVE: Periventricular leukomalacia (PVL), the principal form of brain injury in the premature infant, is characterized by overt focal necrotic lesions in periventricular white matter and diffuse cerebral white matter injury. The early detection of the disease is not consistently possible with cranial ultrasonography or conventional magnetic resonance imaging (MRI). Recently, diffusion-weighted imaging (DWI) has been shown to be highly sensitive in detecting acute ischemic cerebral injury. This study aimed to evaluate possible role of DWI in early diagnosis of PVL. METHODS: Images and clinical data from 12 preterm infants with PVL diagnosed in our NICU from August, 2005 to April, 2007 were reviewed. MRI using conventional and diffusion-weighted imaging, as well as the assays of blood routine test, serum bilirubin, C-reactive protein (CRP), blood culture, blood gas analysis, blood sugar and serum ions were performed in these preterm infants. All examinations were performed on a 3.0-T MRI system (Philips Intera Acheva Magnetom Vision) with echo-planar imaging capability with the use of a standard protocol. The imaging protocol for all the patients contained diffuse weighted images (EPI-SE, TR = 2144 ms, TE = 56 ms), T1-weighted images (TR = 389 ms; TE = 15 ms; slice thickness = 4 mm) as well as T2-weighted images (TR = 3035 ms; TE = 100 ms; slice thickness = 4 mm). The first MR examinations were performed in all these twelve preterm infants (mean age 4.5 days, range 2 - 7 days). Conventional MRI and DWI sequences obtained in the acute phase were compared. All the neonates underwent another two MRI examinations up to 2 and 4 weeks after delivery; five subjects also underwent MRI follow-up for up to 4 - 8 months (in 3 for 4 months, in 1 for 7 months, and in another for 8 months). Qualitative evaluations were performed to assess the presence of DWI changes compatible with PVL. RESULTS: The gestational ages of these twelve patients were from 31 to 35 weeks. None of them had intrauterine distress or birth asphyxia. None of the patients had localized neurological signs in the early course except for abnormal muscular tone of some extent, but seizure and apnea were their major symptoms. No other positive signs of nervous system was found in these preterm infants with PVL. First DWI detection (on the average of 4.5 days) in all these infants showed bilateral, symmetric, diffuse high signal intensity (including genu and plenum of corpus callosum), while conventional MRI showed normal images on both T1- and T2-weighted imaging; two weeks later, DWI showed irregularly high, low mixed signals while conventional MRI showed punctate high signal intensity on T1-weighted imaging and slightly lower signal on T2-weighted imaging. Four weeks later, DWI showed cystic low signal intensity where conventional MRI showed low signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging (cystic PVL). Four months later, DWI showed that the cystic cava became diminished and disappeared, while conventional MRI showed reduced cerebral white matter and dilation of ventricle. CONCLUSION: Bilateral, symmetric, diffuse high signal intensity on DWI seems to be the earliest evidence of PVL; diffusion-weighted imaging performed in the acute phase of the disease may have a higher correlation with later evidence of PVL than does conventional MR imaging. DWI is likely to be a considerable technique in the early assessment of white matter injury and later PVL in preterm infants.