Your browser doesn't support javascript.
loading
Show: 20 | 50 | 100
Results 1 - 3 de 3
Filter
Add more filters











Database
Language
Publication year range
1.
bioRxiv ; 2023 Jan 11.
Article in English | MEDLINE | ID: mdl-36711940

ABSTRACT

Dimension reduction tools preserving similarity and graph structure such as t-SNE and UMAP can capture complex biological patterns in high-dimensional data. However, these tools typically are not designed to separate effects of interest from unwanted effects due to confounders. We introduce the partial embedding (PARE) framework, which enables removal of confounders from any distance-based dimension reduction method. We then develop partial t-SNE and partial UMAP and apply these methods to genomic and neuroimaging data. Our results show that the PARE framework can remove batch effects in single-cell sequencing data as well as separate clinical and technical variability in neuroimaging measures. We demonstrate that the PARE framework extends dimension reduction methods to highlight biological patterns of interest while effectively removing confounding effects.

2.
J Neuroimaging ; 27(5): 447-452, 2017 09.
Article in English | MEDLINE | ID: mdl-28796432

ABSTRACT

BACKGROUND: Cortical pathology in multiple sclerosis (MS) has been associated with prolonged and progressive disease. 7T magnetic resonance imaging (MRI) provides enhanced visualization of cortical lesions (CLs). Hence, we conducted a pilot study to explore whether CLs occur early in MS, as evidenced by pediatric-onset patients. METHODS: A total of 8 pediatric-onset MS patients were imaged using 7T MRI. CLs were annotated on T1-weighted magnetization-prepared rapid acquisition of gradient echoes images as leukocortical (LC), intracortical, or subpial. Total CLs, age at onset, age at scan, disease duration, total relapses, and Expanded Disability Status Scale (EDSS) score were recorded. RESULTS: A median of 120 (range: 48-144) CLs was identified in 8 MS patients (3 female, all with relapsing remitting MS, mean age at scan 21 years ± 3.5 SD, mean age of disease onset 15 years ± 2.3 SD, mean disease duration 5.3 years ± 3.4 SD, median EDSS 2.0). Nearly all the lesions identified were LC. CONCLUSIONS: Many CLs are detectable using 7T MRI in patients with pediatric-onset MS despite relatively brief disease duration, absence of progressive disease, and very limited physical disability-supporting early cortical involvement in MS.


Subject(s)
Cerebral Cortex/diagnostic imaging , Magnetic Resonance Imaging/methods , Multiple Sclerosis, Relapsing-Remitting/diagnostic imaging , Adolescent , Adult , Cerebral Cortex/pathology , Female , Humans , Male , Multiple Sclerosis, Relapsing-Remitting/pathology , Pilot Projects , Time Factors , Young Adult
3.
Mult Scler ; 22(10): 1367-70, 2016 09.
Article in English | MEDLINE | ID: mdl-26552729

ABSTRACT

BACKGROUND: The infratentorial compartment is cardinal for multiple sclerosis (MS) diagnosis. T2-weighted (T2) and proton density-weighted (PD) magnetic resonance imaging (MRI) can visualize infratentorial lesions, but only suboptimally. OBJECTIVE: To combine PD and T2 for better lesion assessment. METHODS: T2 and PD from 35 cases were averaged to form "PT2" images. Two raters counted infratentorial lesions and qualitatively assessed their conspicuity. RESULTS: PT2 showed 244 infratentorial lesions, of which 94% and 74% were seen in PD and T2. PT2 received higher grades for image quality and lesion conspicuity (p < 0.001 for all comparisons). CONCLUSION: PT2 could improve our ability to diagnose and monitor MS.


Subject(s)
Brain/diagnostic imaging , Magnetic Resonance Imaging/methods , Multiple Sclerosis/diagnostic imaging , Adult , Female , Humans , Image Processing, Computer-Assisted , Male , Middle Aged
SELECTION OF CITATIONS
SEARCH DETAIL