Scholarlink

<p>Multiple Sclerosis (MS) is a chronic autoimmune inflammatory disorder of the central nervous system with substantial geographic variation in prevalence, underscoring the need for region-specific, standardized imaging datasets. Despite its clinical significance, imaging datasets from the Middle East—particularly Iraq—remain scarce. This study introduces the Multiple Sclerosis Imaging Dataset (MSID), comprising MRI scans and clinical metadata from patients diagnosed between 2007–2024 across three cities in the Kurdistan Region of Iraq (Sulaymaniyah, Erbil, and Duhok). Data collection was conducted under the approval of the Kurdistan Regional Government’s Ministry of Health, resulting in 52,190 DICOM images, of which 38,105 were standardized for analysis. Imaging modalities include T1-weighted, T2-weighted, and FLAIR sequences. Image quality metrics—Signal-to-Noise Ratio (SNR), Contrast-to-Noise Ratio (CNR), and sharpness—were quantified using Python-based methods. Clinical type information was available for a subset of 43 patients; a chi-square goodness-of-fit test confirmed consistency with established global proportions (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\chi ^{2} = 5.81\)</EquationSource> </InlineEquation>, <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(p = 0.058\)</EquationSource> </InlineEquation>), with relapsing-remitting MS as the predominant form. All DICOM scans were anonymized, spatially standardized, and converted to JPEG, with automated metadata extraction. Regional analysis revealed variations in lesion distribution, potentially linked to environmental or genetic factors. The current release does not include expert lesion segmentations; accordingly, MSID is primarily intended to support imaging-based analyses, automated classification model development, and regional MS characterization rather than supervised lesion segmentation training. As one of the most extensive MS resources in the region, MSID supports regional epidemiological analyses, imaging-based disease classification, and comparative studies with global MS populations, filling a previously unrepresented region in current neuroimaging datasets.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multiple Sclerosis Imaging Dataset (MSID): a standardized multi-center MRI resource

  • 期刊论文

  • 发表日期：

    2026年6月3日

查看全文
  • Mohammed Y. Shakor,
  • Mustafa Ibrahim Khaleel

摘要

Multiple Sclerosis (MS) is a chronic autoimmune inflammatory disorder of the central nervous system with substantial geographic variation in prevalence, underscoring the need for region-specific, standardized imaging datasets. Despite its clinical significance, imaging datasets from the Middle East—particularly Iraq—remain scarce. This study introduces the Multiple Sclerosis Imaging Dataset (MSID), comprising MRI scans and clinical metadata from patients diagnosed between 2007–2024 across three cities in the Kurdistan Region of Iraq (Sulaymaniyah, Erbil, and Duhok). Data collection was conducted under the approval of the Kurdistan Regional Government’s Ministry of Health, resulting in 52,190 DICOM images, of which 38,105 were standardized for analysis. Imaging modalities include T1-weighted, T2-weighted, and FLAIR sequences. Image quality metrics—Signal-to-Noise Ratio (SNR), Contrast-to-Noise Ratio (CNR), and sharpness—were quantified using Python-based methods. Clinical type information was available for a subset of 43 patients; a chi-square goodness-of-fit test confirmed consistency with established global proportions ( \(\chi ^{2} = 5.81\) , \(p = 0.058\) ), with relapsing-remitting MS as the predominant form. All DICOM scans were anonymized, spatially standardized, and converted to JPEG, with automated metadata extraction. Regional analysis revealed variations in lesion distribution, potentially linked to environmental or genetic factors. The current release does not include expert lesion segmentations; accordingly, MSID is primarily intended to support imaging-based analyses, automated classification model development, and regional MS characterization rather than supervised lesion segmentation training. As one of the most extensive MS resources in the region, MSID supports regional epidemiological analyses, imaging-based disease classification, and comparative studies with global MS populations, filling a previously unrepresented region in current neuroimaging datasets.