Scholarlink

Background <p>Cognitive impairment is common in multiple sclerosis (MS), yet comprehensive cognitive assessment is not universally accessible, and patients need to be triaged for referral. Given the links between brain atrophy and cognition, this study investigated whether routinely collected neuroimaging markers could identify MS patients at risk of cognitive impairment.</p> Methods <p>Data were retrospectively analysed from adult MS patients assessed in a specialist cognitive neuroimmunology clinic, who had undergone MRI within 12 months prior to cognitive testing. Normalised brain volume (NBV), normalised grey matter volume (NGMV), normalised white matter volume (NWMV), and corpus callosum index (CCI) were measured using the Siemens MorphoBox automated software. GLMs were estimated to investigate group differences in brain volume metrics between cognitively impaired and non-impaired patients. ROC curves were estimated to investigate screening performance for neuroimaging metrics (Youden’s J). Results are expressed as parameter estimates with 95% bootstrapped confidence intervals (CI).</p> Results <p>120 patients were included (35% cognitively impaired). Cognitively impaired patients had lower NBV (<i>b</i> = − 2.06, 95% CI [− 3.35, − 0.81]) and NWMV (<i>b</i> = − 1.65, 95% CI [− 2.60, − 0.77]). NWMV (area under the curve [AUC] = 0.67, 95% CI [0.57, 0.76]) and CCI (AUC = 0.62, 95% CI [0.51, 0.72]) classified impairment, although sensitivity was low (&lt; 0.70). No clear associations or sufficient classification performance were observed for NGMV. Diagnostic performance improved when neuroimaging markers were statistically combined with relevant demographic information.</p> Conclusion <p>The present study did not find strong evidence supporting routinely collected neuroimaging as standalone cognitive screening tools. Classification performance improved when combined with demographic factors, but remained below thresholds for clinical utility. These findings highlight a gap between group-level associations reported in the literature and their translation to individual-level clinical application.</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

The use of routinely collected structural neuroimaging to identify cognitive impairment in multiple sclerosis

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月17日

查看全文
  • Reine E. Jardine,
  • Valeriya Kuznetsova,
  • Fiore D’Aprano,
  • Tomas Kalinicik,
  • Stefanie Roberts,
  • Charles B. Malpas

摘要

Background

Cognitive impairment is common in multiple sclerosis (MS), yet comprehensive cognitive assessment is not universally accessible, and patients need to be triaged for referral. Given the links between brain atrophy and cognition, this study investigated whether routinely collected neuroimaging markers could identify MS patients at risk of cognitive impairment.

Methods

Data were retrospectively analysed from adult MS patients assessed in a specialist cognitive neuroimmunology clinic, who had undergone MRI within 12 months prior to cognitive testing. Normalised brain volume (NBV), normalised grey matter volume (NGMV), normalised white matter volume (NWMV), and corpus callosum index (CCI) were measured using the Siemens MorphoBox automated software. GLMs were estimated to investigate group differences in brain volume metrics between cognitively impaired and non-impaired patients. ROC curves were estimated to investigate screening performance for neuroimaging metrics (Youden’s J). Results are expressed as parameter estimates with 95% bootstrapped confidence intervals (CI).

Results

120 patients were included (35% cognitively impaired). Cognitively impaired patients had lower NBV (b = − 2.06, 95% CI [− 3.35, − 0.81]) and NWMV (b = − 1.65, 95% CI [− 2.60, − 0.77]). NWMV (area under the curve [AUC] = 0.67, 95% CI [0.57, 0.76]) and CCI (AUC = 0.62, 95% CI [0.51, 0.72]) classified impairment, although sensitivity was low (< 0.70). No clear associations or sufficient classification performance were observed for NGMV. Diagnostic performance improved when neuroimaging markers were statistically combined with relevant demographic information.

Conclusion

The present study did not find strong evidence supporting routinely collected neuroimaging as standalone cognitive screening tools. Classification performance improved when combined with demographic factors, but remained below thresholds for clinical utility. These findings highlight a gap between group-level associations reported in the literature and their translation to individual-level clinical application.