Scholarlink

<p>The Mini Mental State Examination (MMSE) is widely used for first-line cognitive screening. However, the common practice of applying a single cut off score across diverse populations in the MMSE ignores heterogeneity in age, education and language, producing inconsistent thresholds and referral decisions. To integrate the human factors in an objective way, we develop a human factor integrated MMSE (hMMSE) framework that synthesizes the total score, item response profile and routinely available human factors to deliver individualized risk estimates in cognitive screening while preserving standard administration. Using a multi-site cohort from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), the integrated model reduces unnecessary referrals among cognitively normal (CN) individuals by 0.9% (23 of 2615) and reduces missed cognitive impairment by 7.1% (282 of 3965). Improvements are driven primarily by better detection of mild cognitive impairment (MCI) and are accompanied by more accurate downstream differentiation of Alzheimer’s disease (AD) from MCI. Decision curve analysis shows higher net benefit across clinically relevant risk thresholds, supporting flexible local operating points without ad hoc demographic adjustment. Computational calibration can therefore convert the MMSE from a static cut off approach into a personalized decision support tool for precision screening.</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

Human factor integrated MMSE for improved cognitive screening

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月8日

查看全文
  • Meiwei Zhang,
  • Wenyuan Li,
  • Caiyun Hong,
  • Qiushi Cui,
  • Anlong Sun,
  • Honglian Jia,
  • Yuwei Pan

摘要

The Mini Mental State Examination (MMSE) is widely used for first-line cognitive screening. However, the common practice of applying a single cut off score across diverse populations in the MMSE ignores heterogeneity in age, education and language, producing inconsistent thresholds and referral decisions. To integrate the human factors in an objective way, we develop a human factor integrated MMSE (hMMSE) framework that synthesizes the total score, item response profile and routinely available human factors to deliver individualized risk estimates in cognitive screening while preserving standard administration. Using a multi-site cohort from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), the integrated model reduces unnecessary referrals among cognitively normal (CN) individuals by 0.9% (23 of 2615) and reduces missed cognitive impairment by 7.1% (282 of 3965). Improvements are driven primarily by better detection of mild cognitive impairment (MCI) and are accompanied by more accurate downstream differentiation of Alzheimer’s disease (AD) from MCI. Decision curve analysis shows higher net benefit across clinically relevant risk thresholds, supporting flexible local operating points without ad hoc demographic adjustment. Computational calibration can therefore convert the MMSE from a static cut off approach into a personalized decision support tool for precision screening.