<p>Post-stroke rehabilitation is a complex process influenced by several neurophysiological factors. The recovery is traditionally predicted based on initial impairment using linear models. The Proportional Recovery Rule (PRR), developed on the Fugl-Meyer scale, has even been proposed as a therapeutic target. In this framework, patients are classified as “fitters” or “non-fitters”, though this distinction depends on the methodology used. Additionally, issues like mathematical coupling and ceiling effects on clinical scales could raise concerns about the validity of these models. To overcome these issues, Repeated Spectral Clustering (RSC) was used to identify recovery patterns based on NIHSS. We selected 201 patients from the WAKE-UP trail, all moderately impaired at onset and still impaired at 22–36 h. Clustering was performed using a similarity matrix based on pairwise absolute differences between recovery ratios, calculated from 22–36 h to 90 days post-stroke. Cluster differences were tested with prognostic factors, including lesion volume, side, treatment, and the Heidelberg scale. The PRR was fit to the cohort for comparison with clustering results. The linear fit reproduced findings consistent with the literature, such as a correlation of <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\rho (x,\Delta ) = 0.73\)</EquationSource> </InlineEquation> and an average recovery ratio of 70% for the “fitters”. RSC grouped patients into six recovery clusters: <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(C_0\)</EquationSource> </InlineEquation> (full recovery), <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(C_1\)</EquationSource> </InlineEquation> (above average), <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(C_2\)</EquationSource> </InlineEquation> and <InlineEquation ID="IEq5"> <EquationSource Format="TEX">\(C_3\)</EquationSource> </InlineEquation> (average, PRR-aligned), <InlineEquation ID="IEq6"> <EquationSource Format="TEX">\(C_4\)</EquationSource> </InlineEquation> (below average), and <InlineEquation ID="IEq7"> <EquationSource Format="TEX">\(C_5\)</EquationSource> </InlineEquation> (deterioration). NIHSS scores in most patients declined non-proportionally. Lesion volume was not significantly different across clusters, while left-sided strokes were higher in low recovery clusters. Patients with a recovery ratio <InlineEquation ID="IEq8"> <EquationSource Format="TEX">\(\ge 0.3\)</EquationSource> </InlineEquation> within two weeks mostly fell into favorable clusters (<InlineEquation ID="IEq9"> <EquationSource Format="TEX">\(C_0\)</EquationSource> </InlineEquation>–<InlineEquation ID="IEq10"> <EquationSource Format="TEX">\(C_3\)</EquationSource> </InlineEquation>), covering <InlineEquation ID="IEq11"> <EquationSource Format="TEX">\(\approx 90\%\)</EquationSource> </InlineEquation> of such cases. The identified clusters provide a refined view of stroke recovery following wake-up stroke. Clustering better captures patient similarities, enabling the assessment of neurophysiological differences between groups and supporting tailored interventions.</p>

Beyond proportional recovery in wake-up stroke: unsupervised recovery clusters based on the NIHSS

期刊论文
出版模式：

开放获取
发表日期：

2026年1月29日

查看全文

Andrea Zanola,
Antonio Luigi Bisogno,
Veronika Vadinova,
Götz Thomalla,
Bastian Cheng,
Manfredo Atzori,
Maurizio Corbetta

摘要

Post-stroke rehabilitation is a complex process influenced by several neurophysiological factors. The recovery is traditionally predicted based on initial impairment using linear models. The Proportional Recovery Rule (PRR), developed on the Fugl-Meyer scale, has even been proposed as a therapeutic target. In this framework, patients are classified as “fitters” or “non-fitters”, though this distinction depends on the methodology used. Additionally, issues like mathematical coupling and ceiling effects on clinical scales could raise concerns about the validity of these models. To overcome these issues, Repeated Spectral Clustering (RSC) was used to identify recovery patterns based on NIHSS. We selected 201 patients from the WAKE-UP trail, all moderately impaired at onset and still impaired at 22–36 h. Clustering was performed using a similarity matrix based on pairwise absolute differences between recovery ratios, calculated from 22–36 h to 90 days post-stroke. Cluster differences were tested with prognostic factors, including lesion volume, side, treatment, and the Heidelberg scale. The PRR was fit to the cohort for comparison with clustering results. The linear fit reproduced findings consistent with the literature, such as a correlation of \(\rho (x,\Delta ) = 0.73\) and an average recovery ratio of 70% for the “fitters”. RSC grouped patients into six recovery clusters: \(C_0\) (full recovery), \(C_1\) (above average), \(C_2\) and \(C_3\) (average, PRR-aligned), \(C_4\) (below average), and \(C_5\) (deterioration). NIHSS scores in most patients declined non-proportionally. Lesion volume was not significantly different across clusters, while left-sided strokes were higher in low recovery clusters. Patients with a recovery ratio \(\ge 0.3\) within two weeks mostly fell into favorable clusters ( \(C_0\) – \(C_3\) ), covering \(\approx 90\%\) of such cases. The identified clusters provide a refined view of stroke recovery following wake-up stroke. Clustering better captures patient similarities, enabling the assessment of neurophysiological differences between groups and supporting tailored interventions.