Stroke severity classification
Pro Research Analysisby 
Searched over 200M research papers
Stroke Severity Classification: Methods and Key Tools
Stroke severity classification is crucial for diagnosis, treatment planning, and rehabilitation. Multiple approaches and tools are used to assess and categorize stroke severity, each with its own strengths and limitations.
EEG-Based Stroke Severity Classification
Recent research has explored the use of electroencephalogram (EEG) signals for classifying stroke severity. One approach uses a 1D convolutional neural network (CNN) to analyze the power spectral density of EEG recordings, successfully distinguishing between normal, mild, moderate, and severe stroke with high accuracy (up to 97.3%) using data from specific EEG channels . Another method relies on statistical features from EEG signals, such as mean absolute value, standard deviation, and variance, across different frequency sub-bands. These features can effectively identify stroke severity levels and may help guide patient-specific rehabilitation programs .
Clinical Scales and Standardized Measures
Traditional stroke severity classification often uses clinical scales. The National Institutes of Health Stroke Scale (NIHSS), Modified Rankin Scale (mRS), Barthel Index (BI), and Scandinavian Stroke Scale (SSS) are widely used to categorize stroke as mild, moderate, or severe 6810. Studies show that categorizing these scales retains most predictive information and that there is strong agreement among them, making them reliable for clinical and research use .
For pediatric stroke, the Pediatric Stroke Outcome Measure (PSOM) has been validated to classify neurological deficit severity into normal/mild, moderate, and severe categories, showing good agreement with other disability measures .
Gait and Motor Impairment-Based Classification
Gait analysis is another method for classifying stroke severity, especially in chronic stroke patients. Machine learning techniques using symmetric gait features and recursive feature elimination have achieved high accuracy (up to 96%) in distinguishing stroke severity levels, supporting their use in clinical diagnosis .
Upper limb motor impairment is also commonly used to classify severity. Standardized outcome measures like the Fugl-Meyer Upper Extremity section (FMA-UE) and Action Research Arm Test (ARAT) are frequently used, with consensus-based cutoffs for severe, moderate, and mild impairment .
Impairment-Based and Multidimensional Approaches
Some classification systems group stroke-related impairments into categories based on combinations of cognitive, motor, sensory, and consciousness deficits, providing a more nuanced view of stroke effects . However, there is significant variability in how "mild" stroke is defined and classified, with different studies using various assessment tools and cutoff points 6910. This inconsistency can affect treatment decisions and health service delivery.
Challenges and the Need for Consensus
Despite the availability of multiple tools and methods, there is no universal standard for classifying mild stroke, and definitions often vary across studies and clinical settings 6910. This lack of consensus can lead to under-recognition of long-term impairments, especially in those classified as having "mild" strokes, and highlights the need for more comprehensive and standardized classification strategies 6910.
Conclusion
Stroke severity classification relies on a range of methods, including EEG analysis, clinical scales, gait features, and impairment-based systems. While many tools show high accuracy and agreement, inconsistencies—especially in defining mild stroke—remain a challenge. Standardizing classification approaches will improve diagnosis, treatment, and rehabilitation outcomes for stroke survivors.
Sources and full results
Most relevant research papers on this topic
Stroke Severity Classification based on EEG Statistical Features
Stroke severity classification can be identified using statistical features like MA V, STD, and Variance, and EEG sub-band frequencies, potentially aiding in patient-specific rehabilitation programs.
DOI
Upper Limb Motor Impairment and Activity Limitation After Stroke: Standardising Severity Levels for Research in Rehabilitation
This study standardizes severity levels of upper limb motor impairment and activity limitation after stroke, improving research and clinical practice.
DOI