Multivariate prediction of functional outcome using lesion topography characterized by acute diffusion tensor imaging

Artificial intelligence applied in acute ischemic stroke: from child to elderly

This review will summarize artificial intelligence developments in acute ischemic stroke in recent years and forecasts for the future. Stroke is a major healthcare concern due to its effects on the patient's quality of life and its dependence on the timing of the identification as well as the treatment. In recent years, attention increased on the use of artificial intelligence (AI) systems to help categorize, prognosis, and to channel these patients toward the right therapeutic procedure. Machine learning (ML) and in particular deep learning (DL) systems using convoluted neural networks (CNN) are becoming increasingly popular. Various studies over the years evaluated the use of these methods of analysis and prediction in the assessment of stroke patients, and at the same time, several applications and software have been developed to support the neuroradiologists and the stroke team to improve patient outcomes.

Improved prediction of glioma-related aphasia by diffusion MRI metrics, machine learning, and automated fiber bundle segmentation

White matter impairments caused by gliomas can lead to functional disorders. In this study, we predicted aphasia in patients with gliomas infiltrating the language network using machine learning methods. We included 78 patients with left-hemispheric perisylvian gliomas. Aphasia was graded preoperatively using the Aachen aphasia test (AAT). Subsequently, we created bundle segmentations based on automatically generated tract orientation mappings using TractSeg. To prepare the input for the support vector machine (SVM), we first preselected aphasia-related fiber bundles based on the associations between relative tract volumes and AAT subtests. In addition, diffusion magnetic resonance imaging (dMRI)-based metrics [axial diffusivity (AD), apparent diffusion coefficient (ADC), fractional anisotropy (FA), and radial diffusivity (RD)] were extracted within the fiber bundles' masks with their mean, standard deviation, kurtosis, and skewness values. Our model consisted of random forest-based feature selection followed by an SVM. The best model performance achieved 81% accuracy (specificity = 85%, sensitivity = 73%, and AUC = 85%) using dMRI-based features, demographics, tumor WHO grade, tumor location, and relative tract volumes. The most effective features resulted from the arcuate fasciculus (AF), middle longitudinal fasciculus (MLF), and inferior fronto-occipital fasciculus (IFOF). The most effective dMRI-based metrics were FA, ADC, and AD. We achieved a prediction of aphasia using dMRI-based features and demonstrated that AF, IFOF, and MLF were the most important fiber bundles for predicting aphasia in this cohort.

Machine learning prediction of motor function in chronic stroke patients: a systematic review and meta-analysis

Background

Recent studies have reported that machine learning (ML), with a relatively strong capacity for processing non-linear data and adaptive ability, could improve the accuracy and efficiency of prediction. The article summarizes the published studies on ML models that predict motor function 3-6 months post-stroke.

Methods

A systematic literature search was conducted in PubMed, Embase, Cochorane and Web of Science as of April 3, 2023 for studies on ML prediction of motor function in stroke patients. The quality of the literature was assessed using the Prediction model Risk Of Bias Assessment Tool (PROBAST). A random-effects model was preferred for meta-analysis using R4.2.0 because of the different variables and parameters.

Results

A total of 44 studies were included in this meta-analysis, involving 72,368 patients and 136 models. Models were categorized into subgroups according to the predicted outcome Modified Rankin Scale cut-off value and whether they were constructed based on radiomics. C-statistics, sensitivity, and specificity were calculated. The random-effects model showed that the C-statistics of all models were 0.81 (95% CI: 0.79; 0.83) in the training set and 0.82 (95% CI: 0.80; 0.85) in the validation set. According to different Modified Rankin Scale cut-off values, C-statistics of ML models predicting Modified Rankin Scale>2(used most widely) in stroke patients were 0.81 (95% CI: 0.78; 0.84) in the training set, and 0.84 (95% CI: 0.81; 0.87) in the validation set. C-statistics of radiomics-based ML models in the training set and validation set were 0.81 (95% CI: 0.78; 0.84) and 0.87 (95% CI: 0.83; 0.90), respectively.

Conclusion

ML can be used as an assessment tool for predicting the motor function in patients with 3-6 months of post-stroke. Additionally, the study found that ML models with radiomics as a predictive variable were also demonstrated to have good predictive capabilities. This systematic review provides valuable guidance for the future optimization of ML prediction systems that predict poor motor outcomes in stroke patients.

Systematic review registration

https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022335260, identifier: CRD42022335260.

Interpretable deep learning for the prognosis of long-term functional outcome post-stroke using acute diffusion weighted imaging

Advances in deep learning can be applied to acute stroke imaging to build powerful and explainable prediction models that could supersede traditionally used biomarkers. We aimed to evaluate the performance and interpretability of a deep learning model based on convolutional neural networks (CNN) in predicting long-term functional outcome with diffusion-weighted imaging (DWI) acquired at day 1 post-stroke. Ischemic stroke patients (n = 322) were included from the ASTER and INSULINFARCT trials as well as the Pitié-Salpêtrière registry. We trained a CNN to predict long-term functional outcome assessed at 3 months with the modified Rankin Scale (dichotomized as good [mRS ≤ 2] vs. poor [mRS ≥ 3]) and compared its performance to two logistic regression models using lesion volume and ASPECTS. The CNN contained an attention mechanism, which allowed to visualize the areas of the brain that drove prediction. The deep learning model yielded a significantly higher area under the curve (0.83 95%CI [0.78-0.87]) than lesion volume (0.78 [0.73-0.83]) and ASPECTS (0.77 [0.71-0.83]) (p < 0.05). Setting all classifiers to the specificity as the deep learning model (i.e., 0.87 [0.82-0.92]), the CNN yielded a significantly higher sensitivity (0.67 [0.59-0.73]) than lesion volume (0.48 [0.40-0.56]) and ASPECTS (0.50 [0.41-0.58]) (p = 0.002). The attention mechanism revealed that the network learned to naturally attend to the lesion to predict outcome.

Precision medicine in stroke: towards personalized outcome predictions using artificial intelligence

Stroke ranks among the leading causes for morbidity and mortality worldwide. New and continuously improving treatment options such as thrombolysis and thrombectomy have revolutionized acute stroke treatment in recent years. Following modern rhythms, the next revolution might well be the strategic use of the steadily increasing amounts of patient-related data for generating models enabling individualized outcome predictions. Milestones have already been achieved in several health care domains, as big data and artificial intelligence have entered everyday life.

The aim of this review is to synoptically illustrate and discuss how artificial intelligence approaches may help to compute single-patient predictions in stroke outcome research in the acute, subacute and chronic stage. We will present approaches considering demographic, clinical and electrophysiological data, as well as data originating from various imaging modalities and combinations thereof. We will outline their advantages, disadvantages, their potential pitfalls and the promises they hold with a special focus on a clinical audience. Throughout the review we will highlight methodological aspects of novel machine-learning approaches as they are particularly crucial to realize precision medicine. We will finally provide an outlook on how artificial intelligence approaches might contribute to enhancing favourable outcomes after stroke.

Leveraging artificial intelligence in ischemic stroke imaging

Artificial intelligence (AI) is having a disruptive and transformative effect on clinical medicine. Prompt clinical diagnosis and imaging are critical for minimizing the morbidity and mortality associated with ischemic strokes. Clinicians must understand the current strengths and limitations of AI to provide optimal patient care. Ischemic stroke is one of the medical fields that have been extensively evaluated by artificial intelligence. Presented herein is a review of artificial intelligence applied to clinical management of stroke, geared toward clinicians. In this review, we explain the basic concept of AI and machine learning. This review is without coding and mathematical details and targets the clinicians involved in stroke management without any computer or mathematics' background. Here the AI application in ischemic stroke is summarized and classified into stroke imaging (automated diagnosis of brain infarction, automated ASPECT score calculation, infarction segmentation), prognosis prediction, and patients' selection for treatment.

Calhoun V, Grefkes C, Rost NS. Abnormal dynamic functional connectivity is linked to recovery after acute ischemic stroke

The aim of the current study was to explore the whole-brain dynamic functional connectivity patterns in acute ischemic stroke (AIS) patients and their relation to short and long-term stroke severity. We investigated resting-state functional MRI-based dynamic functional connectivity of 41 AIS patients two to five days after symptom onset. Re-occurring dynamic connectivity configurations were obtained using a sliding window approach and k-means clustering. We evaluated differences in dynamic patterns between three NIHSS-stroke severity defined groups (mildly, moderately, and severely affected patients). Furthermore, we built Bayesian hierarchical models to evaluate the predictive capacity of dynamic connectivity and examine the interrelation with clinical measures, such as white matter hyperintensity lesions. Finally, we established correlation analyses between dynamic connectivity and AIS severity as well as 90-day neurological recovery (ΔNIHSS). We identified three distinct dynamic connectivity configurations acutely post-stroke. More severely affected patients spent significantly more time in a configuration that was characterized by particularly strong connectivity and isolated processing of functional brain domains (three-level ANOVA: p < .05, post hoc t tests: p < .05, FDR-corrected). Configuration-specific time estimates possessed predictive capacity of stroke severity in addition to the one of clinical measures. Recovery, as indexed by the realized change of the NIHSS over time, was significantly linked to the dynamic connectivity between bilateral intraparietal lobule and left angular gyrus (Pearson's r = -.68, p = .003, FDR-corrected). Our findings demonstrate transiently increased isolated information processing in multiple functional domains in case of severe AIS. Dynamic connectivity involving default mode network components significantly correlated with recovery in the first 3 months poststroke.

Acute Diffusivity Biomarkers for Prediction of Motor and Language Outcome in Mild-to-Severe Stroke Patients

Background and Purpose—

Early severity of stroke symptoms—especially in mild-to-severe stroke patients—are imperfect predictors of long-term motor and aphasia outcome. Motor function and language processing heavily rely on the preservation of important white matter fasciculi in the brain. Axial diffusivity (AD) from the diffusion tensor imaging model has repeatedly shown to accurately reflect acute axonal damage and is thus optimal to probe the integrity of important white matter bundles and their relationship with long-term outcome. Our aim was to investigate the independent prognostic value of the AD of white matter tracts in the motor and language network evaluated at 24 hours poststroke for motor and aphasia outcome at 3 months poststroke.

Methods—

Seventeen (motor cohort) and 28 (aphasia cohort) thrombolyzed patients with initial mild-to-severe stroke underwent a diffusion tensor imaging sequence at 24 hours poststroke. Motor and language outcome were evaluated at 3 months poststroke with a composite motor score and the aphasia handicap scale. We first used stepwise regression to determine which classic (age, initial motor or aphasia severity, and lesion volume) and imaging (ratio of affected/unaffected AD of motor and language fasciculi) factors were related to outcome. Second, to determine the specificity of our a priori choices of fasciculi, we performed voxel-based analyses to determine if the same, additional, or altogether new regions were associated with long-term outcome.

Results—

The ratio of AD in the corticospinal tract was the sole predictor of long-term motor outcome, and the ratio of AD in the arcuate fasciculus—along with age and initial aphasia severity—was an independent predictor of 3-month aphasia outcome. White matter regions overlapping with these fasciculi naturally emerged in the corresponding voxel-based analyses.

Conclusions—

AD of the corticospinal tract and arcuate fasciculus are effective biomarkers of long-term motor and aphasia outcome, respectively.