Machine Learning Approach to Identify Stroke Within 4.5 Hours

Overcoming big bottlenecks in vascular regeneration

Bioengineering and regenerative medicine strategies are promising for the treatment of vascular diseases. However, current limitations inhibit the ability of these approaches to be translated to clinical practice. Here we summarize some of the big bottlenecks that inhibit vascular regeneration in the disease applications of aortic aneurysms, stroke, and peripheral artery disease. We also describe the bottlenecks preventing three-dimensional bioprinting of vascular networks for tissue engineering applications. Finally, we describe emerging technologies and opportunities to overcome these challenges to advance vascular regeneration.

Artificial intelligence in ischemic stroke images: current applications and future directions

This paper reviews the current research progress in the application of Artificial Intelligence (AI) based on ischemic stroke imaging, analyzes the main challenges, and explores future research directions. This study emphasizes the application of AI in areas such as automatic segmentation of infarct areas, detection of large vessel occlusion, prediction of stroke outcomes, assessment of hemorrhagic transformation risk, forecasting of recurrent ischemic stroke risk, and automatic grading of collateral circulation. The research indicates that Machine Learning (ML) and Deep Learning (DL) technologies have tremendous potential for improving diagnostic accuracy, accelerating disease identification, and predicting disease progression and treatment responses. However, the clinical application of these technologies still faces challenges such as limitations in data volume, model interpretability, and the need for real-time monitoring and updating. Additionally, this paper discusses the prospects of applying large language models, such as the transformer architecture, in ischemic stroke imaging analysis, emphasizing the importance of establishing large public databases and the need for future research to focus on the interpretability of algorithms and the comprehensiveness of clinical decision support. Overall, AI has significant application value in the management of ischemic stroke; however, existing technological and practical challenges must be overcome to achieve its widespread application in clinical practice.

Posterior circulation ischemic stroke: radiomics-based machine learning approach to identify onset time from magnetic resonance imaging

PURPOSE

Posterior circulation ischemic stroke (PCIS) possesses unique features. However, previous studies have primarily or exclusively relied on anterior circulation stroke cases to build machine learning (ML) models for predicting onset time. To date, there is no research reporting the effectiveness and stability of ML in identifying PCIS onset time. We aimed to build diffusion-weighted imaging-based ML models to identify the onset time of PCIS patients.

METHODS

Consecutive PCIS patients within 24 h of definite symptom onset were included (112 in the training set and 49 in the independent test set). Images were processed as follows: volume of interest segmentation, image feature extraction, and feature selection. Five ML models, naïve Bayes, logistic regression, tree ensemble, k-nearest neighbor, and random forest, were built based on the training set to estimate the stroke onset time (binary classification: ≤ 4.5 h or > 4.5 h). Relative standard deviations (RSD), receiver operating characteristic (ROC) curves, and the calibration plot was performed to evaluate the stability and performance of the five models.

RESULTS

The random forest model had the best performance in the test set, with the highest area under the curve (AUC, 0.840; 95% CI: 0.706, 0.974). This model also achieved the highest accuracy, sensitivity, specificity, positive predictive value, and negative predictive value (83.7%, 64.3%, 91.4%, 75.0%, and 86.5%, respectively). Furthermore, the model had high stability (RSD = 0.0094).

CONCLUSION

The PCIS case-based ML model was effective for estimating the symptom onset time and achieved considerably high specificity and stability.

Digital symbol-digit modalities test with modified flexible protocols in patients with CNS demyelinating diseases

Cognitive impairment (CI) is prevalent in central nervous system demyelinating diseases, such as multiple sclerosis (MS) and neuromyelitis optica spectrum disorders (NMOSD). We developed a novel tablet-based modified digital Symbol Digit Modalities Test (MD-SDMT) with adjustable protocols that feature alternating symbol-digit combinations in each trial, lasting one or two minutes. We assessed 144 patients (99 with MS and 45 with NMOSD) using both MD-SDMT protocols and the traditional paper-based SDMT. We also gathered participants' feedback through a questionnaire regarding their preferences and perceived reliability. The results showed strong correlations between MD-SDMT and paper-based SDMT scores (Pearsons correlation: 0.88 for 2 min; 0.85 for 1 min, both p < 0.001). Among the 120 respondents, the majority preferred the digitalized SDMT (55% for the 2 min, 39% for the 1 min) over the paper-based version (6%), with the 2 min MD-SDMT reported as the most reliable test. Notably, patients with NMOSD and older individuals exhibited a preference for the paper-based test, as compared to those with MS and younger patients. In summary, even with short test durations, the digitalized SDMT effectively evaluates cognitive function in MS and NMOSD patients, and is generally preferred over the paper-based method, although preferences may vary with patient characteristics.

Ischemia-reperfusion injury in a salvaged penumbra: Longitudinal high-tesla perfusion magnetic resonance imaging in a rat model

INTRODUCTION

Although ischemia-reperfusion (I/R) injury varies between cortical and subcortical regions, its effects on specific regions remain unclear. In this study, we used various magnetic resonance imaging (MRI) techniques to examine the spatiotemporal dynamics of I/R injury within the salvaged ischemic penumbra (IP) and reperfused ischemic core (IC) of a rodent model, with the aim of enhancing therapeutic strategies by elucidating these dynamics.

MATERIALS AND METHODS

A total of 17 Sprague-Dawley rats were subjected to 1 h of transient middle cerebral artery occlusion with a suture model. MRI, including diffusion tensor imaging (DTI), T2-weighted imaging, perfusion-weighted imaging, and T1 mapping, was conducted at multiple time points for up to 5 days during the I/R phases. The spatiotemporal dynamics of blood-brain barrier (BBB) modifications were characterized through changes in T1 within the IP and IC regions and compared with mean diffusivity (MD), T2, and cerebral blood flow.

RESULTS

During the I/R phases, the MD of the IC initially decreased, normalized after recanalization, decreased again at 24 h, and peaked on day 5. By contrast, the IP remained relatively stable. Both the IP and IC exhibited hyperperfusion, with the IP reaching its peak at 24 h, followed by resolution, whereas hyperperfusion was maintained in the IC until day 5. Despite hyperperfusion, the IP maintained an intact BBB, whereas the IC experienced persistent BBB leakage. At 24 h, the IC exhibited an increase in the T2 signal, corresponding to regions exhibiting BBB disruption at 5 days.

CONCLUSIONS

Hyperperfusion and BBB impairment have distinct patterns in the IP and IC. Quantitative T1 mapping may serve as a supplementary tool for the early detection of malignant hyperemia accompanied by BBB leakage, aiding in precise interventions after recanalization. These findings underscore the value of MRI markers in monitoring ischemia-specific regions and customizing therapeutic strategies to improve patient outcomes.

Artificial Intelligence and Machine Learning Applications in Critically Ill Brain Injured Patients

The utilization of Artificial Intelligence (AI) and Machine Learning (ML) is paving the way for significant strides in patient diagnosis, treatment, and prognostication in neurocritical care. These technologies offer the potential to unravel complex patterns within vast datasets ranging from vast clinical data and EEG (electroencephalogram) readings to advanced cerebral imaging facilitating a more nuanced understanding of patient conditions. Despite their promise, the implementation of AI and ML faces substantial hurdles. Historical biases within training data, the challenge of interpreting multifaceted data streams, and the "black box" nature of ML algorithms present barriers to widespread clinical adoption. Moreover, ethical considerations around data privacy and the need for transparent, explainable models remain paramount to ensure trust and efficacy in clinical decision-making.This article reflects on the emergence of AI and ML as integral tools in neurocritical care, discussing their roles from the perspective of both their scientific promise and the associated challenges. We underscore the importance of extensive validation in diverse clinical settings to ensure the generalizability of ML models, particularly considering their potential to inform critical medical decisions such as withdrawal of life-sustaining therapies. Advancement in computational capabilities is essential for implementing ML in clinical settings, allowing for real-time analysis and decision support at the point of care. As AI and ML are poised to become commonplace in clinical practice, it is incumbent upon health care professionals to understand and oversee these technologies, ensuring they adhere to the highest safety standards and contribute to the realization of personalized medicine. This engagement will be pivotal in integrating AI and ML into patient care, optimizing outcomes in neurocritical care through informed and data-driven decision-making.

Estimating the volume of penumbra in rodents using DTI and stack-based ensemble machine learning framework

BACKGROUND

This study investigates the potential of diffusion tensor imaging (DTI) in identifying penumbral volume (PV) compared to the standard gadolinium-required perfusion-diffusion mismatch (PDM), utilizing a stack-based ensemble machine learning (ML) approach with enhanced explainability.

METHODS

Sixteen male rats were subjected to middle cerebral artery occlusion. The penumbra was identified using PDM at 30 and 90 min after occlusion. We used 11 DTI-derived metrics and 14 distance-based features to train five voxel-wise ML models. The model predictions were integrated using stack-based ensemble techniques. ML-estimated and PDM-defined PVs were compared to evaluate model performance through volume similarity assessment, the Pearson correlation analysis, and Bland-Altman analysis. Feature importance was determined for explainability.

RESULTS

In the test rats, the ML-estimated median PV was 106.4 mL (interquartile range 44.6-157.3 mL), whereas the PDM-defined median PV was 102.0 mL (52.1-144.9 mL). These PVs had a volume similarity of 0.88 (0.79-0.96), a Pearson correlation coefficient of 0.93 (p < 0.001), and a Bland-Altman bias of 2.5 mL (2.4% of the mean PDM-defined PV), with 95% limits of agreement ranging from -44.9 to 49.9 mL. Among the features used for PV prediction, the mean diffusivity was the most important feature.

CONCLUSIONS

Our study confirmed that PV can be estimated using DTI metrics with a stack-based ensemble ML approach, yielding results comparable to the volume defined by the standard PDM. The model explainability enhanced its clinical relevance. Human studies are warranted to validate our findings.

RELEVANCE STATEMENT

The proposed DTI-based ML model can estimate PV without the need for contrast agent administration, offering a valuable option for patients with kidney dysfunction. It also can serve as an alternative if perfusion map interpretation fails in the clinical setting.

KEY POINTS

• Penumbral volume can be estimated by DTI combined with stack-based ensemble ML. • Mean diffusivity was the most important feature used for predicting penumbral volume. • The proposed approach can be beneficial for patients with kidney dysfunction.

Automated Assessment of the DWI-FLAIR Mismatch in Patients with Acute Ischemic Stroke: Added Value to Routine Clinical Practice

BACKGROUND AND PURPOSE

The DWI-FLAIR mismatch is used to determine thrombolytic eligibility in patients with acute ischemic stroke when the time since stroke onset is unknown. Commercial software packages have been developed for automated DWI-FLAIR classification. We aimed to use e-Stroke software for automated classification of the DWI-FLAIR mismatch in a cohort of patients with acute ischemic stroke and in a comparative analysis with 2 expert neuroradiologists.

MATERIALS AND METHODS

In this retrospective study, patients with acute ischemic stroke who had MR imaging and known time since stroke onset were included. The DWI-FLAIR mismatch was evaluated by 2 neuroradiologists blinded to the time since stroke onset and automatically by the e-Stroke software. After 4 weeks, the neuroradiologists re-evaluated the MR images, this time equipped with automated predicted e-Stroke results as a computer-assisted tool. Diagnostic performances of e-Stroke software and the neuroradiologists were evaluated for prediction of DWI-FLAIR mismatch status.

RESULTS

A total of 157 patients met the inclusion criteria. A total of 82 patients (52%) had a time since stroke onset of ≤4.5 hours. By means of consensus reads, 81 patients (51.5%) had a DWI-FLAIR mismatch. The diagnostic accuracy (area under the curve/sensitivity/specificity) of e-Stroke software for the determination of the DWI-FLAIR mismatch was 0.72/90.0/53.9. The diagnostic accuracy (area under the curve/sensitivity/specificity) for neuroradiologists 1 and 2 was 0.76/69.1/84.2 and 0.82/91.4/73.7, respectively; both significantly (P < .05) improved to 0.83/79.0/86.8 and 0.89/92.6/85.5, respectively, following the use of e-Stroke predictions as a computer-assisted tool. The interrater agreement (κ) for determination of DWI-FLAIR status was improved from 0.49 to 0.57 following the use of the computer-assisted tool.

CONCLUSIONS

This automated quantitative approach for DWI-FLAIR mismatch provides results comparable with those of human experts and can improve the diagnostic accuracies of expert neuroradiologists in the determination of DWI-FLAIR status.

Contributions of Machine Learning in the Management of Stroke: A Bibliometric Analysis of the 50 Most Cited Articles

BACKGROUND

Though currently considered a 'black box,' machine learning (ML) has a promising future to ameliorate the health-care burden of stroke which is the second leading cause of mortality worldwide. Through this study, we sought to review the most influential articles on the applications of ML in stroke.

METHODS

Web of Sciences database was searched, and a list of the top 50 most cited articles, assessing the application of ML in stroke, was prepared by 2 authors, independently. Subsequently, a detailed analysis was performed to characterize the most impactful studies.

RESULTS

The total number of citations to the top 50 articles were 2959 (range 35-243 citations) with a median of 47 citations. Highest number of articles were published in the journal Stroke and the United States was the major contributing country. The majority of the studies focused on the utilization of ML to improve stroke risk prediction, diagnosis, and outcome prediction. Statistical analysis revealed an insignificant association between the total and mean number of citations and the impact factor of the journal (P = 0.516 and 0.987, respectively).

CONCLUSIONS

Recent years have witnessed a surge in the application of ML in stroke, with an enhancement in interest and funding over the years. ML has revolutionized the management of stroke and continues to aid in the neurosurgical decision-making and care in stroke patients.

Biochemical, biomechanical and imaging biomarkers of ischemic stroke: Time for integrative thinking

Stroke is one of the leading causes of adult disability affecting millions of people worldwide. Post-stroke cognitive and motor impairments diminish quality of life and functional independence. There is an increased risk of having a second stroke and developing secondary conditions with long-term social and economic impacts. With increasing number of stroke incidents, shortage of medical professionals and limited budgets, health services are struggling to provide a care that can break the vicious cycle of stroke. Effective post-stroke recovery hinges on holistic, integrative and personalized care starting from improved diagnosis and treatment in clinics to continuous rehabilitation and support in the community. To improve stroke care pathways, there have been growing efforts in discovering biomarkers that can provide valuable insights into the neural, physiological and biomechanical consequences of stroke and how patients respond to new interventions. In this review paper, we aim to summarize recent biomarker discovery research focusing on three modalities (brain imaging, blood sampling and gait assessments), look at some established and forthcoming biomarkers, and discuss their usefulness and complementarity within the context of comprehensive stroke care. We also emphasize the importance of biomarker guided personalized interventions to enhance stroke treatment and post-stroke recovery.

Diffusion-/perfusion-weighted imaging fusion to automatically identify stroke within 4.5 h

OBJECTIVES

We aimed to develop machine learning (ML) models based on diffusion- and perfusion-weighted imaging fusion (DP fusion) for identifying stroke within 4.5 h, to compare them with DWI- and/or PWI-based ML models, and to construct an automatic segmentation-classification model and compare with manual labeling methods.

METHODS

ML models were developed from multimodal MRI datasets of acute stroke patients within 24 h of clear symptom onset from two centers. The processes included manual segmentation, registration, DP fusion, feature extraction, and model establishment (logistic regression (LR) and support vector machine (SVM)). A segmentation-classification model (X-Net) was proposed for automatically identifying stroke within 4.5 h. The area under the receiver operating characteristic curve (AUC), sensitivity, Dice coefficients, decision curve analysis, and calibration curves were used to evaluate model performance.

RESULTS

A total of 418 patients (≤ 4.5 h: 214; > 4.5 h: 204) were evaluated. The DP fusion model achieved the highest AUC in identifying the onset time in the training (LR: 0.95; SVM: 0.92) and test sets (LR: 0.91; SVM: 0.90). The DP fusion-LR model displayed consistent positive and greater net benefits than other models across a broad range of risk thresholds. The calibration curve demonstrated the good calibration of the DP fusion-LR model (average absolute error: 0.049). The X-Net model obtained the highest Dice coefficients (DWI: 0.81; Tmax: 0.83) and achieved similar performance to manual labeling (AUC: 0.84).

CONCLUSIONS

The automatic segmentation-classification models based on DWI and PWI fusion images had high performance in identifying stroke within 4.5 h.

CLINICAL RELEVANCE STATEMENT

Perfusion-weighted imaging (PWI) fusion images had high performance in identifying stroke within 4.5 h. The automatic segmentation-classification models based on DWI and PWI fusion images could provide clinicians with decision-making guidance for acute stroke patients with unknown onset time.

KEY POINTS

• The diffusion/perfusion-weighted imaging fusion model had the best performance in identifying stroke within 4.5 h. • The X-Net model had the highest Dice and achieved performance close to manual labeling in segmenting lesions of acute stroke. • The automatic segmentation-classification model based on DP fusion images performed well in identifying stroke within 4.5 h.

Machine learning-based gait analysis to predict clinical frailty scale in elderly patients with heart failure

Aims

Although frailty assessment is recommended for guiding treatment strategies and outcome prediction in elderly patients with heart failure (HF), most frailty scales are subjective, and the scores vary among raters. We sought to develop a machine learning-based automatic rating method/system/model of the clinical frailty scale (CFS) for patients with HF.

Methods and results

We prospectively examined 417 elderly (≥75 years) with symptomatic chronic HF patients from 7 centres between January 2019 and October 2023. The patients were divided into derivation (n = 194) and validation (n = 223) cohorts. We obtained body-tracking motion data using a deep learning-based pose estimation library, on a smartphone camera. Predicted CFS was calculated from 128 key features, including gait parameters, using the light gradient boosting machine (LightGBM) model. To evaluate the performance of this model, we calculated Cohen's weighted kappa (CWK) and intraclass correlation coefficient (ICC) between the predicted and actual CFSs. In the derivation and validation datasets, the LightGBM models showed excellent agreements between the actual and predicted CFSs [CWK 0.866, 95% confidence interval (CI) 0.807-0.911; ICC 0.866, 95% CI 0.827-0.898; CWK 0.812, 95% CI 0.752-0.868; ICC 0.813, 95% CI 0.761-0.854, respectively]. During a median follow-up period of 391 (inter-quartile range 273-617) days, the higher predicted CFS was independently associated with a higher risk of all-cause death (hazard ratio 1.60, 95% CI 1.02-2.50) after adjusting for significant prognostic covariates.

Conclusion

Machine learning-based algorithms of automatically CFS rating are feasible, and the predicted CFS is associated with the risk of all-cause death in elderly patients with HF.

Medical image identification methods: A review

The identification of medical images is an essential task in computer-aided diagnosis, medical image retrieval and mining. Medical image data mainly include electronic health record data and gene information data, etc. Although intelligent imaging provided a good scheme for medical image analysis over traditional methods that rely on the handcrafted features, it remains challenging due to the diversity of imaging modalities and clinical pathologies. Many medical image identification methods provide a good scheme for medical image analysis. The concepts pertinent of methods, such as the machine learning, deep learning, convolutional neural networks, transfer learning, and other image processing technologies for medical image are analyzed and summarized in this paper. We reviewed these recent studies to provide a comprehensive overview of applying these methods in various medical image analysis tasks, such as object detection, image classification, image registration, segmentation, and other tasks. Especially, we emphasized the latest progress and contributions of different methods in medical image analysis, which are summarized base on different application scenarios, including classification, segmentation, detection, and image registration. In addition, the applications of different methods are summarized in different application area, such as pulmonary, brain, digital pathology, brain, skin, lung, renal, breast, neuromyelitis, vertebrae, and musculoskeletal, etc. Critical discussion of open challenges and directions for future research are finally summarized. Especially, excellent algorithms in computer vision, natural language processing, and unmanned driving will be applied to medical image recognition in the future.

Determining acute ischemic stroke onset time using machine learning and radiomics features of infarct lesions and whole brain

Accurate determination of the onset time in acute ischemic stroke (AIS) patients helps to formulate more beneficial treatment plans and plays a vital role in the recovery of patients. Considering that the whole brain may contain some critical information, we combined the Radiomics features of infarct lesions and whole brain to improve the prediction accuracy. First, the radiomics features of infarct lesions and whole brain were separately calculated using apparent diffusion coefficient (ADC), diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) sequences of AIS patients with clear onset time. Then, the least absolute shrinkage and selection operator (Lasso) was used to select features. Four experimental groups were generated according to combination strategies: Features in infarct lesions (IL), features in whole brain (WB), direct combination of them (IW) and Lasso selection again after direct combination (IWS), which were used to evaluate the predictive performance. The results of ten-fold cross-validation showed that IWS achieved the best AUC of 0.904, which improved by 13.5% compared with IL (0.769), by 18.7% compared with WB (0.717) and 4.2% compared with IW (0.862). In conclusion, combining infarct lesions and whole brain features from multiple sequences can further improve the accuracy of AIS onset time.

An Automatic DWI/FLAIR Mismatch Assessment of Stroke Patients

DWI/FLAIR mismatch assessment for ischemic stroke patients shows promising results in determining if patients are eligible for recombinant tissue-type plasminogen activator (r-tPA) treatment. However, the mismatch criteria suffer from two major issues: binary classification of a non-binary problem and the subjectiveness of the assessor. In this article, we present a simple automatic method for segmenting stroke-related parenchymal hyperintensities on FLAIR, allowing for an automatic and continuous DWI/FLAIR mismatch assessment. We further show that our method's segmentations have comparable inter-rater agreement (DICE 0.820, SD 0.12) compared to that of two neuro-radiologists (DICE 0.856, SD 0.07), that our method appears robust to hyper-parameter choices (suggesting good generalizability), and lastly, that our methods continuous DWI/FLAIR mismatch assessment correlates to mismatch assessments made for a cohort of wake-up stroke patients at hospital submission. The proposed method shows promising results in automating the segmentation of parenchymal hyperintensity within ischemic stroke lesions and could help reduce inter-observer variability of DWI/FLAIR mismatch assessment performed in clinical environments as well as offer a continuous assessment instead of the current binary one.

Automatic triaging of acute ischemic stroke patients for reperfusion therapies using Artificial Intelligence methods and multiple MRI features: A review

BACKGROUND

The selection of appropriate treatments for Acute Ischemic Stroke (AIS), including Intravenous (IV) tissue plasminogen activator (tPA) and Mechanical thrombectomy, is a critical aspect of clinical decision-making. Timely treatment is essential, with recommended administration of therapies within 4.5 h of symptom onset. However, patients with unknown Time Since Stroke (TSS), are often excluded from thrombolysis, even if the stroke onset exceeds 6 h. Current clinical guidelines propose using multimodal Magnetic Resonance Imaging (MRI) to assess various mismatches.

METHODS

The review explores the significance of automatic methods based on Artificial Intelligence (AI) algorithms that utilize multiple MRI features to identify patients who are most likely to benefit from acute reperfusion therapies. These AI methods include TSS classification and patient selection for therapies in the late time window (>6 h) using MRI images to provide detailed stroke information.

RESULTS

The review discusses the challenges and limitations in the existing mismatch methods, which may lead to missed opportunities for reperfusion therapy. To address these limitations, AI approaches have been developed to enhance accuracy and support clinical decision-making. These AI methods have shown promising results, outperforming traditional mismatch assessments and providing improved sensitivity and specificity in identifying patients eligible for reperfusion therapies.

DISCUSSION

In summary, the integration of AI algorithms utilizing multiple MRI features has the potential to enhance accuracy, improve patient outcomes, and positively influence the decision-making process in AIS. However, ongoing research and collaboration among clinicians, researchers, and technologists are vital to realize the full potential of AI in optimizing stroke management.

A deep learning analysis of stroke onset time prediction and comparison to DWI-FLAIR mismatch

INTRODUCTION

When time since stroke onset is unknown, DWI-FLAIR mismatch rating is an established technique for patient stratification. A visible DWI lesion without corresponding parenchymal hyperintensity on FLAIR suggests time since onset of under 4.5 h and thus a potential benefit from intravenous thrombolysis. To improve accuracy and availability of the mismatch concept, deep learning might be able to augment human rating and support decision-making in these cases.

METHODS

We used unprocessed DWI and coregistered FLAIR imaging data to train a deep learning model to predict dichotomized time since ischemic stroke onset. We analyzed the performance of Group Convolutional Neural Networks compared to other deep learning methods. Unlabeled imaging data was used for pre-training. Prediction performance of the best deep learning model was compared to the performance of four independent junior and senior raters. Additionally, in cases deemed indeterminable by human raters, model ratings were used to augment human performance. Post-hoc gradient-based explanations were analyzed to gain insights into model predictions.

RESULTS

Our best predictive model performed comparably to human raters. Using model ratings in cases deemed indeterminable by human raters improved rating accuracy and interrater agreement for junior and senior ratings. Post-hoc explainability analyses showed that the model localized stroke lesions to derive predictions.

DISCUSSION

Our analysis shows that deep learning based clinical decision support has the potential to improve the accessibility of the DWI-FLAIR mismatch concept by supporting patient stratification.

Computed Tomography Angiography-Based Thrombus Radiomics for Predicting the Time Since Stroke Onset

RATIONALE AND OBJECTIVES

The measurement of the time since stroke onset (TSS) is crucial for decision-making in the treatment of acute ischemic stroke (AIS). This study assessed the utility of computed tomography angiography (CTA) radiomics features (RFs) to estimate TSS.

MATERIALS AND METHODS

A total of 221 patients with AIS were enrolled in this retrospective study and were divided into a training group (n = 154) and a test group (n = 67). Thrombi in CTA images were manually outlined using ITK-SNAP. Images were aligned, normalized, and pre-processed to extract RFs. The TSS was calculated as the time from stroke onset to CTA completion. The patients were classified into two groups according to estimated TSS: ≤4.5 and >4.5 hours. A total of 944 RFs were extracted from CTA images. Clinical factors associated with TSS were identified using multivariate logistic regression, and a combined model (clinical data and RFs) was constructed. The predictive value of the models was assessed by the area under the receiver operating characteristic curve (AUC). The performance of the models was compared using the DeLong test, and clinical utility was evaluated by decision curve analysis.

RESULTS

The AUC of the radiomics model was 0.803 (95% confidence interval [CI]: 0.733-0.873) and 0.803 (95% CI: 0.698-0.908) in the training and test cohorts, respectively. The AUC of the combined model (containing data on age, diabetes, and atrial fibrillation) in the training and test sets was 0.813 (95% CI: 0.750-0.889) and 0.803 (95% CI: 0.699-0.907), respectively. The DeLong test showed no significant difference between the radiomics and combined models. Decision curve analysis showed that both models had clinical utility.

CONCLUSION

CTA-based thrombus radiomics can estimate TSS in patients with AIS. The addition of clinical data to the model does not improve predictive performance.

Proposed Protocols for Artificial Intelligence Imaging Database in Acute Stroke Imaging

PURPOSE

To propose standardized and feasible imaging protocols for constructing artificial intelligence (AI) database in acute stroke by assessing the current practice at tertiary hospitals in South Korea and reviewing evolving AI models.

MATERIALS AND METHODS

A nationwide survey on acute stroke imaging protocols was conducted using an electronic questionnaire sent to 43 registered tertiary hospitals between April and May 2021. Imaging protocols for endovascular thrombectomy (EVT) in the early and late time windows and during follow-up were assessed. Clinical applications of AI techniques in stroke imaging and required sequences for developing AI models were reviewed. Standardized and feasible imaging protocols for data curation in acute stroke were proposed.

RESULTS

There was considerable heterogeneity in the imaging protocols for EVT candidates in the early and late time windows and posterior circulation stroke. Computed tomography (CT)-based protocols were adopted by 70% (30/43), and acquisition of noncontrast CT, CT angiography and CT perfusion in a single session was most commonly performed (47%, 14/30) with the preference of multiphase (70%, 21/30) over single phase CT angiography. More hospitals performed magnetic resonance imaging (MRI)-based protocols or additional MRI sequences in a late time window and posterior circulation stroke. Diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) were most commonly performed MRI sequences with considerable variation in performing other MRI sequences. AI models for diagnostic purposes required noncontrast CT, CT angiography and DWI while FLAIR, dynamic susceptibility contrast perfusion, and T1-weighted imaging (T1WI) were additionally required for prognostic AI models.

CONCLUSION

Given considerable heterogeneity in acute stroke imaging protocols at tertiary hospitals in South Korea, standardized and feasible imaging protocols are required for constructing AI database in acute stroke. The essential sequences may be noncontrast CT, DWI, CT/MR angiography and CT/MR perfusion while FLAIR and T1WI may be additionally required.

The Value of Applying Machine Learning in Predicting the Time of Symptom Onset in Stroke Patients: Systematic Review and Meta-Analysis

BACKGROUND

Machine learning is a potentially effective method for identifying and predicting the time of the onset of stroke. However, the value of applying machine learning in this field remains controversial and debatable.

OBJECTIVE

We aimed to assess the value of applying machine learning in predicting the time of stroke onset.

METHODS

PubMed, Web of Science, Embase, and Cochrane were comprehensively searched. The C index and sensitivity with 95% CI were used as effect sizes. The risk of bias was evaluated using PROBAST (Prediction Model Risk of Bias Assessment Tool), and meta-analysis was conducted using R (version 4.2.0; R Core Team).

RESULTS

Thirteen eligible studies were included in the meta-analysis involving 55 machine learning models with 41 models in the training set and 14 in the validation set. The overall C index was 0.800 (95% CI 0.773-0.826) in the training set and 0.781 (95% CI 0.709-0.852) in the validation set. The sensitivity and specificity were 0.76 (95% CI 0.73-0.80) and 0.79 (95% CI 0.74-0.82) in the training set and 0.81 (95% CI 0.68-0.90) and 0.83 (95% CI 0.73-0.89) in the validation set, respectively. Subgroup analysis revealed that the accuracy of machine learning in predicting the time of stroke onset within 4.5 hours was optimal (training: 0.80, 95% CI 0.77-0.83; validation: 0.79, 95% CI 0.71-0.86).

CONCLUSIONS

Machine learning has ideal performance in identifying the time of stroke onset. More reasonable image segmentation and texture extraction methods in radiomics should be used to promote the value of applying machine learning in diverse ethnic backgrounds.

TRIAL REGISTRATION

PROSPERO CRD42022358898; https://www.crd.york.ac.uk/Prospero/display_record.php?RecordID=358898.

Integrating Artificial Intelligence and Machine Learning Into Cancer Clinical Trials

The practice of oncology requires analyzing and synthesizing abundant data. From the patient's workup to determine eligibility to the therapies received to the post-treatment surveillance, practitioners must constantly juggle, evaluate, and weigh decision-making based on their best understanding of information at hand. These complex, multifactorial decisions have a tremendous opportunity to benefit from data-driven machine learning (ML) methods to drive opportunities in artificial intelligence (AI). Within the past 5 years, we have seen AI move from simply a promising opportunity to being used in prospective trials. Here, we review recent efforts of AI in clinical trials that have moved the needle towards improved prediction of actionable outcomes, such as predicting acute care visits, short term mortality, and pathologic extranodal extension. We then pause and reflect on how these AI models ask a different question than traditional statistics models that readers may be more familiar with; how then should readers conceptualize and interpret AI models that they are not as familiar with. We end with what we believe are promising future opportunities for AI in oncology, with an eye towards allowing the data to inform us through unsupervised learning and generative models, rather than asking AI to perform specific functions.

MSA-YOLOv5: Multi-scale attention-based YOLOv5 for automatic detection of acute ischemic stroke from multi-modality MRI images

BACKGROUND AND OBJECTIVE

Acute ischemic stroke (AIS) is a common neurological disorder characterized by the sudden onset of cerebral ischemia, leading to functional impairments. Swift and precise detection of AIS lesions is crucial for stroke diagnosis and treatment but poses a significant challenge. This study aims to leverage multimodal fusion technology to combine complementary information from various modalities, thereby enhancing the detection performance of AIS target detection models.

METHODS

In this retrospective study of AIS, we collected data from 316 AIS patients and created a multi-modality magnetic resonance imaging (MRI) dataset. We propose a Multi-Scale Attention-based YOLOv5 (MSA-YOLOv5), targeting challenges such as small lesion size and blurred borders at low resolutions. Specifically, we augment YOLOv5 with a prediction head to detect objects at various scales. Next, we replace the original prediction head with a Multi-Scale Swin Transformer Prediction Head (MS-STPH), which reduces computational complexity to linear levels and enhances the ability to detect small lesions. We incorporate a Second-Order channel attention (SOCA) module to adaptively rescale channel features by employing second-order feature statistics for more discriminative representations. Finally, we further validate the effectiveness of our method using the ISLES 2022 dataset.

RESULTS

On our in-house AIS dataset, MSA-YOLOv5 achieves a 79.0% mAP0.5, substantially surpassing other single-stage models. Compared to two-stage models, it maintains a comparable performance level while significantly reducing the number of parameters and resolution. On the ISLES 2022 dataset, MSA-YOLOv5 attains an 80.0% mAP0.5, outperforming other network models by a considerable margin. MS-STPH and SOCA modules can significantly increase mAP0.5 by 2.7% and 1.9%, respectively. Visualization interpretability results show that the proposed MSA-YOLOv5 restricts high attention in the small regions of AIS lesions.

CONCLUSIONS

The proposed MSA-YOLOv5 is capable of automatically and effectively detecting acute ischemic stroke lesions in multimodal images, particularly for small lesions and artifacts. Our enhanced model reduces the number of parameters while improving detection accuracy. This model can potentially assist radiologists in providing more accurate diagnosis, and enable clinicians to develop better treatment plans.

Artificial Intelligence for Automated DWI/FLAIR Mismatch Assessment on Magnetic Resonance Imaging in Stroke: A Systematic Review

We conducted this Systematic Review to create an overview of the currently existing Artificial Intelligence (AI) methods for Magnetic Resonance Diffusion-Weighted Imaging (DWI)/Fluid-Attenuated Inversion Recovery (FLAIR)-mismatch assessment and to determine how well DWI/FLAIR mismatch algorithms perform compared to domain experts. We searched PubMed Medline, Ovid Embase, Scopus, Web of Science, Cochrane, and IEEE Xplore literature databases for relevant studies published between 1 January 2017 and 20 November 2022, following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. We assessed the included studies using the Quality Assessment of Diagnostic Accuracy Studies 2 tool. Five studies fit the scope of this review. The area under the curve ranged from 0.74 to 0.90. The sensitivity and specificity ranged from 0.70 to 0.85 and 0.74 to 0.84, respectively. Negative predictive value, positive predictive value, and accuracy ranged from 0.55 to 0.82, 0.74 to 0.91, and 0.73 to 0.83, respectively. In a binary classification of ±4.5 h from stroke onset, the surveyed AI methods performed equivalent to or even better than domain experts. However, using the relation between time since stroke onset (TSS) and increasing visibility of FLAIR hyperintensity lesions is not recommended for the determination of TSS within the first 4.5 h. An AI algorithm on DWI/FLAIR mismatch assessment focused on treatment eligibility, outcome prediction, and consideration of patient-specific data could potentially increase the proportion of stroke patients with unknown onset who could be treated with thrombolysis.

Prediction of subjective cognitive decline after corpus callosum infarction by an interpretable machine learning-derived early warning strategy

Background and purpose

Corpus callosum (CC) infarction is an extremely rare subtype of cerebral ischemic stroke, however, the symptoms of cognitive impairment often fail to attract early attention of patients, which seriously affects the long-term prognosis, such as high mortality, personality changes, mood disorders, psychotic reactions, financial burden and so on. This study seeks to develop and validate models for early predicting the risk of subjective cognitive decline (SCD) after CC infarction by machine learning (ML) algorithms.

Methods

This is a prospective study that enrolled 213 (only 3.7%) CC infarction patients from a nine-year cohort comprising 8,555 patients with acute ischemic stroke. Telephone follow-up surveys were carried out for the patients with definite diagnosis of CC infarction one-year after disease onset, and SCD was identified by Behavioral Risk Factor Surveillance System (BRFSS) questionnaire. Based on the significant features selected by the least absolute shrinkage and selection operator (LASSO), seven ML models including Extreme Gradient Boosting (XGBoost), Logistic Regression (LR), Light Gradient Boosting Machine (LightGBM), Adaptive Boosting (AdaBoost), Gaussian Naïve Bayes (GNB), Complement Naïve Bayes (CNB), and Support vector machine (SVM) were established and their predictive performances were compared by different metrics. Importantly, the SHapley Additive exPlanations (SHAP) was also utilized to examine internal behavior of the highest-performance ML classifier.

Results

The Logistic Regression (LR)-model performed better than other six ML-models in SCD predictability after the CC infarction, with the area under the receiver characteristic operator curve (AUC) of 77.1% in the validation set. Using LASSO and SHAP analysis, we found that infarction subregions of CC infarction, female, 3-month modified Rankin Scale (mRS) score, age, homocysteine, location of angiostenosis, neutrophil to lymphocyte ratio, pure CC infarction, and number of angiostenosis were the top-nine significant predictors in the order of importance for the output of LR-model. Meanwhile, we identified that infarction subregion of CC, female, 3-month mRS score and pure CC infarction were the factors which independently associated with the cognitive outcome.

Conclusion

Our study firstly demonstrated that the LR-model with 9 common variables has the best-performance to predict the risk of post-stroke SCD due to CC infarcton. Particularly, the combination of LR-model and SHAP-explainer could aid in achieving personalized risk prediction and be served as a decision-making tool for early intervention since its poor long-term outcome.

Artificial Intelligence for Clinical Decision Support in Acute Ischemic Stroke: A Systematic Review

BACKGROUND

Established randomized trial-based parameters for acute ischemic stroke group patients into generic treatment groups, leading to attempts using various artificial intelligence (AI) methods to directly correlate patient characteristics to outcomes and thereby provide decision support to stroke clinicians. We review AI-based clinical decision support systems in the development stage, specifically regarding methodological robustness and constraints for clinical implementation.

METHODS

Our systematic review included full-text English language publications proposing a clinical decision support system using AI techniques for direct decision support in acute ischemic stroke cases in adult patients. We (1) describe data and outcomes used in those systems, (2) estimate the systems' benefits compared with traditional stroke diagnosis and treatment, and (3) reported concordance with reporting standards for AI in healthcare.

RESULTS

One hundred twenty-one studies met our inclusion criteria. Sixty-five were included for full extraction. In our sample, utilized data sources, methods, and reporting practices were highly heterogeneous.

CONCLUSIONS

Our results suggest significant validity threats, dissonance in reporting practices, and challenges to clinical translation. We outline practical recommendations for the successful implementation of AI research in acute ischemic stroke treatment and diagnosis.

Big Data in Stroke: How to Use Big Data to Make the Next Management Decision

The last decade has seen significant advances in the accumulation of medical data, the computational techniques to analyze that data, and corresponding improvements in management. Interventions such as thrombolytics and mechanical thrombectomy improve patient outcomes after stroke in selected patients; however, significant gaps remain in our ability to select patients, predict complications, and understand outcomes. Big data and the computational methods needed to analyze it can address these gaps. For example, automated analysis of neuroimaging to estimate the volume of brain tissue that is ischemic and salvageable can help triage patients for acute interventions. Data-intensive computational techniques can perform complex risk calculations that are too cumbersome to be completed by humans, resulting in more accurate and timely prediction of which patients require increased vigilance for adverse events such as treatment complications. To handle the accumulation of complex medical data, a variety of advanced computational techniques referred to as machine learning and artificial intelligence now routinely complement traditional statistical inference. In this narrative review, we explore data-intensive techniques in stroke research, how it has informed the management of stroke patients, and how current work could shape clinical practice in the future.

A clinical-radiomics model based on noncontrast computed tomography to predict hemorrhagic transformation after stroke by machine learning: a multicenter study

OBJECTIVE

To build a clinical-radiomics model based on noncontrast computed tomography images to identify the risk of hemorrhagic transformation (HT) in patients with acute ischemic stroke (AIS) following intravenous thrombolysis (IVT).

MATERIALS AND METHODS

A total of 517 consecutive patients with AIS were screened for inclusion. Datasets from six hospitals were randomly divided into a training cohort and an internal cohort with an 8:2 ratio. The dataset of the seventh hospital was used for an independent external verification. The best dimensionality reduction method to choose features and the best machine learning (ML) algorithm to develop a model were selected. Then, the clinical, radiomics and clinical-radiomics models were developed. Finally, the performance of the models was measured using the area under the receiver operating characteristic curve (AUC).

RESULTS

Of 517 from seven hospitals, 249 (48%) had HT. The best method for choosing features was recursive feature elimination, and the best ML algorithm to build models was extreme gradient boosting. In distinguishing patients with HT, the AUC of the clinical model was 0.898 (95% CI 0.873-0.921) in the internal validation cohort, and 0.911 (95% CI 0.891-0.928) in the external validation cohort; the AUC of radiomics model was 0.922 (95% CI 0.896-0.941) and 0.883 (95% CI 0.851-0.902), while the AUC of clinical-radiomics model was 0.950 (95% CI 0.925-0.967) and 0.942 (95% CI 0.927-0.958) respectively.

CONCLUSION

The proposed clinical-radiomics model is a dependable approach that could provide risk assessment of HT for patients who receive IVT after stroke.

Common clinical blood and urine biomarkers for ischemic stroke: an Estonian Electronic Health Records database study

BACKGROUND

Ischemic stroke (IS) is a major health risk without generally usable effective measures of primary prevention. Early warning signals that are easy to detect and widely available can save lives. Estonia has one nation-wide Electronic Health Record (EHR) database for the storage of medical information of patients from hospitals and primary care providers.

METHODS

We extracted structured and unstructured data from the EHRs of participants of the Estonian Biobank (EstBB) and evaluated different formats of input data to understand how this continuously growing dataset should be prepared for best prediction. The utility of the EHR database for finding blood- and urine-based biomarkers for IS was demonstrated by applying different analytical and machine learning (ML) methods.

RESULTS

Several early trends in common clinical laboratory parameter changes (set of red blood indices, lymphocyte/neutrophil ratio, etc.) were established for IS prediction. The developed ML models predicted the future occurrence of IS with very high accuracy and Random Forests was proved as the most applicable method to EHR data.

CONCLUSIONS

We conclude that the EHR database and the risk factors uncovered are valuable resources in screening the population for risk of IS as well as constructing disease risk scores and refining prediction models for IS by ML.

CDC42 Might Be a Molecular Signature of DWI-FLAIR Mismatch in a Nonhuman Primate Stroke Model

No definitive blood markers of DWI-FLAIR mismatch, a pivotal indicator of salvageable ischemic penumbra brain tissue, are known. We previously reported that CDC42 and RHOA are associated with the ischemic penumbra. Here, we investigated whether plasma CDC42 and RHOA are surrogate markers of DWI-FLAIR mismatch. Sixteen cynomolgus macaques (3 as controls and 13 for the stroke model) were included. Guided by digital subtraction angiography (DSA), a middle cerebral artery occlusion (MCAO) model was established by occluding the middle cerebral artery (MCA) with a balloon. MRI and neurological deficit scoring were performed to evaluate postinfarction changes. Plasma CDC42 and RHOA levels were measured by enzyme-linked immunosorbent assay (ELISA). The stroke model was successfully established in eight monkeys. Based on postinfarction MRI images, experimental animals were divided into a FLAIR (-) group (N = 4) and a FLAIR (+) group (N = 4). Plasma CDC42 in the FLAIR (-) group showed a significant decrease compared with that in the FLAIR (+) group (p < 0.05). No statistically significant difference was observed for plasma RHOA. The FLAIR (-) group showed a milder neurological function deficit and a smaller infarct volume than the FLAIR (+) group (p < 0.05). Therefore, plasma CDC42 might be a new surrogate marker for DWI-FLAIR mismatch.

Machine learning and acute stroke imaging

BACKGROUND

In recent years, machine learning (ML) has had notable success in providing automated analyses of neuroimaging studies, and its role is likely to increase in the future. Thus, it is paramount for clinicians to understand these approaches, gain facility with interpreting ML results, and learn how to assess algorithm performance.

OBJECTIVE

To provide an overview of ML, present its role in acute stroke imaging, discuss methods to evaluate algorithms, and then provide an assessment of existing approaches.

METHODS

In this review, we give an overview of ML techniques commonly used in medical imaging analysis and methods to evaluate performance. We then review the literature for relevant publications. Searches were run in November 2021 in Ovid Medline and PubMed. Inclusion criteria included studies in English reporting use of artificial intelligence (AI), machine learning, or similar techniques in the setting of, and in applications for, acute ischemic stroke or mechanical thrombectomy. Articles that included image-level data with meaningful results and sound ML approaches were included in this discussion.

RESULTS

Many publications on acute stroke imaging, including detection of large vessel occlusion, detection and quantification of intracranial hemorrhage and detection of infarct core, have been published using ML methods. Imaging inputs have included non-contrast head CT, CT angiograph and MRI, with a range of performances. We discuss and review several of the most relevant publications.

CONCLUSIONS

ML in acute ischemic stroke imaging has already made tremendous headway. Additional applications and further integration with clinical care is inevitable. Thus, facility with these approaches is critical for the neurointerventional clinician.

Models of ultrasonic radiomics and clinical characters for lymph node metastasis assessment in thyroid cancer: a retrospective study

Background

Preoperative prediction of cervical lymph node metastasis in papillary thyroid carcinoma provided a basis for tumor staging and treatment decision. This study aimed to investigate the utility of machine learning and develop different models to preoperatively predict cervical lymph node metastasis based on ultrasonic radiomic features and clinical characteristics in papillary thyroid carcinoma nodules.

Methods

Data from 400 papillary thyroid carcinoma nodules were included and divided into training and validation group. With the help of machine learning, clinical characteristics and ultrasonic radiomic features were extracted and selected using randomforest and least absolute shrinkage and selection operator regression before classified by five classifiers. Finally, 10 models were built and their area under the receiver operating characteristic curve, accuracy, sensitivity, specificity, positive predictive value and negative predictive value were measured.

Results

Among the 10 models, RF-RF model revealed the highest area under curve (0.812) and accuracy (0.7542) in validation group. The top 10 variables of it included age, seven textural features, one shape feature and one first-order feature, in which eight were high-dimensional features.

Conclusions

RF-RF model showed the best predictive performance for cervical lymph node metastasis. And the importance features selected by it highlighted the unique role of higher-dimensional statistical methods for radiomics analysis.

Predicting Functional Outcome Using 24-Hour Post-Treatment Characteristics: Application of Machine Learning Algorithms in the STRATIS Registry

FOR SOCIAL MEDIA: @AliciaCastongu2, @FazalZaidi9, @oozaidat, @Mouhammad_Jumaa OBJECTIVE: Machine learning (ML) algorithms have emerged as powerful predictive tools in the field stroke. Here, we examine the predictive accuracy of ML models for predicting functional outcomes using 24-hour post-treatment characteristics in the Systematic Evaluation of Patients Treated With Neurothrombectomy Devices for Acute Ischemic Stroke (STRATIS) Registry.

METHODS

ML models, adaptive boost, random forest (RF), classification and regression trees (CART), C5.0 decision tree (C5.0), support vector machine (SVM), least absolute shrinkage and selection operator (LASSO), and logistic regression (LR), and traditional LR models were used to predict 90-day functional outcome (modified Rankin Scale score 0-2). Twenty-four-hour National Institutes of Health Stroke Scale (NIHSS) was examined as a continuous or dichotomous variable in all models. Model accuracy was assessed using the area under characteristic curve (AUC).

RESULTS

The 24-hour NIHSS score was a top-predictor of functional outcome in all models. ML models using the continuous 24-hour NIHSS scored showed moderate-to-good predictive performance (range mean AUC: 0.76-0.92); however, RF (AUC: 0.92 ± 0.028) outperformed all ML models, except LASSO (AUC: 0.89 ± 0.023, p = 0.0958). Importantly, RF demonstrated a significantly higher predictive value than LR (AUC: 0.87 ± 0.031, p = 0.048) and traditional LR (AUC: 85 ± 0.06, p = 0.035) when using the 24-hour continuous NIHSS score. Predictive accuracy was similar between the 24-hour NIHSS score dichotomous and continuous ML models.

INTERPRETATION

In this substudy, we found similar predictive accuracy for functional outcome when using the 24-hour NIHSS score as a continuous or dichotomous variable in ML models. ML models had moderate-to-good predictive accuracy, with RF outperforming LR models. External validation of these ML models is warranted. ANN NEUROL 2023;93:40-49.

Comparison of semi-quantitative and visual assessment of early MRI signal evolution in acute ischaemic stroke

Background

The evaluation of DWI/FLAIR mismatch in ischaemic stroke patients with unknown, time from onset can determine the treatment strategy. This approach is based on, visual assessment and may be subject to insufficient inter-rater agreement.

Objective

To compare the inter-rater agreement of visual evaluation of FLAIR MRI and proposed region of interest (ROI) semiquantitative method in large vessel occlusion (LVO) strokes.

Methods

Five readers have analysed MRIs of 104 patients obtained within six hours of the onset of stroke symptoms resulting from LVO visually and semi-quantitatively. For the semiquantitative analysis, a ROI method was used to obtain relative signal intensity compared to the unaffected side. Cut-off values of 1.15 and 1.10 were tested. The analysis yielded FLAIR-positive (abnormal) and negative (normal) findings. Percentage agreement and Fleiss kappa coefficients were calculated.

Results

The visual agreement of 5/5 readers and ≥ 4/5 readers occurred in 31% and 59% of cases respectively. Semi-quantitative evaluation using a cut-off value of 1.15 increased the agreements to 67% and 88% respectively. The agreement of visual evaluation was fair. The semi-quantitative method utilising the cut-off of 1.15 had moderate agreement although it increased the number of FLAIR-negative results compared to the visual evaluation. A low cut-off value of 1.10 didn't improve the agreement significantly.

Conclusion

The inter-rater agreement of visual evaluation of FLAIR in patients with short-duration large vessel occlusion stroke was fair. The high cut-off value of semiquantitative evaluation increased the agreement although it changed the proportion of FLAIR positive and negative results.

Explainable machine learning for long-term outcome prediction in two-center stroke patients after intravenous thrombolysis

Objective

Neurological outcome prediction in patients with ischemic stroke is very critical in treatment strategy and post-stroke management. Machine learning techniques with high accuracy are increasingly being developed in the medical field. We studied the application of machine learning models to predict long-term neurological outcomes in patients with after intravenous thrombolysis.

Methods

A retrospective cohort study was performed to review all stroke patients with intravenous thrombolysis. Patients with modified Rankin Score (mRs) less than two at three months post-thrombolysis were considered as good outcome. The clinical features between stroke patients with good and with poor outcomes were compared using three different machine learning models (Random Forest, Support Vector Machine and Logistic Regression) to identify which performed best. Two datasets from the other stroke center were included accordingly for external verification and performed with explainable AI models.

Results

Of the 488 patients enrolled in this study, and 374 (76.6%) patients had favorable outcomes. Patients with higher mRs at 3 months had increased systolic pressure, blood glucose, cholesterol (TC), and 7-day National Institute of Health Stroke Scale (NIHSS) score compared to those with lower mRs. The predictability and the areas under the curves (AUC) for the random forest model was relatively higher than support vector machine and LR models. These findings were further validated in the external dataset and similar results were obtained. The explainable AI model identified the risk factors as well.

Conclusion

Explainable AI model is able to identify NIHSS_Day7 is independently efficient in predicting neurological outcomes in patients with ischemic stroke after intravenous thrombolysis.

Using Artificial Intelligence to Develop a Multivariate Model with a Machine Learning Model to Predict Complications in Mexican Diabetic Patients without Arterial Hypertension (National Nested Case-Control Study): Metformin and Elevated Normal Blood Pressure Are Risk Factors, and Obesity Is Protective

Diabetes mellitus is a disease with no cure that can cause complications and even death. Moreover, over time, it will lead to chronic complications. Predictive models have been used to identify people with a tendency to develop diabetes mellitus. At the same time, there is limited information regarding the chronic complications of patients with diabetes. Our study is aimed at creating a machine-learning model that will be able to identify the risk factors of a diabetic patient developing chronic complications such as amputations, myocardial infarction, stroke, nephropathy, and retinopathy. The design is a national nested case-control study with 63,776 patients and 215 predictors with four years of data. Using an XGBoost model, the prediction of chronic complications has an AUC of 84%, and the model has identified the risk factors for chronic complications in patients with diabetes. According to the analysis, the most crucial risk factors based on SHAP values (Shapley additive explanations) are continued management, metformin treatment, age between 68 and 104 years, nutrition consultation, and treatment adherence. But we highlight two exciting findings. The first is a reaffirmation that high blood pressure figures across patients with diabetes without hypertension become a significant risk factor at diastolic > 70 mmHg (OR: 1.095, 95% CI: 1.078-1.113) or systolic > 120 mmHg (OR: 1.147, 95% CI: 1.124-1.171). Furthermore, people with diabetes with a BMI > 32 (overall obesity) (OR: 0.816, 95% CI: 0.8-0.833) have a statistically significant protective factor, which the paradox of obesity may explain. In conclusion, the results we have obtained show that artificial intelligence is a powerful and feasible tool to use for this type of study. However, we suggest that more studies be conducted to verify and elaborate upon our findings.

Identifying patients with acute ischemic stroke within a 6-h window for the treatment of endovascular thrombectomy using deep learning and perfusion imaging

Introduction

It is critical to identify the stroke onset time of patients with acute ischemic stroke (AIS) for the treatment of endovascular thrombectomy (EVT). However, it is challenging to accurately ascertain this time for patients with wake-up stroke (WUS). The current study aimed to construct a deep learning approach based on computed tomography perfusion (CTP) or perfusion weighted imaging (PWI) to identify a 6-h window for patients with AIS for the treatment of EVT.

Methods

We collected data from 377 patients with AIS, who were examined by CTP or PWI before making a treatment decision. Cerebral blood flow (CBF), time to maximum peak (Tmax), and a region of interest (ROI) mask were preprocessed from the CTP and PWI. We constructed the classifier based on a convolutional neural network (CNN), which was trained by CBF, Tmax, and ROI masks to identify patients with AIS within a 6-h window for the treatment of EVT. We compared the classification performance among a CNN, support vector machine (SVM), and random forest (RF) when trained by five different types of ROI masks. To assess the adaptability of the classifier of CNN for CTP and PWI, which were processed respectively from CTP and PWI groups.

Results

Our results showed that the CNN classifier had a higher performance with an area under the curve (AUC) of 0.935, which was significantly higher than that of support vector machine (SVM) and random forest (RF) (p = 0.001 and p = 0.001, respectively). For the CNN classifier trained by different ROI masks, the best performance was trained by CBF, Tmax, and ROI masks of Tmax > 6 s. No significant difference was detected in the classification performance of the CNN between CTP and PWI (0.902 vs. 0.928; p = 0.557).

Discussion

The CNN classifier trained by CBF, Tmax, and ROI masks of Tmax > 6 s had good performance in identifying patients with AIS within a 6-h window for the treatment of EVT. The current study indicates that the CNN model has potential to be used to accurately estimate the stroke onset time of patients with WUS.

Comparison of ischemic stroke diagnosis models based on machine learning

Background

The incidence, prevalence, and mortality of ischemic stroke (IS) continue to rise, resulting in a serious global disease burden. The prediction models have a great value in the early prediction and diagnosis of IS.

Methods

The R software was used to screen the differentially expressed genes (DEGs) of IS and control samples in the datasets GSE16561, GSE58294, and GSE37587 and analyze DEGs for enrichment analysis. The feature genes of IS were obtained by several machine learning algorithms, including the least absolute shrinkage and selector operation (LASSO) logistic regression, the support vector machine-recursive feature elimination (SVM-RFE), and the Random Forest (RF). The IS diagnostic models were constructed based on transcriptomics by machine learning and artificial neural network (ANN).

Results

A total of 69 DEGs, mainly involved in immune and inflammatory responses, were identified. The pathways enriched in the IS group were complement and coagulation cascades, lysosome, PPAR signaling pathway, regulation of autophagy, and toll-like receptor signaling pathway. The feature genes selected by LASSO, SVM-RFE, and RF were 17, 10, and 12, respectively. The area under the curve (AUC) of the LASSO model in the training dataset, GSE22255, and GSE195442 was 0.969, 0.890, and 1.000. The AUC of the SVM-RFE model was 0.957, 0.805, and 1.000, respectively. The AUC of the RF model was 0.947, 0.935, and 1.000, respectively. The models have good sensitivity, specificity, and accuracy. The AUC of the LASSO+ANN, SVM-RFE+ANN, and RF+ANN models was 1.000, 0.995, and 0.997, respectively, in the training dataset. However, the AUC of LASSO+ANN, SVM-RFE+ANN, and RF+ANN models was 0.688, 0.605, and 0.619, respectively, in the GSE22255 dataset. The AUC of the LASSO+ANN and RF+ANN models was 0.740 and 0.630, respectively, in the GSE195442 dataset. In the training dataset, the sensitivity, specificity, and accuracy of the LASSO+ANN model were 1.000, 1.000, and 1.000, respectively; of the SVM-RFE+ANN model were 0.946, 0.982, and 0.964, respectively; and of the RF+ANN model were 0.964, 1.000, and 0.982, respectively. In the test datasets, the sensitivity was very satisfactory; however, the specificity and accuracy were not good.

Conclusion

The LASSO, SVM-RFE, and RF models have good prediction abilities. However, the ANN model is efficient at classifying positive samples and is unsuitable at classifying negative samples.

Deep-Learning-Based Stroke Screening Using Skeleton Data from Neurological Examination Videos

According to the Korea Institute for Health and Social Affairs, in 2017, the elderly, aged 65 or older, had an average of 2.7 chronic diseases per person. The concern for the medical welfare of the elderly is increasing due to a low birth rate, an aging population, and the lack of medical personnel. The demand for services that take user age, cognitive capacity, and difficulty into account is rising. As a result, there is an increased demand for smart healthcare systems that can lower hospital admissions and offer patients individualized care. This has motivated us to develop an AI system that can easily screen and manage neurological diseases through videos. As neurological diseases can be diagnosed by visual analysis to some extent, in this study, we set out to estimate the possibility of a person having a neurological disease from videos. Among neurological diseases, we focus on stroke because it is a common condition in the elderly population and results in high mortality and morbidity worldwide. The proposed method consists of three steps: (1) transforming neurological examination videos into landmark data, (2) converting the landmark data into recurrence plots, and (3) estimating the possibility of a stroke using deep neural networks. Major features, such as the hand, face, pupil, and body movements of a person are extracted from test videos taken under several neurological examination protocols using deep-learning-based landmark extractors. Sequences of these landmark data are then converted into recurrence plots, which can be interpreted as images. These images can be fed into convolutional neural networks to classify stroke using feature-fusion techniques. A case study of the application of a disease screening test to assess the capability of the proposed method is presented.

Neuroimaging in Pediatric Stroke

Pediatric stroke is unfortunately not a rare condition. It is associated with severe disability and mortality because of the complexity of potential clinical manifestations, and the resulting delay in seeking care and in diagnosis. Neuroimaging plays an important role in the multidisciplinary response for pediatric stroke patients. The rapid development of adult endovascular thrombectomy has created a new momentum in health professionals caring for pediatric stroke patients. Neuroimaging is critical to make decisions of identifying appropriate candidates for thrombectomy. This review article will review current neuroimaging techniques, imaging work-up strategies and special considerations in pediatric stroke. For resources limited areas, recommendation of substitute imaging approaches will be provided. Finally, promising new techniques and hypothesis-driven research protocols will be discussed.

Identifying acute ischemic stroke patients within the thrombolytic treatment window using deep learning

BACKGROUND AND PURPOSE

Treatment of acute ischemic stroke is heavily contingent upon time, as there is a strong relationship between time clock and tissue progression. Work has established imaging biomarker assessments as surrogates for time since stroke (TSS), namely, by comparing signal mismatch between diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) imaging. Our goal was to develop an automatic technique for determining TSS from imaging that does not require subspecialist radiology expertise.

METHODS

Using 772 patients (66 ± 9 years, 319 women), we developed and externally evaluated a deep learning network for classifying TSS from MR images and compared algorithm predictions to neuroradiologist assessments of DWI-FLAIR mismatch. Models were trained to classify TSS within 4.5 hours and performance metrics with confidence intervals were reported on both internal and external evaluation sets.

RESULTS

Three board-certified neuroradiologists' DWI-FLAIR mismatch assessments, based on majority vote, yielded a sensitivity of .62, a specificity of .86, and a Fleiss' kappa of .46 when used to classify TSS. The deep learning method performed similarly to radiologists and outperformed previously reported methods, with the best model achieving an average evaluation accuracy, sensitivity, and specificity of .726, .712, and .741, respectively, on an internal cohort and .724, .757, and .679, respectively, on an external cohort.

CONCLUSION

Our model achieved higher generalization performance on external evaluation datasets than the current state-of-the-art for TSS classification. These results demonstrate the potential of automatic assessment of onset time from imaging without the need for expertly trained radiologists.

Improving interobserver agreement and performance of deep learning models for segmenting acute ischemic stroke by combining DWI with optimized ADC thresholds

OBJECTIVES

To examine the role of ADC threshold on agreement across observers and deep learning models (DLMs) plus segmentation performance of DLMs for acute ischemic stroke (AIS).

METHODS

Twelve DLMs, which were trained on DWI-ADC-ADC combination from 76 patients with AIS using 6 different ADC thresholds with ground truth manually contoured by 2 observers, were tested by additional 67 patients in the same hospital and another 78 patients in another hospital. Agreement between observers and DLMs were evaluated by Bland-Altman plot and intraclass correlation coefficient (ICC). The similarity between ground truth (GT) defined by observers and between automatic segmentation performed by DLMs was evaluated by Dice similarity coefficient (DSC). Group comparison was performed using the Mann-Whitney U test. The relationship between the DSC and ADC threshold as well as AIS lesion size was evaluated by linear regression analysis. A p < .05 was considered statistically significant.

RESULTS

Excellent interobserver agreement and intraobserver repeatability in the manual segmentation (all ICC > 0.98, p < .001) were achieved. The 95% limit of agreement was reduced from 11.23 cm2 for GT on DWI to 0.59 cm2 for prediction at an ADC threshold of 0.6 × 10-3 mm2/s combined with DWI. The segmentation performance of DLMs was improved with an overall DSC from 0.738 ± 0.214 on DWI to 0.971 ± 0.021 on an ADC threshold of 0.6 × 10-3 mm2/s combined with DWI.

CONCLUSIONS

Combining an ADC threshold of 0.6 × 10-3 mm2/s with DWI reduces interobserver and inter-DLM difference and achieves best segmentation performance of AIS lesions using DLMs.

KEY POINTS

• Higher Dice similarity coefficient (DSC) in predicting acute ischemic stroke lesions was achieved by ADC thresholds combined with DWI than by DWI alone (all p < .05). • DSC had a negative association with the ADC threshold in most sizes, both hospitals, and both observers (most p < .05) and a positive association with the stroke size in all ADC thresholds, both hospitals, and both observers (all p < .001). • An ADC threshold of 0.6 × 10-3 mm2/s eliminated the difference of DSC at any stroke size between observers or between hospitals (p = .07 to > .99).

The integration of radiology curriculum in undergraduate medical education

Knowledge of basic radiology is an essential component of the undergraduate medical curriculum. Pre-clinical education introduces medical students to essential knowledge and skills. However, the current curriculum and radiology teaching are not without inherent limitations. This article explores the essential role of radiology education for medical students and discusses the current state of affairs. It also highlights the limitations and associated challenges and proposes solutions.

•

Fulfilling the goal of integrating radiology into undergraduate medical curriculam is a real challenge due to the enduring faith assuming that traditional medical disciplines are worthy of consuming the available study time.

•

Radiology is an essential division of medical science with a wide range of clinical applications in the diagnosis, prognosis, and treatment efficacy.

•

Poor integration of the radiology curriculum in undergraduate medical education can lead to poor image interpretation skills and poor use of imaging algorithms in medical students and post-graduate doctors.

Stroke Risk Prediction with Machine Learning Techniques

A stroke is caused when blood flow to a part of the brain is stopped abruptly. Without the blood supply, the brain cells gradually die, and disability occurs depending on the area of the brain affected. Early recognition of symptoms can significantly carry valuable information for the prediction of stroke and promoting a healthy life. In this research work, with the aid of machine learning (ML), several models are developed and evaluated to design a robust framework for the long-term risk prediction of stroke occurrence. The main contribution of this study is a stacking method that achieves a high performance that is validated by various metrics, such as AUC, precision, recall, F-measure and accuracy. The experiment results showed that the stacking classification outperforms the other methods, with an AUC of 98.9%, F-measure, precision and recall of 97.4% and an accuracy of 98%.

Prediction of Hidden Coronary Artery Disease Using Machine Learning in Patients With Acute Ischemic Stroke

BACKGROUND AND OBJECTIVES

A machine learning technique for identifying hidden coronary artery disease (CAD) might be useful. We developed and validated machine learning models to predict patients with hidden CAD and assess long-term outcomes in patients with acute ischemic stroke.

METHODS

Multidetector coronary computed tomography was performed for patients without known history of CAD. Primary outcomes were defined as having any degree of CAD and having obstructive CAD (≥50% stenosis). Demographic variables, risk factors, laboratory results, Trial of ORG 10172 in Acute Stroke Treatment (TOAST) classification, NIH Stroke Scale score, blood pressure, and carotid artery stenosis were used to develop and validate machine learning models to predict CAD. Area under the receiver operating characteristic curves (AUC) was calculated for performance analysis, and Kaplan-Meier and Cox survival analyses of long-term outcomes were performed. Major adverse cardiovascular events (MACE) were defined as ischemic stroke, myocardial infarction, unstable angina, urgent coronary revascularization, and cardiovascular mortality.

RESULTS

Overall, 1,710 patients were included for the training dataset and 348 patients for the validation dataset. An Extreme Gradient Boosting model was developed to predict any degree of CAD, which showed an AUC of 0.763 (95% CI 0.711-0.814) on validation. A logistic regression model was used to predict obstructive CAD and had an AUC of 0.714 (95% CI 0.692-0.799). During the first 5 years of follow-up, MACE occurred more frequently when predicted of any CAD (P = 0.022) or obstructive CAD (P < 0.001). Cox proportional analysis showed that the hazard ratio of MACE was 1.5 (95% CI 1.1-2.2; P = 0.016) when predicted of any CAD, whereas it was 1.9 (95% CI 1.3-2.6; P < 0.001) for obstructive CAD.

DISCUSSION

We demonstrated that machine learning may help identify hidden CAD in patients with acute ischemic stroke. Long-term outcomes were also associated with prediction results.

CLASSIFICATION OF EVIDENCE

This study provides Class II evidence that in patients with acute ischemic stroke with CAD risk factors but no known history of CAD, a machine learning model predicts CAD on multidetector coronary computed tomography with an AUC of 0.763 (95% CI 0.711-0.814).

Deep Transfer Learning for Automatic Prediction of Hemorrhagic Stroke on CT Images

Intracerebral hemorrhage (ICH) is the most common type of hemorrhagic stroke which occurs due to ruptures of weakened blood vessel in brain tissue. It is a serious medical emergency issues that needs immediate treatment. Large numbers of noncontrast-computed tomography (NCCT) brain images are analyzed manually by radiologists to diagnose the hemorrhagic stroke, which is a difficult and time-consuming process. In this study, we propose an automated transfer deep learning method that combines ResNet-50 and dense layer for accurate prediction of intracranial hemorrhage on NCCT brain images. A total of 1164 NCCT brain images were collected from 62 patients with hemorrhagic stroke from Kalinga Institute of Medical Science, Bhubaneswar and used for evaluating the model. The proposed model takes individual CT images as input and classifies them as hemorrhagic or normal. This deep transfer learning approach reached 99.6% accuracy, 99.7% specificity, and 99.4% sensitivity which are better results than that of ResNet-50 only. It is evident that the deep transfer learning model has advantages for automatic diagnosis of hemorrhagic stroke and has the potential to be used as a clinical decision support tool to assist radiologists in stroke diagnosis.

Causative Classification of Ischemic Stroke by the Machine Learning Algorithm Random Forests

Background

Prognosis, recurrence rate, and secondary prevention strategies differ by different etiologies in acute ischemic stroke. However, identifying its cause is challenging.

Objective

This study aimed to develop a model to identify the cause of stroke using machine learning (ML) methods and test its accuracy.

Methods

We retrospectively reviewed the data of patients who had determined etiology defined by the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) from CASE-II (NCT04487340) to train and evaluate six ML models, namely, Random Forests (RF), Logistic Regression (LR), Extreme Gradient Boosting (XGBoost), K-Nearest Neighbor (KNN), Ada Boosting, Gradient Boosting Machine (GBM), for the detection of cardioembolism (CE), large-artery atherosclerosis (LAA), and small-artery occlusion (SAO). Between October 2016 and April 2020, patients were enrolled consecutively for algorithm development (phase one). Between June 2020 and December 2020, patients were enrolled consecutively in a test set for algorithm test (phase two). Area under the curve (AUC), precision, recall, accuracy, and F1 score were calculated for the prediction model.

Results

Finally, a total of 18,209 patients were enrolled in phase one, including 13,590 patients (i.e., 6,089 CE, 4,539 LAA, and 2,962 SAO) in the model, and a total of 3,688 patients were enrolled in phase two, including 3,070 patients (i.e., 1,103 CE, 1,269 LAA, and 698 SAO) in the model. Among the six models, the best models were RF, XGBoost, and GBM, and we chose the RF model as our final model. Based on the test set, the AUC values of the RF model to predict CE, LAA, and SAO were 0.981 (95%CI, 0.978-0.986), 0.919 (95%CI, 0.911-0.928), and 0.918 (95%CI, 0.908-0.927), respectively. The most important items to identify CE, LAA, and SAO were atrial fibrillation and degree of stenosis of intracranial arteries.

Conclusion

The proposed RF model could be a useful diagnostic tool to help neurologists categorize etiologies of stroke.

Clinical Trial Registration

[www.ClinicalTrials.gov], identifier [NCT01274117].

Applications and Techniques for Fast Machine Learning in Science

In this community review report, we discuss applications and techniques for fast machine learning (ML) in science-the concept of integrating powerful ML methods into the real-time experimental data processing loop to accelerate scientific discovery. The material for the report builds on two workshops held by the Fast ML for Science community and covers three main areas: applications for fast ML across a number of scientific domains; techniques for training and implementing performant and resource-efficient ML algorithms; and computing architectures, platforms, and technologies for deploying these algorithms. We also present overlapping challenges across the multiple scientific domains where common solutions can be found. This community report is intended to give plenty of examples and inspiration for scientific discovery through integrated and accelerated ML solutions. This is followed by a high-level overview and organization of technical advances, including an abundance of pointers to source material, which can enable these breakthroughs.

Deep Learning Approach Using Diffusion-Weighted Imaging to Estimate the Severity of Aphasia in Stroke Patients

Background and Purpose

This study aimed to investigate the applicability of deep learning (DL) model using diffusion-weighted imaging (DWI) data to predict the severity of aphasia at an early stage in acute stroke patients.

Methods

We retrospectively analyzed consecutive patients with aphasia caused by acute ischemic stroke in the left middle cerebral artery territory, who visited Asan Medical Center between 2011 and 2013. To implement the DL model to predict the severity of post-stroke aphasia, we designed a deep feed-forward network and utilized the lesion occupying ratio from DWI data and established clinical variables to estimate the aphasia quotient (AQ) score (range, 0 to 100) of the Korean version of the Western Aphasia Battery. To evaluate the performance of the DL model, we analyzed Cohen’s weighted kappa with linear weights for the categorized AQ score (0–25, very severe; 26–50, severe; 51–75, moderate; ≥76, mild) and Pearson’s correlation coefficient for continuous values.

Results

We identified 225 post-stroke aphasia patients, of whom 176 were included and analyzed. For the categorized AQ score, Cohen’s weighted kappa coefficient was 0.59 (95% confidence interval [CI], 0.42 to 0.76; P<0.001). For continuous AQ score, the correlation coefficient between true AQ scores and model-estimated values was 0.72 (95% CI, 0.55 to 0.83; P<0.001).

Conclusions

DL approaches using DWI data may be feasible and useful for estimating the severity of aphasia in the early stage of stroke.