Value of machine learning to predict functional outcome of endovascular treatment for acute ischaemic stroke of the posterior circulation.
05 Oct 2021-Rivista Di Neuroradiologia (SAGE PublicationsSage UK: London, England)-pp 19714009211049088
TL;DR: In this article, the authors evaluated machine learning algorithms in their ability to detect vessel occlusion of the posterior circulation in individuals with vessel stroke, and found that they performed well.
Abstract: PurposeClinical outcomes vary considerably among individuals with vessel occlusion of the posterior circulation. In the present study we evaluated machine learning algorithms in their ability to di...
Citations
More filters
TL;DR: In this article , Wang et al. developed an effective machine learning model to preoperatively predict the occurrence of futile recanalization (FR) of acute basilar artery occlusion (ABAO) patients with endovascular treatment (EVT).
16 Jan 2023-Journal of stroke and cerebrovascular diseases : official publication of the National Stroke Association
TL;DR: In this article , the authors trained and internally validated a ML model that predicts malignant middle cerebral artery infarction (MMI) following mechanical thrombectomy (MT) for ACLVO.
Abstract: Prediction of malignant middle cerebral artery infarction (MMI) could identify patients for early intervention. We trained and internally validated a ML model that predicts MMI following mechanical thrombectomy (MT) for ACLVO.All patients who underwent MT for ACLVO between 2015 - 2021 at a single institution were reviewed. Data was divided into 80% training and 20% test sets. 10 models were evaluated on the training set. The top 3 models underwent hyperparameter tuning using grid search with nested 5-fold CV to optimize the area under the receiver operating curve (AUROC). Tuned models were evaluated on the test set and compared to logistic regression.A total of 381 patients met the inclusion criteria. There were 50 (13.1%) patients who developed MMI. Out of the 10 ML models screened on the training set, the top 3 performing were neural network (median AUROC 0.78, IQR 0.72 - 0.83), support vector machine ([SVM] median AUROC 0.77, IQR 0.72 - 0.83), and random forest (median AUROC 0.75, IQR 0.68 - 0.81). On the test set, random forest (median AUROC 0.78, IQR 0.73 - 0.83) and neural network (median AUROC 0.78, IQR 0.73 - 0.83) were the top performing models, followed by SVM (median AUROC 0.77, IQR 0.70 - 0.83). These scores were significantly better than those for logistic regression (AUROC 0.72, IQR 0.66 - 0.78), individual risk factors, and the Malignant Brain Edema score (p < 0.001 for all).ML models predicted MMI with good discriminative ability. They outperformed standard statistical techniques and individual risk factors.
TL;DR: In this article , the authors evaluated various machine learning algorithms in their ability to classify between patients with favorable (defined as 0-2 points on the modified Rankin scale [mRS]), unfavorable (mRS 3-6), poor (m RS 5-6) and nonpoor (n = 0-4) outcomes at dismissal.
Abstract: PURPOSE
Various studies have identified prognostic factors for a favorable outcome of endovascular treatment in posterior circulation. We evaluated various machine learning algorithms in their ability to classify between patients with favorable (defined as 0-2 points on the modified Rankin scale [mRS]), unfavorable (mRS 3-6), poor (mRS 5-6), and nonpoor (mRS 0-4) outcomes at dismissal.
METHODS
We retrospectively analyzed data from 415 patients that were treated between 2018 and 2021 from the multicentric DGNR registry. Five models (random forest, support vector machine, k-nearest neighbor, neural network [NN], and generalized linear model [GLM]) were trained with clinical input variables and evaluated with a test dataset of 82 patients. The model with the highest accuracy on the training dataset was defined as the best model.
RESULTS
A total of 132 patients showed poor and 162 patients showed favorable outcome. All baseline variables except sex were highly significantly different between patients with favorable and unfavorable outcomes. The variables NIHSS, the presence of wake-up stroke, the administration of IV-thrombolysis and mRS pretreatment were significantly different between patients with poor and nonpoor outcomes. The best-performing NN achieved a sensitivity of 0.56, a specificity of 0.86 and an area under the curve (AUC) of 0.77 on the test dataset in the classification analysis between favorable and unfavorable outcomes. The best-performing GLM achieved a sensitivity of 0.65, a specificity of 0.91 and an AUC of 0.81 in the classification analysis between poor and nonpoor outcomes.
CONCLUSION
Short-term favorable and poor outcomes in patients with acute ischemic stroke of the posterior circulation can be predicted prior to thrombectomy with moderate sensitivity and high specificity with machine learning models.
References
More filters
[...]
TL;DR: Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the forest, and are also applicable to regression.
Abstract: Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The generalization error for forests converges a.s. to a limit as the number of trees in the forest becomes large. The generalization error of a forest of tree classifiers depends on the strength of the individual trees in the forest and the correlation between them. Using a random selection of features to split each node yields error rates that compare favorably to Adaboost (Y. Freund & R. Schapire, Machine Learning: Proceedings of the Thirteenth International conference, aaa, 148–156), but are more robust with respect to noise. Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the splitting. Internal estimates are also used to measure variable importance. These ideas are also applicable to regression.
79,257 citations
University of Calgary1, Maastricht University2, Erasmus University Rotterdam3, Royal Melbourne Hospital4, University of Amsterdam5, Bellvitge University Hospital6, Florey Institute of Neuroscience and Mental Health7, UCLA Medical Center8, University Hospital Bonn9, State University of New York System10, University of Toronto11, Beaumont Hospital12, Philadelphia College of Osteopathic Medicine13, Altair Engineering14, University of California, Los Angeles15, University of Pittsburgh16
TL;DR: Endovascular thrombectomy is of benefit to most patients with acute ischaemic stroke caused by occlusion of the proximal anterior circulation, irrespective of patient characteristics or geographical location, and will have global implications on structuring systems of care to provide timely treatment.
4,846 citations
TL;DR: It is found that the Support Vector Machine (SVM) algorithm is applied most frequently (in 29 studies) followed by the Naïve Bayes algorithm (in 23 studies), however, the Random Forest algorithm showed superior accuracy comparatively.
Abstract: Supervised machine learning algorithms have been a dominant method in the data mining field. Disease prediction using health data has recently shown a potential application area for these methods. This study aims to identify the key trends among different types of supervised machine learning algorithms, and their performance and usage for disease risk prediction. In this study, extensive research efforts were made to identify those studies that applied more than one supervised machine learning algorithm on single disease prediction. Two databases (i.e., Scopus and PubMed) were searched for different types of search items. Thus, we selected 48 articles in total for the comparison among variants supervised machine learning algorithms for disease prediction. We found that the Support Vector Machine (SVM) algorithm is applied most frequently (in 29 studies) followed by the Naive Bayes algorithm (in 23 studies). However, the Random Forest (RF) algorithm showed superior accuracy comparatively. Of the 17 studies where it was applied, RF showed the highest accuracy in 9 of them, i.e., 53%. This was followed by SVM which topped in 41% of the studies it was considered. This study provides a wide overview of the relative performance of different variants of supervised machine learning algorithms for disease prediction. This important information of relative performance can be used to aid researchers in the selection of an appropriate supervised machine learning algorithm for their studies.
580 citations
TL;DR: It is argued that opaque decisions are more common in medicine than critics realize and that ceding medical decision-making to black box systems as contravening the profound moral responsibilities of clinicians should be considered.
Abstract: Although decision-making algorithms are not new to medicine, the availability of vast stores of medical data, gains in computing power, and breakthroughs in machine learning are accelerating the pace of their development, expanding the range of questions they can address, and increasing their predictive power. In many cases, however, the most powerful machine learning techniques purchase diagnostic or predictive accuracy at the expense of our ability to access "the knowledge within the machine." Without an explanation in terms of reasons or a rationale for particular decisions in individual cases, some commentators regard ceding medical decision-making to black box systems as contravening the profound moral responsibilities of clinicians. I argue, however, that opaque decisions are more common in medicine than critics realize. Moreover, as Aristotle noted over two millennia ago, when our knowledge of causal systems is incomplete and precarious-as it often is in medicine-the ability to explain how results are produced can be less important than the ability to produce such results and empirically verify their accuracy.
333 citations
TL;DR: Stent retriever thrombectomy is a safe treatment modality for patients with stroke presenting with BAO, and although the stent retrievers showed a good recanalisation rate, there are currently no randomised clinical trials to assess its clinical efficacy in comparison with the reference treatment.
Abstract: Background Basilar artery occlusion (BAO) remains one of the most devastating subtypes of stroke with high mortality and poor outcome. Early recanalisation is the most powerful predictor of favourable outcome in patients with stroke, and may be improved with mechanical thrombectomy using stent retriever devices. However, the benefit in functional outcome and safety of stent retrievers are not yet well known. The aim of this study was to assess efficacy and safety profiles of stent retriever thrombectomy in BAO patients with stroke. Methods We analysed data retrospectively from our consecutive clinical series and conducted a systematic review and meta-analysis of all previous studies of stent retriever thrombectomy in BAO patients with stroke between November 2010 and April 2014. Results From March 2010 to March 2013, 22 patients with acute BAO were treated with a Solitaire stent retriever in our series. Favourable outcome was significantly associated with younger age and distal BAO. The literature search identified 15 previous studies involving a total of 312 subjects. In the meta-analysis, including our series data, the recanalisation rate (Thrombolysis In Cerebral Infarction (TICI) score ≥2b) reached 81% (95% CI 73% to 87%). The rate of symptomatic intracranial haemorrhage was 4% (95% CI 2% to 8%), favourable outcome (modified Rankin Scale (mRS) ≤2 at 3 months) was found in 42% (95% CI 36% to 48%) and mortality rate was 30% (95% CI 25% to 36%). Conclusions Stent retriever thrombectomy is a safe treatment modality for patients with stroke presenting with BAO. Although the stent retrievers showed a good recanalisation rate, there are currently no randomised clinical trials to assess its clinical efficacy in comparison with the reference treatment.
136 citations