Revascularization Outcome Prediction for A Direct Aspiration-First Pass Technique (ADAPT) from Pre-Treatment Imaging and Machine Learning

• DOI: 10.3390/brainsci11101321
• Publication Date: 2021-10-06T14:36:57Z
• Authors: Tatsat R. Patel; Muhammad Waqas; Seyyed M. M. J. S...; Zeguang Ren; Cesario V. Borlongan; Rimal Dossani; Elad I. Levy; Adnan H. Siddiqui; Kenneth V. Snyder; Jason M. Davies; Maxim Mokin; Vincent M. Tutino
• Keywords: acute ischemic stroke; 03 medical and health sciences; machine learning; 0302 clinical medicine; first pass effect; Neurosciences. Biological psychiatry. Neuropsychiatry; ADAPT; Article; RC321-571

A direct aspiration-first pass technique (ADAPT) has recently gained popularity for the treatment of large vessel ischemic stroke. Here, we sought to create a machine learning-based model that uses pre-treatment imaging metrics predict successful outcomes ADAPT in middle cerebral artery (MCA) stroke cases. In 119 MCA strokes treated by ADAPT, calculated four parameters—clot length, perviousness, distance from internal carotid (ICA) and angle interaction (AOI) between clot/catheter. We determined success first effect (FPE), performed univariate analyses. further built validated multivariate learning models random train-test split (75%:25%) our data. To test stability, repeated procedure over 100 randomizations, reported average performances. Our results show perviousness (p = 0.002) AOI 0.031) were significantly higher clot length 0.007) was lower cases with FPE. logistic regression achieved highest accuracy (74.2%) testing cohort, an AUC 0.769. The had similar performance randomizations (average 0.768 ± 0.026). This study provides feasibility imaging-based predictors outcome. Such may help operators select most adequate thrombectomy approach.