Prediction of Tissue Outcome and Assessment of Treatment Effect in Acute Ischemic Stroke Using Deep Learning

Oh fuck, another prediction tool. Survivors don't care about prediction you blithering idiots. They want 100% recovery. GET THERE! Have you ever talked to survivors?
 http://stroke.ahajournals.org/content/49/6/1394?etoc=

Anne Nielsen, Mikkel Bo Hansen, Anna Tietze, Kim Mouridsen

https://doi.org/10.1161/STROKEAHA.117.019740

Stroke. 2018;49:1394-1401

Originally published May 2, 2018

• Article
• Figures & Tables
• Supplemental Materials
• Info & Metrics

This article requires a subscription to view the full text. If you have a subscription you may use the login form below to view the article. Access to this article can also be purchased.

---

Abstract

Background and Purpose—Treatment options for patients with acute ischemic stroke depend on the volume of salvageable tissue. This volume assessment is currently based on fixed thresholds and single imagine modalities, limiting accuracy. We wish to develop and validate a predictive model capable of automatically identifying and combining acute imaging features to accurately predict final lesion volume.

Methods—Using acute magnetic resonance imaging, we developed and trained a deep convolutional neural network (CNN_deep) to predict final imaging outcome. A total of 222 patients were included, of which 187 were treated with rtPA (recombinant tissue-type plasminogen activator). The performance of CNN_deep was compared with a shallow CNN based on the perfusion-weighted imaging biomarker Tmax (CNN_Tmax), a shallow CNN based on a combination of 9 different biomarkers (CNN_shallow), a generalized linear model, and thresholding of the diffusion-weighted imaging biomarker apparent diffusion coefficient (ADC) at 600×10⁻⁶ mm²/s (ADC_thres). To assess whether CNN_deep is capable of differentiating outcomes of ±intravenous rtPA, patients not receiving intravenous rtPA were included to train CNN_{deep, −rtpa} to access a treatment effect. The networks’ performances were evaluated using visual inspection, area under the receiver operating characteristic curve (AUC), and contrast.

Results—CNN_deep yields significantly better performance in predicting final outcome (AUC=0.88±0.12) than generalized linear model (AUC=0.78±0.12; Tmax

(AUC=0.72±0.14; thres (AUC=0.66±0.13; P<0.0001) and a substantially better performance than CNN_shallow (AUC=0.85±0.11; P=0.063). Measured by contrast, CNN_deep improves the predictions significantly, showing superiority to all other methods (P≤0.003). CNN_deep also seems to be able to differentiate outcomes based on treatment strategy with the volume of final infarct being significantly different (P=0.048).

Conclusions—The considerable prediction improvement accuracy over current state of the art increases the potential for automated decision support in providing recommendations for personalized treatment plans.