Both Infarcted and Noninfarcted Brain Regions Contribute to Deep Learning-Based MRI Prediction of Acute Stroke Outcome.
Authors
Affiliations (7)
Affiliations (7)
- From the Department of Radiology (Y.L., B.J., H.v.V., H.F., Z.Z., S. Luo, M.E.M., J.J.H., G.Z.), Stanford University, Stanford, California [email protected].
- From the Department of Radiology (Y.L., B.J., H.v.V., H.F., Z.Z., S. Luo, M.E.M., J.J.H., G.Z.), Stanford University, Stanford, California.
- Department of Radiology (Y.Y.), University of California, San Francisco, San Francisco, California.
- Department of Computer and Data Sciences, Case School of Engineering (S. Li), Case Western Reserve University, Cleveland, Ohio.
- Department of Neurology (D.S.L.), UCLA, Los Angeles, California.
- Department of Neuroradiology (G.A., M.W.), University of Texas MD Anderson Center, Houston, Texas.
- Department of Neurology (M.G.L.), Stanford University, Stanford, California.
Abstract
Predicting long-term clinical outcomes based on early acute ischemic stroke (AIS) information would be useful for many reasons, including patient counseling and clinical trial execution. This study investigates how different regions in brain imaging, including noninfarcted areas, contribute to the accuracy of predicting 90-day stroke outcomes by using deep learning (DL). We developed and validated DL models in 449 patients with AIS, by using MRI DWI scans from 1-7 days poststroke and 90-day mRS outcome data. These models were trained on various inputs: infarct volumes, full-brain images, infarct masks, intensity-preserved infarct masks, and images in which the infarct region is removed, which we call lesion-neutralized images. Performance was assessed by using accuracy of predicting the specific mRS score, accuracy within ±1 mRS category, mean absolute error (MAE), and the area under the curve (AUC) to predict unfavorable outcome (mRS > 2). The model trained by using only infarct volume size reported the highest (worst) MAE of 1.51 points (95% CI, 1.40-1.61; <i>P</i> < .001), while the model trained with full-brain images achieved the lowest MAE of 1.07 points (95% CI, 0.99-1.16). Models with intermediate amounts of imaging information each improved on the volume-only predictions but did not reach the performance of the full brain images; infarct masks, intensity-preserved infarct masks, and lesion-neutralized images demonstrated MAEs of 1.25 (95% CI, 1.15-1.34; <i>P</i> = .002), 1.21 (95% CI, 1.11-1.30; <i>P</i> = .008), and 1.35 (95% CI, 1.24-1.45; <i>P</i> < .001), respectively. Similar results were seen for other prediction tasks, including AUC to predict unfavorable outcomes, ranging from 0.68 (95% CI, 0.63-0.73) for infarct volume to 0.86 (95% CI, 0.82-0.89) for full brain inputs. While the best performance came from by using the full brain imaging volume, we demonstrate that the infarct location, its signal characteristics, and importantly, the noninfarcted regions all contribute to the predictions. The noninfarcted areas may be a proxy for overall brain health and resilience, containing important information about potential outcomes.