Population-based Stroke Atlas for outcome prediction: method and preliminary results for ischemic stroke from CT

Abstract

Background and purpose: Knowledge of outcome prediction is important in stroke management. We propose a lesion size and location-driven method for stroke outcome prediction using a Population-based Stroke Atlas (PSA) linking neurological parameters with neuroimaging in population. The PSA aggregates data from previously treated patients and applies them to currently treated patients. The PSA parameter distribution in the infarct region of a treated patient enables prediction. We introduce a method for PSA calculation, quantify its performance, and use it to illustrate ischemic stroke outcome prediction of modified Rankin Scale (mRS) and Barthel Index (BI).

Methods: The preliminary PSA was constructed from 128 ischemic stroke cases calculated for 8 variants (various data aggregation schemes) and 3 case selection variables (infarct volume, NIHSS at admission, and NIHSS at day 7), each in 4 ranges. Outcome prediction for 9 parameters (mRS at 7th, and mRS and BI at 30th, 90th, 180th, 360th day) was studied using a leave-one-out approach, requiring 589,824 PSA maps to be analyzed.

Results: Outcomes predicted for different PSA variants are statistically equivalent, so the simplest and most efficient variant aiming at parameter averaging is employed. This variant allows the PSA to be pre-calculated before prediction. The PSA constrained by infarct volume and NIHSS reduces the average prediction error (absolute difference between the predicted and actual values) by a fraction of 0.796; the use of 3 patient-specific variables further lowers it by 0.538. The PSA-based prediction error for mild and severe outcomes (mRS = [2]-[5]) is (0.5-0.7). Prediction takes about 8 seconds.

Conclusions: PSA-based prediction of individual and group mRS and BI scores over time is feasible, fast and simple, but its clinical usefulness requires further studies. The case selection operation improves PSA predictability. A multiplicity of PSAs can be computed independently for different datasets at various centers and easily merged, which enables building powerful PSAs over the community.

Figure 1. Illustration of PSA calculation and outcome prediction.

Top) processing of a single patient (case) contributing to formation of the PSA. Bottom) formation of the PSA from its contributing patients (left) and PSA-based prediction (right). The horizontal arrow represents weighting dependency between the PSA and a predicted case.

Figure 2. The software platform for PSA calculation and illustration of PSA-based prediction.

The calculated maps of interest and the cases (patients) to be predicted are selectable from the first two top panels on the right. For illustration, the mRS90 map is selected here and shown as an axial image along with the superimposed normalized contour of the case under prediction (in the left hemisphere). The results of mRS90 prediction (the mean value of 4.25) along with the actual value for this patient (of 4) are shown in the right-bottom panel.

Figure 3. Examples of PSA maps calculated for w₁ weighting (i.e., parameter and scan averaging): a) mRS (from the left to the right mRS7, mRS30, mRS90, mRS180, mRS360); b) BI (from the left to the right BI30, BI90, BI180, BI360); c) NIHSS at admission (note that the left hemisphere intensity of the NIHSS map is higher than that of the right hemisphere corresponding to the fact that patients with a right sided ischemic stroke are associated with a lower NIHSS score [45]); d) NCCT image intensity (infarct frequency) distribution.

Image intensity, proportional to map value, was normalized to 0–255 range. Note the trends over time in the similar locations of the mRS and BI maps (demonstrating the decrease in intensity for mRS and the increase in intensity for BI) which correspond to the improvement of outcomes over time (as the patients with up to one year survival were included). The images are in the radiological convention.