Table 2 Observation† counts per prediction class for each task

From: Hybrid machine learning for real-time prediction of edema trajectory in large middle cerebral artery stroke

Data set

Task

Maximum MLS class

  

0mm

0–3mm

3–8mm

> 8mm

Mass General Brigham

8-h

2012

1568

3405

1530

(derivation cohort)

24-h

2638

2713

6867

3478

Boston Medical Center

8-h

523

309

620

439

(validation cohort)

24-h

873

696

1223

921

  1. †Observations are transformed hourly sets of data for a given patient. Each patient may account for up to 8 observations per scan on the 8-h task, and up to 24 observations per scan on the 24-h task.
Back to article page