Send to

Choose Destination
J Stroke Cerebrovasc Dis. 2012 Nov;21(8):704-11. doi: 10.1016/j.jstrokecerebrovasdis.2011.03.004. Epub 2011 Apr 20.

Motor outcome for patients with acute intracerebral hemorrhage predicted using diffusion tensor imaging: an application of ordinal logistic modeling.

Author information

Department of Rehabilitation Medicine, Nishinomiya Kyoritsu Neurosurgical Hospital, Nishinomiya, Hyogo, Japan.


This study examined the clinical usefulness of magnetic resonance-diffusion tensor imaging (DTI) for predicting motor outcome in patients with intracerebral hemorrhage. We studied 15 subjects (age range, 31-81 years) diagnosed by conventional computed tomography with thalamic hemorrhage, putaminal hemorrhage, or both. DTI data were obtained on days 14-18 after diagnosis. Mean fractional anisotropy (FA) values within the right and left cerebral peduncles were estimated by a computer-automated method. Using logistic regression analyses, the ratios of FA values in the affected and unaffected hemispheres (rFA) were modeled in relation to motor outcome scores at 1 month after onset, assessed using the Medical Research Council (MRC) scale (0 = null to 5 = full). The rFA values ranged from 0.628 to 1.001 (median value, 0.856). Analyses showed that the relationships between rFA and MRC scale matched the logistic probabilities for both the upper extremities (R(2) = 0.272; P < .001) and lower extremities (R(2) = 0.247; P < .001). When estimated rFA values were <0.7, the estimated probability of an MRC score of 0-1 was close to 80% for the upper extremities and 65% for the lower extremities. Meanwhile, when estimated rFA values were >0.9, the estimated probability of an MRC score of 3-5 was close to 60% for the upper extremities and 80% for the lower extremities. Our data indicate that for patients with intracerebral hemorrhage, DTI is a useful tool for quantitatively predicting motor outcome, suggesting wider clinical applicability of this method for outcome prediction.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center