Detection of Hyperacute Primary Intraparenchymal Hemorrhage by Magnetic Resonance Imaging
Background MRI has become increasingly used in the acute setting to manage patients with stroke. There has been concern that MRI may not be sensitive in the detection of acute intracranial hemorrhage. We assessed whether strongly susceptibility-weighted MRI would be sensitive to intraparenchymal hemorrhage in the first few hours.
Case Descriptions In the course of our ongoing studies of MRI of acute ischemic stroke in more than 200 patients, 35 patients had MR studies within 6 hours. Six of these patients who presented with acute focal symptoms with definite time of clinical onset (2.5 to 5 hours) were found to have evidence of intraparenchymal hemorrhage. Standard T1- and T2-weighted MR scans were performed. In 5 of the patients, echo-planar imaging and gradient-echo sequences were performed to increase the sensitivity of magnetic susceptibility effects of the pulse sequences. Four of the cases were of putaminal hemorrhage and 2 were lobar hemorrhages. The hemorrhage was most evident as foci of T2* hypointensity (signal loss) and unambiguous on the more susceptibility-weighted sequences, particularly echo-planar gradient-echo images.
Conclusions MRI can detect hemorrhage within 2.5 to 5 hours of onset of clinical symptoms as regions of marked signal loss due to susceptibility effects, whereas conventional MR scans of ischemic stroke may appear normal. These results demonstrate that MR susceptibility sequences may be sensitive to hyperacute hemorrhage and suggest that MR may be an adequate screen for primary intraparenchymal hemorrhage.
The intracranial manifestation of blood on MRI has a complicated appearance, dependent on its evolution over time and its location.1 2 3 4 5 6 7 8 9 10 11 12 13 Although the sensitivity and efficacy of MRI to detect acute intracranial hemorrhage have been evaluated to some extent,2 3 4 5 6 7 its ability to ascertain the presence of hyperacute hemorrhage (within the first 6 hours) has been less well studied.
After intraparenchymal hemorrhage, blood goes through several stages of breakdown that are detectable by MRI. The earliest is the transition from oxyhemoglobin to deoxyhemoglobin. The latter is a paramagnetic molecule that causes signal loss (darkening) in images due to magnetic susceptibility effects, which are best seen on T2*-weighted (or susceptibility-weighted) pulse sequences.14 Although little data exist regarding the time of transition to deoxyhemoglobin, most authorities have stated that it does not become apparent until several hours up to a few days.8 9
A number of previous studies claiming that MRI is insensitive to the detection of acute intracranial hemorrhage were performed at a low field strength (≤0.6 T),4 5 6 7 and the results may not be valid at a field strength of 1.5 T since susceptibility effects, and therefore sensitivity to hemorrhage, increase with magnetic field strength. We report our experience in identifying hyperacute intraparenchymal hemorrhage with MRI.
Subjects and Methods
As part of a longitudinal study of diffusion and perfusion MRI in acute stroke, we have evaluated more than 200 patients from 1990 to 1996. Within this sample, 35 patients had MR studies within 6 hours. Of these patients, six were found to have MR evidence of primary intraparenchymal hemorrhage within 2.5 to 5 hours after the onset symptoms, confirmed by CT findings. No cases of hemorrhage were missed by MRI based on CT, follow-up MRI, and/or clinical course. In all cases the patients were awake when symptoms began, and an accurate time of onset was determined.
MR studies of the brain were performed on a 1.5-T Siemens SP Magnetom (for patient 3), a 1.0-T Siemens Expert (for patient 6), or a modified 1.5-T Siemens Vision scanner with echo-planar imaging (EPI) capability. For each patient we obtained an axial conventional or turbo spin-echo T2-weighted sequence with the following typical parameters: repetition time (TR), 2570 ms; echo time (TE), 22 or 90 ms; excitations, 1; slice thickness, 5 mm; matrix size, 190×256; and field of view, 250×250 mm. We also obtained axial T1-weighted images with the following typical parameters: TR, 600 ms; TE, 15 ms; excitations, 1; matrix size, 128×256; and slice thickness, 5 mm. Susceptibility-weighted gradient-echo images (subsequently referred to as gradient-echo) were obtained with the following parameters: TR, 800 ms; TE, 26 ms; excitations, 1; and flip angle, 30 degrees.
Additional EPI sequences were acquired with the use of single-shot EPI with TE of 60 ms for the EPI gradient-echo sequence and an effective TE of 80, 100, or 118 ms for the EPI T2-weighted spin-echo sequences. The EPI sequences, including the spin-echo EPI sequences, intrinsically have greater sensitivity to magnetic susceptibility effects.
The diagnosis of intraparenchymal hemorrhage was made on the basis of location and marked regions of signal loss due to susceptibility effects characteristic of MR appearance of hemorrhage. The diagnosis was also confirmed by CT in all cases.
A 67-year-old right-handed woman, with a history of hypertension and hypercholesterolemia, noticed a sudden onset of left upper extremity weakness and left facial numbness, with headache and nausea. The physical examination was notable for a moderately severe paresis of the left face and left upper extremity and some decrease in pin perception in the left face (all three divisions). An MRI study was performed 2.5 hours after the event with the susceptibility-weighted images demonstrating internal components of marked signal loss, with some central regions of T2 hyperintensity (Figure⇓, panels A and B). T2-weighted images (panels C and D) revealed T2 hyperintensity centrally with a thin peripheral rim of hypointensity and a further surrounding ring of T2 hyperintensity. The latter likely represented edema. The T1-weighted images (panel E) revealed a 2-cm isointense to white matter lesion in the right central gyrus with surrounding T1 hypointensity comparable to gray matter. A CT scan 18 hours later confirmed hemorrhage with high-attenuation blood in the same area (panel F).
A 46-year-old right-handed black man, while sitting at his desk at work, developed a sudden onset of progressive right-sided weakness and sensory loss. In the emergency department his blood pressure was 230/120 mm Hg. On neurological examination, a right hemiplegia with right-sided sensory loss was noted. The patient also had word-finding difficulties.
An MRI study was performed after 3 hours of onset. Gradient-echo images were not obtained. Axial EPI T2-weighted images, which are sensitive to susceptibility effects, demonstrated a heterogeneous lesion within the left putamen, with a rim of T2 hypointensity as well as internal components of signal loss consistent with susceptibility. A small amount of surrounding T2 hyperintensity was consistent with edema. An axial T2-weighted image demonstrated some susceptibility-induced signal loss, although less than the EPI T2-weighted image. A T1-weighted image demonstrated a 2.5-cm isointense to white matter lesion within the left putamen with a peripheral rim of T1 hypointensity. A CT scan performed 9.5 hours later confirmed a high-attenuation left putaminal hypertensive hemorrhage with a small amount of surrounding edema.
An 80-year-old right-handed man with a history of alcohol abuse and a seizure disorder had been swimming with his personal physician and developed difficulty with speech and the use of his right hand while buttoning his sleeve. In the emergency department he exhibited fluent aphasia with poor repetition and comprehension, as well as multiple paraphasias and a right homonymous hemianopsia.
An MRI study 4 hours after the event revealed a 5-cm left posterior parietal T2 lesion with components of signal loss due to susceptibility, and a T1 heterogeneous lesion with a fluid/fluid level with the dependent component predominantly isointense to white matter and the more superior portion hypointense to gray matter but slightly hyperintense to cerebrospinal fluid. A surrounding rim of T1 hypointensity was seen. A CT scan performed 4 days latter demonstrated a 5-cm left parietal hemorrhage with a small amount of surrounding edema without evidence of a fluid level.
A 54-year-old right-handed woman developed numbness involving the left hand, and upon standing for a few minutes her left leg gave way with numbness and weakness. Upon getting out of bed, she was unable to use the left side of the body. Her blood pressure on admission was 170/100 mm Hg, and she had a dense left hemianesthesia with mild weakness of the left upper and lower extremities.
Axial EPI susceptibility-weighted images 5 hours after the event demonstrated marked signal loss in the right posterior putamen consistent with a hypertensive hemorrhage. A conventional gradient-echo axial image demonstrated clearly seen signal loss, but it was slightly diminished from the EPI susceptibility-weighted image. An EPI T2-weighted image demonstrated signal loss within the right putamen, although it was diminished in amount from the two prior sequences. A T2-weighted image failed to convincingly demonstrate susceptibility-induced signal loss. A T1-weighted axial image revealed the majority of right putaminal lesion to be isointense to white matter with a faint peripheral rim of hypointensity.
An 81-year-old right-handed man with a history of hypertension who had been taking aspirin and nonsteroidal anti-inflammatory drugs for elbow inflammation was speaking on the telephone to his daughter when he suddenly developed slurred speech and right upper extremity weakness. In the emergency department he had a blood pressure of 203/85 mm Hg. His examination was notable for moderate dysarthria and right facial weakness with a mild upper motor neuron pattern of weakness in the right upper extremity.
MRI was performed 5 hours after onset of symptoms and revealed a 3-cm left putaminal lesion with marked foci of hypointensity on the T2-weighted EPI sequence consistent with hemorrhage, with a small amount of surrounding T2 hyperintensity. T1-weighted images revealed the lesion to be T1 hypointense with a small peripheral rim of hypointensity. EPI susceptibility sequences could not be obtained because the patient became short of breath.
A 55-year-old black left-handed man with long-standing poorly controlled hypertension was returning home from work and collapsed at the door of his house, with sudden onset of right-sided weakness and numbness. In the emergency department his blood pressure was 260/140 mm Hg, and on examination he had a right hemiplegia and hemisensory loss.
An MR study at 1.0 T performed 5 hours later demonstrated a 5-cm left putaminal lesion with foci of hypointensity on the gradient-echo images. Axial T2-weighted images demonstrated a hyperintense lesion without evidence of internal foci of signal loss. A T1- weighted image revealed an isointense lesion with a peripheral rim of hypointensity.
A summary of the typical imaging findings in hyperacute hemorrhage is demonstrated in the Table⇓.
All cases of intracranial hemorrhage could be detected on the MR scans as foci of increased susceptibility, causing signal loss (darkening), particularly on the gradient-echo susceptibility-weighted and EPI T2-weighted images. Our observations demonstrate that in the clinical setting, it is possible to detect hyperacute intracranial hemorrhage on MRI. The earliest increased susceptibility is probably due to the paramagnetic effects of deoxyhemoglobin, which is the earliest observable step in hemoglobin breakdown as seen by MRI.1
Intracranial hemorrhage less than 24 hours old has been postulated to be in the oxyhemoglobin phase and would be isointense on T1-weighted images and slightly hyperintense on T2-weighted images.1 Our results suggest that the transition to deoxyhemoglobin occurs within the first few hours of a hematoma and is detectable by susceptibility-weighted MR pulse sequences. Prior pessimism about MR sensitivity to the detection of acute intraparenchymal hemorrhage was based largely on investigations at lower field strengths or pulse sequences not as weighted toward susceptibility effects.4 5 6 7 Few clinical observations by MRI of acute bleeds within the first 5 hours are found in the literature to refute or support that view. We have shown in six consecutive cases of intraparenchymal bleeds within 5 hours of onset that MRI, particularly if susceptibility weighted with EPI, is sensitive to acute blood and has a characteristic appearance that makes confusion with other etiologies unlikely. Susceptibility weighting can cause artifactual signal loss and anatomic distortion in the inferior frontal and inferior temporal lobes, although that did not interfere with diagnosis in these cases.
An additional feature of all the cases in this report is a peripheral rim of hypointensity in the acute phase on both T1- and T2-weighted images. This has been previously reported and has been speculated by Thulborn and Atlas9 to represent a rim of storage iron due to early phagocytic activity, or it may represent a region of dephasing at the boundary of two tissues (blood and brain) with differing magnetic susceptibility.14 None of the hemorrhages had T1 hyperintensity, suggesting that evolution to met-hemoglobin blood breakdown products had not yet occurred within 5 hours.
We suggest that MRI at 1.5 T may be more sensitive to a hyperacute intraparenchymal hemorrhage than has previously been thought. With study of a larger patient group, the need for a CT to exclude an intraparenchymal hemorrhage in addition to an emergency MR study may prove unnecessary. A single MR examination to rule out hemorrhage as well as to provide standard, more sensitive diagnostic information15 could be cost-effective in the management of stroke patients as opposed to a head CT to rule out hemorrhage followed by MRI. This is particularly important in view of emerging therapies, including tissue plasminogen activator in the management of acute stroke, that optimally involve a short time window.16 In the setting of acute stroke, this would further enhance the utility of MRI, which is more sensitive than CT in detecting the area of affected brain, particularly when used in conjunction with innovative techniques such as diffusion and perfusion imaging.17 18 19
Reprint requests to Mahesh R. Patel, MD, Department of Radiology, Beth Israel Hospital, 330 Brookline Ave, Boston, MA 02115. E-mail firstname.lastname@example.org.
- Received June 17, 1996.
- Revision received September 3, 1996.
- Accepted September 3, 1996.
- Copyright © 1996 by American Heart Association
Hayman LA, Taber KH, Ford JJ, Bryan RN. Mechanisms of MR signal alteration by acute intracerebral blood: old concepts and new theories. AJNR Am J Neuroradiol. 1991;12:899-907.
Wingarten K, Zimmerman RD, Cahill PT, Deck MD. Detection of acute intracerebral hemorrhage on MR imaging: ineffectiveness of prolonged interecho interval pulse sequences. AJNR Am J Neuroradiol.. 1991;12:475-479.
Weingarten K, Zimmerman RD, Deo-Narine V, Markisz J, Cahill PT, Deck MD. MR imaging of acute intracranial hemorrhage: findings on sequential spin-echo and gradient-echo images in a dog model. AJNR Am J Neuroradiol.. 1991;12:457-467.
Di Chiro G, Brooks RA, Girton ME, Caporale T, Wright DC, Dwyer AJ, Horne MK III. Sequential MR studies of intracerebral hematomas in monkeys. AJNR Am J Neuroradiol.. 1986;7:193-199.
Mattle HP, Edelman RR, Schroth G, O'Reilly GV. Spontaneous and Traumatic Hemorrhage in Clinical and Magnetic Resonance Imaging, Vol 1. 2nd ed. Philadelphia, Pa: WB Saunders Co; 1996:652-702.
Thulborn KR, Atlas SW. Intracranial hemorrhage. In: Atlas SW, ed. Magnetic Resonance Imaging of the Brain and Spine. New York, NY: Raven Press; 1991:175-222.
Mittl RL Jr, Gomori JM, Schnall MD, Holland GA, Grossman RI, Atlas SW. Magnetization transfer effects in MR imaging of in vivo intracranial hemorrhage. AJNR Am J Neuroradiol.. 1993;14:881-891.
Edelman RR, Johnson K, Buxton R, Shoukimas G, Rosen BR, Davis KR, Brady TJ. MR of hemorrhage: a new approach. AJNR Am J Neuroradiol.. 1986;7:751-756.
Bryan RN, Levy LM, Whitlow WD, Killian JM, Preziosi TJ, Rosario JA. Diagnosis of acute cerebral infarction: comparison of CT and MR imaging. AJNR Am J Neuroradiol.. 1991;12:611-620.
Warach S, Chien D, Li W, Ronthal M, Edelman RR. Fast magnetic resonance diffusion-weighted imaging of acute human stroke [published erratum appears in Neurology.. 1992;42:2192]. Neurology. 1992;42:1717-1723.