Susceptibility-Based Differentiation of Intracranial Calcification and Hemorrhage in Pediatric Patients

Abstract

Differential diagnosis of intracranial hemorrhage versus calcification on conventional magnetic resonance images (MRIs) is often challenging. Although computed tomography (CT) confirms calcification, phase information obtained during susceptibility-weighted imaging can be useful in distinguishing between 2 pathologies. Fourteen patients previously diagnosed to have hemorrhage or calcification with imaging were included in the study retrospectively. Phase shift values of hemorrhage and calcification were compared by using Student t test. The pathologies identified were tuberous sclerosis, Sturge-Weber syndrome, craniopharyngioma, congenital cytomegalovirus, subependymal hemorrhages, and hemorrhagic microembolic infarction. Calcifications appeared hypointense whereas hemorrhages were hyperintense on phase maps (left-handed magnetic resonance system). Statistical comparison of phase shift values yielded significant difference between hemorrhage versus calcification (P < .01). Phase maps were found to offer valuable data to differentiate 2 pathologies when used complementary to conventional magnetic resonance images. Considering the relatively higher risks of radiation exposure in children, susceptibility-weighted imaging with phase maps may help to waive radiation exposure from CT.

Keywords

intracranial hemorrhage calcification phase image susceptibility-weighted imaging

Differentiation of intracranial calcification and hemorrhage is important in many pediatric diseases for accurate diagnosis and treatment. Intracranial calcifications show varying signal intensities on conventional T1- and T2-weighted magnetic resonance images (MRIs) depending on the amount of calcium deposition in a lesion.¹ As a result, calcification and hemorrhage may appear similar in some cases. This makes distinguishing the 2 pathologies difficult and hence leads to misinterpretation. Computed tomography (CT) is considered to be the criterion standard in identification of calcification. However, CT employs ionizing radiation and constitutes risk particularly for the pediatric age group.

The distinction of the 2 pathologies in magnetic resonance imaging (MRI) can be performed by using phase images of gradient echo-based acquisitions.^2
–4 Gradient echo phase images are sensitive to magnetic field changes induced by different tissue susceptibilities. Tissues with calcium deposition such as tumoral or dystrophic calcifications exhibit diamagnetic characteristics that cause phase delay, whereas tissues with iron such as ferritin or hemosiderin display paramagnetic property that induce phase advance. The phase shift of a proton (δφ) depends on echo time (TE) and perturbation of magnetic field (δB) due to local gradients induced by susceptibility differences in tissues:

δ φ = - γ . T E . δ B,

where γ is the gyromagnetic ratio. Thus, different phase shifts in the voxels generate opposite signal intensities for tissues with calcification or hemorrhage in the gradient echo phase images. Calcification appears hypointense, whereas hemorrhage looks hyperintense on the phase maps produced in a left-handed magnetic resonance system such as a Siemens scanner.⁵

Susceptibility-weighted imaging is a relatively new MRI technique established based on a 3-D gradient echo sequence.^5

–10 A susceptibility-weighted image is produced by combining the magnitude image with the filtered phase image, thus enhancing the contrast based on local susceptibility differences in the tissue. It offers valuable information on various neurologic disorders including traumatic brain injury and vascular malformation.⁵ The phase maps obtained from the raw data of susceptibility-weighted imaging was shown to detect and differentiate intracranial calcification and hemorrhage in adult patients.⁴

The objective of this paper is to extend the application to pediatric patients and demonstrate feasibility and merits of using phase maps produced by susceptibility-weighted imaging acquisition to differentiate calcification and hemorrhage in this population. Susceptibility-weighted imaging has been a part of routine pediatric magnetic resonance protocol in our institution for almost a year. In the following, we present our experience with 8 cases of calcification and 6 cases of hemorrhage where the differentiation was attainable by utilizing phase maps complementary to T1-, T2-, and susceptibility-weighted images and discuss the results within the context of this objective.

Methods

We retrospectively analyzed medical records of pediatric patients scanned with susceptibility-weighted imaging sequence between March 2013 and March 2014. A total of 14 patients diagnosed with either intracranial calcification or hemorrhage were included in the study. All patients underwent brain MRI on a 1.5-Tesla clinical scanner (Magnetom Aera, Siemens, Erlangen, Germany) in our institution. The age and gender of each patient are given in Table 1. The magnetic resonance protocol consisted of (a) 3-dimensional T1-weighted imaging, (b) coronal and axial T2-weighted imaging, (c) axial fluid-attenuated inversion recovery, (d) axial diffusion-weighted imaging, and (e) susceptibility-weighted imaging. The parameters for susceptibility-weighted imaging acquisition were as follows: repetition time = 49 ms, time to echo = 40 ms, flip angle = 15°, field of view = 260 × 260 mm, resolution: 256 × 256, and acquisition time = 2 minutes 47 seconds. Susceptibility-weighted imaging acquisition produced magnitude and phase images. Phase data were high-pass filtered by the scanner’s software to create a spatial map depicting the smooth variation of phase for each slice for evaluation by the radiologists. Susceptibility-weighted imaging routine involved further postprocessing of the phase information.

Table 1.

Patients With Intracranial Hemorrhages and Calcifications.

Case	Age	Gender	Diagnosis	Differentiation	Lesion appearance
1	28 d	Male	Bilateral subependymal hemorrhage and prematurity	√	Homogeneous
2	2 y	Male	Bilateral subependymal hemorrhage and prematurity	√	Homogeneous
3	10 d	Male	Intraventricular hemorrhage and prematurity	√	Heterogeneous
4	1 d	Female	Intraventricular and choroid plexus hemorrhage and prematurity	√	Heterogeneous
5	32 d	Male	Hemorrhagic embolic infarcts and infective endocarditis	√	Homogeneous
6	9 y	Male	Subdural hemorrhage and trauma	×	Heterogeneous
7	4 mo	Male	Calcification and congenital cytomegalovirus infection	√	Homogeneous
8	8 y	Female	Calcification and tuberous sclerosis	√	Homogeneous
9	13 y	Female	Calcification and tuberous sclerosis	√	Homogeneous
10	14 y	Female	Calcification and tuberous sclerosis	√	Homogeneous
11	12 y	Male	Calcification and tuberous sclerosis	√	Homogeneous
12	4 y	Male	Calcification and tuberous sclerosis	√	Heterogeneous
13	15 y	Female	Gyral calcification and Sturge-Weber syndrome	√	Heterogeneous
14	9 y	Female	Tumoral calcification and craniopharyngioma	√	Heterogeneous

Two radiologists with 3 and 5 years of experience in pediatric radiology evaluated the images with consensus. Intracranial lesions were detected and diagnosed based on conventional MRIs (T1-weighted, T2-weighted, and fluid-attenuated inversion recovery), diffusion-weighted imaging, susceptibility-weighted imaging, and phase maps. The lesions with hyperintense appearance on the phase maps were considered as paramagnetic based on the bright appearance of veins on our images. On the other hand, the hypointense lesions were interpreted as diamagnetic. The lesions with mixed signals of opposite intensities were diagnosed based on the pattern of dominant signal. Two calcification cases were compared with previously acquired CT images. Because of the concern of radiation exposure caused by CT acquisitions, rest of the patients did not have cranial CT scans.

Region of interest was manually placed on hemorrhage and calcification to include a major part of the lesion. Average phase shifts were calculated and Student t test was used to compare the phase shifts of hemorrhage and calcifications, and P <.05 was considered significant.

Results

A total of 19 lesions (8 hemorrhage and 11 calcifications) were detected using MRIs. Seven hemorrhage and 11 calcifications were differentiated by use of phase maps together with conventional MRIs (Table 1). Discrimination of 1 hemorrhage close to the skull failed (Figure 2). Of all hemorrhage lesions, 5 showed homogenous signal intensity, whereas 3 were heterogeneous. Of all the calcifications, 8 showed homogenous signal intensity in contrast to 3 lesions that presented heterogeneous signal. The calcifications associated with tuberous sclerosis were observed to show mostly homogenous intensity (Table 1). Student t test between phase shifts of hemorrhage and calcification yielded significant difference (P < .01), indicating that hemorrhage and calcification show clear difference on susceptibility-weighted imaging phase maps (Figures 1 –5).

Magnetic resonance images of 5 children with intracranial hemorrhage are shown in Figure 1: left column depicts T2-weighted images, middle column shows susceptibility-weighted imaging images and right column displays the phase maps. The hemorrhages detected are shown with yellow arrows.

A 28-day-old neonate having abnormal signal intensities with relatively small size resembling calcifications on bilateral subependymal regions that line the lateral ventricles is shown (Figure 1, first row). T2-weighted image was not sufficient enough for a trustable diagnosis. For this case, the phase map excluded the possibility of calcification based on the bright appearance of the pathology.

T2-weighted image of a 2-year-old boy with the history of prematurity and epilepsy exhibits nodular contour of bilateral subependymal regions of lateral ventricles that would mimic subependymal calcified nodules of tuberous sclerosis (Figure 1, 2nd row). Susceptibility-weighted imaging image and phase map helped detect subependymal lesions that were not visible on T2-weighted images. The diagnosis of subependymal hemorrhage was confirmed with increased signals of the identified lesions.

Neonates (<1-month-old) were diagnosed with intraventricular and choroid plexus hemorrhage (Figure 1, third and fourth rows, respectively). The pathologic regions appeared bright on the phase maps, resulting in the diagnosis of hemorrhage.

A 1-month-old baby with congenital heart disease had microembolic hemorrhagic periventricular white matter lesions that could be distinguished from calcification owing to the brightness of the lesions on the phase maps (Figure 1, fifth row). The lesions were detected on susceptibility-weighted imaging, whereas all were isointense on the T2-weighted images.

The pathology located in the subdural space of the right temporal region, close to the calvarium, in a 9-year-old child is seen on the T2-weighted fluid-attenuated inversion recovery image (Figure 2). Phase map gave a somewhat bright appearance, indicating to the possibility of hemorrhage. However, it was not a strong diagnosis because of the location of pathology where the magnetic susceptibility difference between the skull and cortex was large. This patient was included in our study to show the limitation of the phase map as addressed in the discussion section below.

Calcification cases are displayed in Figure 3. In the first row, T2-weighted magnetic resonance image of a 4-month-old boy diagnosed with congenital cytomegalovirus infection revealed hypointense lesions of bilateral periventricular white matter that resembled either calcification or hemorrhage. Susceptibility-weighted image and phase map helped detect the number of lesions more accurately than the conventional sequences. Hypointense appearance on the phase map eliminated the possibility of hemorrhage.

The 5 of the cases in which calcification was detected had been diagnosed with tuberous sclerosis. In the second row of Figure 3, a case with subependymal calcified nodules (Table 1, case 11) and in the third row, a cortical hamartoma of the tuberous sclerosis (Table 1, case 13) were shown by the yellow arrows and appear dark on the phase maps. The white dots indicated by red arrows in the phase maps show blood vessels. Veins can be used as a reference marker if handedness of the scanner is not known by the radiologist. The lesion with the same intensity as vein is the paramagnetic one, whereas the one with the opposite intensity is diamagnetic. In the second row of Figure 3, the vein appears bright (red arrow) on the phase map, whereas the lesions appear dark, indicating calcification. This well illustrates the differentiation of blood products and calcifications on the phase maps of susceptibility-weighted imaging acquisitions.

Left parietooccipital calcified leptomeningeal angioma of an 8-year-old girl with Sturge-Weber syndrome was detected on the phase map (Figure 4). The previously acquired CT examination before the MRI scan of the patient validated the calcification of leptomeningeal angioma. The CT density of cortical angioma was measured as 473 Hounsfield units (HU). Note that the T2-weighted image revealed global atrophy of the left cerebral hemisphere and slight hypointensity of the gyral structures but was insufficient to make an accurate diagnosis of calcification. Susceptibility-weighted image and phase map played a critical role in making the correct diagnosis.

A mass lesion located within the suprasellar region and accompanying the cystic component in the left temporal lobe was detected on T2-weighted images (Figure 5). The phase map was more sensitive to reveal the extent of the hypointense components in the solid suprasellar part, and lining of the large cyst that was not possible to manifest on the T2-weighted image. The calcified components were proven to be calcifications with the phase maps because of their hypointense appearance. The previously acquired CT also validated the calcifications. The CT density of lesions located within the suprasellar region and temporal lobe were 660 and 261 HU, respectively. The diagnosis of craniopharyngioma was histologically established.

Discussion

Susceptibility-weighted imaging is an MRI technique that offers valuable information such as iron imaging in brain, evaluation of stroke, vasculopathies, trauma, epilepsy, and the characterization of the tumors.^5,11

–14 It provides a complementary contrast to T1-, T2-, T2*-, and proton density–weighted imaging. The contrast is based on the magnetic susceptibility differences of the tissues and enhanced by the phase information after it is processed. Phase data processing mainly involves high-pass filtering correction and mask creation to multiply with magnitude images.⁷ Processed phase images constitute the core of contrast enhancement in final susceptibility-weighted imaging reconstructed images. Moreover, processed phase images of gradient echo-based sequences themselves are also excellent tools in differentiating calcification and hemorrhage^2,3 since the 2 pathologies cause different phase shifts owing to their different magnetic susceptibilities. This characteristics of phase images of susceptibility-weighted imaging acquisitions were reported for adult patients.⁴ Gradient echo phase images were compared with quantitative susceptibility maps in characterization of intracranial calcifications and hemorrhages in patients with ages starting from 6 to 75.¹⁵ A recent publication reported that phase images of susceptibility-weighted imaging acquisitions yielded significant phase shift values for calcification and hemorrhage in gliomas and validated the results against histopathologic examination.¹⁰ Herein, we utilized filtered phase maps obtained from susceptibility-weighted imaging sequence to differentiate hemorrhage and calcification in pediatric cases.

Chronic hemorrhage and calcification may both appear hypointense on T2-weighted images. Some imaging characteristics including location, size, and multiplicity of the aforementioned lesions are mostly not sufficient enough to make the differential diagnosis. Therefore, CT is the gold standard for the detection of calcification and often employed to make an accurate diagnosis. The lesion is interpreted as calcification when Hounsfield units (HU) are above a certain threshold (100 HU).^16
–18 However, CT utilizes radiation as an imaging technique and should be avoided particularly in children and fetal imaging whenever possible. We show that phase maps can be an alternative method to CT imaging to differentiate these 2 pathologies when evaluated in conjunction with routine MRI and susceptibility-weighted imaging. One limitation of MRI compared to CT is the longer scan time. The acquisition of susceptibility-weighted imaging takes about 2 minutes 47 seconds. Although short scan times are crucial especially for pediatric acquisition because of patient motion, susceptibility-weighted imaging with scan duration less than 3 minutes may not be that long when considering the aforementioned benefits.¹⁹

Tuberous sclerosis is a phacomatosis that commonly reveals calcified cortical hamartomas and subependymal tubers.²⁰ Calcifications of the subependymal nodules located along the atrium and caudothalamic groove in tuberous sclerosis are pathognomonic. Cortical tubers may also calcify and frequency of the calcification for both lesions increases with the patient age. CT has been the imaging modality of choice to detect those calcified lesions and suggested to be beneficial to define undiagnosed family members with subtle symptoms.^21,22 However, susceptibility-weighted imaging was suggested as an alternative to CT in the course of diagnosis of tuberous sclerosis, as a modality that is as sensitive as CT in detecting the intracranial calcifications.¹⁸ Although the lack of cranial CT imaging of our current tuberous sclerosis cases made it difficult to compare the sensitivity, the calcifications of subependymal and cortical hamartomas could be consistently detected and differentiated using susceptibility-weighted imaging and phase maps (Figure 3, second row).

Sturge-Weber syndrome is a sporadic neurocutaneous disease characterized with facial port wine stain in the trigeminal nerve distribution and ipsilateral choroidal-leptomeningeal angiomas.²³ Susceptibility-weighted imaging is extremely sensitive to enlarged transmedullary veins by detecting deoxygenated blood in without contrast administration, whereas conventional MRI is not sufficient to show structural changes.¹ As seen in Figure 4, calcified leptomeningeal angiomas might easily be manifested in phase maps.

The TORCH group agents including toxoplasmosis, rubella, cytomegalovirus, and herpes simplex virus are the most common cause of intrauterine infections. Intracranial calcifications that involve the basal ganglia, brain stem, and cortical and periventricular areas may be encountered. Because susceptibility-weighted imaging and phase maps have been reported to be more sensitive than conventional sequences, incorporation of susceptibility-weighted imaging to MRI protocol can provide additional information for the diagnosis of intrauterine infection by differentiating between intracranial calcifications and hemorrhage. The periventricular white matter calcifications even in millimeter in size could be detected on phase maps more exquisitely than T1- and T2-weighted sequences in the patient with congenital cytomegalovirus (Figure 3, first row).

The application of susceptibility-weighted imaging in pediatric brain tumors is helpful to differentiate the blood products and the calcifications within the tumors. The reliable differentiation of both entities may facilitate to grade the tumor as well as monitoring the treatment response.⁴ Besides, the estimation of the tumor type often containing calcifications like teratoma, oligodendroglioma, ependymoma, and craniopharyngioma as depicted in the current study is possible using the phase maps (Figure 5).

In some circumstances, phase maps may present some limitations: When the pathology is adjacent to the skull, interpretation may be challenging because of imperfections induced by large magnetic susceptibility difference between the skull and cortex. As seen in Figure 2, the pathology was observed on a T2-weighted image, whereas hemorrhage was not obvious on the phase map, in spite of the brightness observed therein. Another limitation of the phase map is the mixed appearance of opposite signals within the lesion in some situations. This may be caused by aliasing or imperfections in detecting the phase of the magnetic resonance signal. Aliasing results in opposite intensities at the edges and the center of the lesion that is relatively easy to recognize (Figure 1, third row).¹⁸ The use of shorter echo time or double echo can help to minimize aliasing. However, phase detection may also fail when the magnetic resonance signal is low in the hypointense regions. Both hemorrhage and calcification may yield low magnetic resonance signals that make phase detection difficult. In cases with heterogeneous signals, dominant signal pattern is recommended for consideration in the diagnosis⁴ or quantitative susceptibility mapping can be used to correctly deconvolve the local susceptibilities.^15,24

–27 Quantitative susceptibility mapping is a new postprocessing technique that inverts the phase data to magnetic susceptibility maps by accounting the shape and the orientation of the lesion relative to main magnetic field. It is a promising technique that successfully differentiates intracranial calcification and hemorrhage and will be the subject of future work.

This work has some limitations. One is that CT images were only available for 2 cases of calcifications (Figures 4 and 5). The diagnoses were based on conventional magnetic resonance images, diffusion-weighted imaging, susceptibility-weighted imaging, phase images, and clinical data. Because of the lack of complete set of CT images, no sensitivity or specificity analyses are presented. However, we performed a t test on the phase shifts of hemorrhage and calcifications and showed that they appear significantly different (opposite intensity) on the phase images (P < .01).

In conclusion, phase maps can be a useful tool to distinguish hemorrhage and calcification in children when used in conjunction with conventional magnetic resonance and susceptibility-weighted imaging images. They can provide critical information for the diagnosis of calcifications in phacomatosis as tuberous sclerosis and Sturge-Weber syndrome, and pediatric brain tumors. This capacity can obviate the need for confirmatory CT imaging for the diagnosis and thus eliminate unnecessary radiation exposure to the children.

Footnotes

Acknowledgments

The study was performed at Erciyes University Children Hospital.

Author Contributions

MC clinically evaluated the patients. KG, GK, MC, MSD, SBG, and SD collected patient information. GK and SD read the radiological images. KG and GK wrote the paper. KG, GK, MSD, SBG, MC, SD, MB, and AC revised the manuscript.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Ethical Approval

This retrospective study was approved by Erciyes University Institutional Review Board (2014/525).

References

Henkelman

Watts

Kucharczyk

. High signal intensity in MR images of calcified brain tissue. Radiology. 1991;179:199–206.

Yamada

Imakita

Sakuma

Takamiya

. Intracranial calcification on gradient-echo phase image: depiction of diamagnetic susceptibility. Radiology. 1996;198:171–178.

Gupta

Rao

Jain

. Differentiation of calcification from chronic hemorrhage with corrected gradient echo phase imaging. J Comput Assist Tomogr. 2001;25:698–704.

Zhu

Zhan

. Magnetic resonance susceptibility weighted imaging in detecting intracranial calcification and hemorrhage. Chin Med J (Engl). 2008;121:2021–2025.

Mittal

Neelavalli

Haacke

. Susceptibility-weighted imaging: technical aspects and clinical applications, part 2. AJNR Am J Neuroradiol. 2009;30:232–252.

Haacke

Mittal

Neelavalli

Cheng

. Susceptibility-weighted imaging: technical aspects and clinical applications, part 1. AJNR Am J Neuroradiol. 2009;30:19–30.

Haacke

Cheng

Reichenbach

. Susceptibility weighted imaging (SWI). Magn Reson Med. 2004;52:612–618.

Docampo

Gonzalez

Bravo

Sarroca

Morales

Bruno

. Susceptibility-weighted angiography of intracranial blood products and calcifications compared to gradient echo sequence. Neuroradiol J. 2013;26:493–500.

Tsai

Kao

Tsui

Hasso

Greensite

. Susceptibility weighted imaging and cerebrovascular disorders. Neuroradiol J. 2011;24:121–127.

10.

Berberat

Grobholz

Boxheimer

Rogers

Remonda

Roelcke

. Differentiation between calcification and hemorrhage in brain tumors using susceptibility-weighted imaging: a pilot study. AJR Am J Roentgenol. 2014;202:847–850.

11.

Zulfiqar

Dumrongpisutikul

Intrapiromkul

Yousem

. Detection of intratumoral calcification in oligodendrogliomas by susceptibility-weighted MR imaging. AJNR Am J Neuroradiol. 2012;33:858–864.

12.

Thomas

Somasundaram

Thamburaj

. Clinical applications of susceptibility weighted MR imaging of the brain—a pictorial review. Neuroradiology. 2008;50:105–116.

13.

Bai

Wang

Han

. Susceptibility weighted imaging: a new tool in the diagnosis of prostate cancer and detection of prostatic calcification. PLoS One. 2013;8:e53237.

14.

Tong

Ashwal

Obenaus

Nickerson

Kido

Haacke

. Susceptibility-weighted MR imaging: a review of clinical applications in children. AJNR Am J Neuroradiol. 2008;29:9–17.

15.

Chen

Zhu

Kovanlikaya

. Intracranial calcifications and hemorrhages: characterization with quantitative susceptibility mapping. Radiology. 2014;270:496–505.

16.

Kucharczyk

Henkelman

. Visibility of calcium on MR and CT: can MR show calcium that CT cannot? AJNR Am J Neuroradiol. 1994;15:1145–1148.

17.

Tsuruda

Bradley

. MR detection of intracranial calcification: a phantom study. AJNR Am J Neuroradiol. 1987;8:1049–1055.

18.

Mittal

Kish

Haacke

. Identification of calcification with MRI using susceptibility-weighted imaging: a case study. J Magn Reson Imaging. 2009;29:177–182.

19.

Tsai

Shih

Chan

Tsai

Chung

. Practical aspects of shortening acquisition time in brain mr susceptibility-weighted imaging. Neuroradiol J. 2012;25:649–656.

20.

Roach

Sparagana

. Diagnosis of tuberous sclerosis complex. J Child Neurol. 2004;19:643–649.

21.

Altman

Purser

Post

. Tuberous sclerosis: characteristics at CT and MR imaging. Radiology. 1988;167:527–532.

22.

Inoue

Nemoto

Murata

. CT and MR imaging of cerebral tuberous sclerosis. Brain Dev. 1998;20:209–221.

23.

Jagtap

Srinivas

Harsha

Radhakrishnan

. Sturge-Weber syndrome: clinical spectrum, disease course, and outcome of 30 patients. J Child Neurol. 2013;28:725–731.

24.

Deistung

Schweser

Wiestler

. Quantitative susceptibility mapping differentiates between blood depositions and calcifications in patients with glioblastoma. PLoS One. 2013;8:e57924.

25.

de Rochefort

Liu

Kressler

. Quantitative susceptibility map reconstruction from MR phase data using bayesian regularization: validation and application to brain imaging. Magn Reson Med. 2010;63:194–206.

26.

Liu

de Rochefort

. Morphology enabled dipole inversion for quantitative susceptibility mapping using structural consistency between the magnitude image and the susceptibility map. NeuroImage. 2012;59:2560–2568.

27.

Xia

Zheng

Shen

. Quantitative measurements of brain iron deposition in cirrhotic patients using susceptibility mapping. Acta Radiol. 2014.