Subdural hematomas (SDH) are 1 of the 3 types of extra-axial intracranial hemorrhages (along with subarachnoid and epidural hemorrhages) and usually occur as a result of trauma. Deceleration injuries are often the cause of subdural bleeding from rupturing of veins via a shearing mechanism. Other entities, such as child abuse and ventricular decompression, also can result in subdural bleeding, and spontaneous hemorrhages may occur in patients receiving anticoagulants or patients with a coagulopathy condition. Compression of a dural sinus does not directly cause a subdural hematoma, although compression may result in a venous infarction.
Chronic subdural hematoma is more common in elderly individuals because of the age-associated decrease in brain volume and the increase in venous fragility. In one study, the mean age of patients with bilateral CSDH was 77 years, and the mean age with unilateral CSDH was 72 years. Although CSDH is considered to be a somewhat benign disease of the elderly, mortality of up to 13% has been reported. [1, 2, 3]
Overall incidence of bilateral CSDH is reported to be 16-20%. 
Some subdural hematomas are clinically silent, whereas others cause symptoms as a result of mass effect on the adjacent brain. Some hematomas can grow large enough to result in herniation of cerebral tissue. Before computed tomography (CT) scanning and magnetic resonance imaging (MRI) technology, subdural hematomas were diagnosed only on the basis of this mass effect, which was depicted as displacement of the blood vessels on angiograms or as a calcified pituitary gland on skull radiographs. The advent of CT scan and MRI studies has made the diagnosis of even small hemorrhages routine (see the image below). CSH images may appear as hypodense, isodense, hyperdense, or mixed density. [4, 5, 6, 7, 1, 3, 8]
CT scanning is usually the first evaluation in patients with suspected acute subdural hematoma because CT scans depict acute hemorrhage and skull fractures well, they are relatively fast to obtain, and CT scanning is more readily available than MRI. Smaller hemorrhages may be missed on CT scans inn the nonacute setting, MRI is the study of choice because of its high sensitivity and specificity. [9, 10, 11, 12] CT scanning may fail to depict small hemorrhages because of the similarity in attenuation between blood and adjacent bone and because of streak artifacts in the posterior fossa and inferior middle cranial fossa. MRI aids in the detection of small hematomas because of its multiplanar capabilities. [4, 5, 6, 7, 3, 8]
CT scan findings in subdural hematomas depend on the age of the hemorrhage (see the image below). [13, 14, 15] Differentiating subdural from epidural hematomas may be difficult when the hemorrhage is small, because the image of the blood may not demonstrate a typical shape in either condition. Follow-up imaging to ensure that the hematoma is not expanding and to check for an adjacent skull fracture is typical. CT findings may appear as hypodense, isodense, hyperdense, or mixed density, with correlations with vascular endothelial growth factor (VEGF). A study by Weigel et found that mean VEGF concentration were highly correlated with exudation rates; VEGF values were highest in mixed-density hematomas, followed by isodense and hypodense hematomas. 
Small subdural hematomas may not be depicted because the attenuation may be similar to the adjacent inner table of the skull. Viewing the images with a wider window and level (eg, 240 and 80 HU) assists in detection in these cases; however, CT scanning fails to depict a certain number of small hemorrhages. Gentry et al found that only 53% of acute and subacute subdural hematomas were revealed on CT scan studies compared with MRI; however, this study was performed using older CT technology. [16, 17, 7] Measurement of white matter in HU was found to be a helpful predictor of outcome in patients with subdural hematoma with cerebral edema. A cutoff value of 31.5 HU of white matter had an 80% sensitivity and a 99.9% specificity for death. 
In the acute phase, subdural hematomas appear as a crescent-shaped extra-axial collection with increased attenuation that, when large enough, causes effacement of the adjacent sulci and midline shift. The attenuation changes as the hematoma ages (see the images below). [18, 19]
Subacute subdural hematomas may be difficult to detect because they may have isoattenuation compared with adjacent gray matter. Displacement of the gray matter–white matter junction is an important sign that indicates the presence of a space-occupying lesion. Although often administered in the past to help detect displacement of cortical vessels, contrast medium is not necessary with the capabilities of current scanners.
Chronic subdural hematomas (see the image below) have isoattenuation relative to the cerebrospinal fluid (CSF). In rare cases, such hematomas may calcify, resulting in an unusual appearance that can be mistaken for a calcified mass.
Unlike epidural hematomas, subdural hematomas are not restricted by dural tethering at the cranial sutures; they can cross suture lines and continue along the falx and tentorium (see the image below). However, they do not cross the midline because of the meningeal reflections.
When a subdural hematoma is discovered on a CT scan, it is important to check for the presence of other related injuries, such as skull fracture (see the first image below), intraparenchymal contusions, and subarachnoid blood (see the second image below). The presence of adjacent parenchymal injury in patients with a subdural hematoma is the most important factor in predicting their clinical outcome.
Rebleeding into subdural hematomas also may occur and is depicted as a layer of high-attenuation hemorrhage within a lower attenuation hematoma.
In older patients with cerebral atrophy, an appearance of bilateral frontal subdural hygromas may be seen when the patient is in the supine position. However, the lack of mass effect and the presence of general atrophy suggest that this appearance is merely the result of settling of the atrophic brain rather than a pathologic subdural collection. A similar finding can be seen in young children (benign enlargement of the subarachnoid space), which should resolve in the first few years of life (see the image below).
Posttraumatic subdural hygromas can also be confused with chronic subdural hematomas. These develop days or weeks following trauma and result from tears in the arachnoid and resulting leakage of CSF into the subdural space. They are self-limited and usually resolve after several months.
Magnetic Resonance Imaging
MRI is the most sensitive imaging test available for the detection of subdural hematomas. Small subdural hematomas are occasionally difficult to distinguish from epidural hemorrhages.
MRI is more sensitive than CT scanning in the detection of subdural hematomas because the multiplanar and superior tissue differentiation of MRI makes detection easier. In particular, a sensitivity of more than 95% has been described with T2-weighted images of subdural hematomas because of the marked difference in signal intensity between blood products and adjacent structures (see the images below). [4, 5]
The presence of IL-6 and IL-8 on hyperintense T1-weighted images and the presence of beta-trace proteins on hyperintense T2-weighted images appear to be associated with rebleeding and cerebrospinal fluid admixure in chronic SDHs, according to Park et al. 
Preoperative findings on MRI, particularly T1-weighted classification, have been found to be a significant indicator of recurrence of chronic SDH, with T1-iso/hypointensity determined to be a high indicator of risk (18.2% recurrence rate vs 5.2% for other indicators).  Size of the middle meningeal artery (MMA) has been shown to be larger on magnetic resonance angiography (MRA) in patients who developed chronic SDH. 
The shape of the subdural hematoma on axial images is the same crescent-shaped pattern seen on CT scan images. Coronal images are useful in evaluating the extent of subdural hematomas and in detecting temporal and tentorial hemorrhages, 2 aspects that are poorly depicted on CT scans.
In subdural hematomas, the signal depends on the age of the hemorrhage and follows the signal pattern of intraparenchymal hematomas in acute and subacute cases (see the image below). Chronic subdural hematomas, which appear as isoattenuation relative to CSF on CT scans, often demonstrate increased signal intensity on T1-weighted images because of the presence of free methemoglobin, though the intensity decreases over time. Hemosiderin is usually not present and is believed to result from the lack of a dural blood–brain barrier. 
When hemorrhages of differing ages exist within a subdural collection, septae may separate the different blood products (see the first image below). In addition, a blood–fluid level may be seen. When blood products of various ages are depicted on MRIs in a child, particularly when the blood is at multiple sites, child abuse must be suspected (see the second image below). Posterior interhemispheric and tentorial subdural hematomas are also suggestive of child abuse because they are associated with shaken baby syndrome.