Some bleeding is undetectable by CT scans

X-ray, CT or MRI?

People fall or hit their heads. This is so common that there should be no ambiguity in the outpatient clinic or practice as to whether and how diagnostic measures should be initiated in the event of such an event. Unfortunately, everyday life in German clinics looks different. Even in the vast majority of specialized hospital departments, there are no clear regulations as to the severity of the symptoms and when not to be diagnosed. If one looks at all of the patients with traumatic brain injury (TBI), it is not even clear which type of diagnostic imaging should be used for this. This article is intended to provide more clarity here.

When we discuss traumatic brain injuries with our students in the lecture, they consistently report that they have got to know the practice during clinical traineeships in emergency hospital departments, that for mild TBI an X-ray of the skull overview in two planes and an X-ray of the cervical spine is carried out in two planes with target recording. When discussing the question of what TBI is really about, even the students realize that the diagnosis of TBI actually depends on brain damage, which naturally cannot be shown on a conventional X-ray.

X-ray not indicated

It follows from this that the production of a conventional X-ray in TBI is an unjustified indication for the use of X-rays in medicine. In addition, the conventional x-ray of the skull is a very difficult image to interpret, even for experts, as numerous anatomical structures overlap and thus around 30% of the fractures are overlooked.

The sensitivity of the diagnosis from this recording is therefore low. If a fracture is found on an X-ray overview, a computed tomography (CT) scan of the skull must be carried out in order to rule out brain damage. If a patient has neurological abnormalities and no fracture is found on the conventional X-ray, a CT must also be performed.

CT also necessary in children

In the 2003 amendment to the X-ray Ordinance, the use of X-rays in children has been restricted by using the method that applies the lowest X-ray radiation. In imaging diagnostics, this is ultrasound and magnetic resonance tomography (MRT). This statement is confirmed by the fact that recent publications have shown that CT examinations in children cause significantly more tumor diseases. As neuroradiologists and paediatricians, we take this instruction very seriously, as we generally use ultrasound and MRI in children, for example for headache clarification, for checking hydrocephalus, for demyelinating diseases and for tumors of the CNS.

With SHT, however, these methods have their limits. Even small epidural hematomas must be safely ruled out here, which is not always possible with ultrasound. Even a non-traumatized child under the age of six cannot undergo an MRI scan without sedation. However, the sedation makes it difficult to monitor the children later. In addition, the MRI examination takes significantly more time than the CT examination, if only because of the preparation of the children. When using the CT, the ambulance service can enter the CT room with the ambulance stretcher, the child can be repositioned immediately and two to three minutes later the decision on further therapy can be made based on the image information. This speed is, for example, brain- and life-saving in the case of the space-consuming epidural hematoma. This is why CT is the imaging method of choice for acute diagnosis of TBI in children with TBI.

Use the Glasgow Coma Score as an indication

Ultimately, the question of how (it has to be CT) is no longer to be answered in acute diagnostics, but only the question of the indication. Unfortunately, that's not that easy. There are numerous recommendations here, but they are not that easy to implement in practice. Who decides, for example, whether it was a high-speed accident? Is that a fall from 2 m or just 1.50 m? Does the car have to be 50 km / h or 70 km / h? A classification that is recommended by Rieger et al. was published in "Der Radiologe" 2002. This classification is based on the GCS (Glascow Coma Score), which is used consistently in emergency medicine. Although the GCS has weaknesses in neurological symptoms, it is always collected by the emergency doctors in an emergency and is used in everyday clinical practice. Rieger et al. do not recommend performing imaging tests if the GCS is 15, with the exception of two collectives: patients over 65 years of age and patients on anticoagulant therapy or on antiplatelet drugs. No diagnostic imaging is performed for patients with a GCS of 14 who improve to a GCS of 15 within two hours.

CT of the skull is indicated in all patients with a GCS of 13 and less. Of course, this classification cannot be used consistently (for example if a fracture is felt or if the patient was unconscious at the scene of the accident for a long time), but it is a relatively simple guide that can be used in everyday clinical practice and when making acute decisions .

The problem of timing

In addition to the very easy to diagnose epi- and subdural hematomas (Table 1), neuroradiologists sometimes have problems with very recent contusion bleeding in acute diagnosis. Typically if the TBI was only a very short time ago (patient had an accident in front of the hospital) and the CT is made a few minutes after the trauma.

To put it simply, fresh blood on the CT is bright (hyperdense), but very fresh blood still has the density of brain tissue (isodense) and can therefore not be recognized (especially in places where partial volume effects occur, such as frontobasal). So if the CT does not diagnose any bleeding and the patient shows neuropsychological or focal neurological symptoms, an MRI examination should be added in the further course (in the days after).

Neuroradiologists have been using T2 * -weighted sequences for years. Recently, however, SWI sequences (susceptibility weighted images) have become established for this question, as they are even more sensitive than T2 * -weighted sequences. With SWI sequences, diffuse axonal injuries can also be detected, the diagnosis of which plays an important role in the rehabilitation process. However, since only about 30% of axonal shear injuries have a hemorrhagic component, the SWI sequence must be combined with the diffusion-weighted sequence (DWI) in order to be able to diagnose non-hemorrhagic axonal injuries as well.

Indications for cervical spine imaging

When should imaging diagnostics of the cervical spine be performed in TBI? This, too, has been clearer for a long time. In 2002, a meta-analysis was published in Neurosurgery, which showed that in conscious patients with isolated TBI and no clinical complaints of the cervical spine (Cervical spine), no findings requiring therapy such as fractures or dislocation were recorded during imaging diagnostics.

If the patient has cervical spine complaints, a conventional image in two planes and a dens target image can be made. For greater security, they can be combined with oblique shots. If there is an urgent clinical suspicion of a fracture, a CT of the cervical spine from occiput to BWK1 with coronary and sagittal reconstructions should be performed. If the patient has a neurological deficit that could be related to the cervical trauma, an MRI of the cervical spine is also indicated in the acute phase.


If imaging diagnostics is indicated in the acute situation of a traumatic brain injury, a computed tomography of the skull should be performed. Conventional x-rays of the skull are obsolete and contraindicated. Magnetic resonance tomography is not associated with any x-ray exposure, but it takes longer and offers little additional information in acute diagnosis. In the further course, however, it has advantages in the sensitivity of the diagnosis of brain contusions and diffuse axonal injuries.

Rieger J, Linsenmaier U, Pfeifer K, Reiser M (2002) Radiological diagnosis of traumatic brain injury. The Radiologist; 42: 547-555.
Guidelines for Management of acute cervical spinal injuries (2002) Neurosurgery; 50 Supplement: 30-35
Parizel P, Makkat S, Van Miert E, Goethem J, Hauwe L, Schepper A (2001) Intracranial hemorrhage. European Radiology; 11: 1770-1783
Kidwell C. (2004) Comparison of MRI and CT for detection of acute intracerebral hemorrhage. JAMA; 292: 1823-1830
Soman S, Hoidsworth S, Barnes P, Rosenberg J, Andre J, Bammer R, Yeom K (2013) Improved T2 * imaging without increase in scan time: SWI processing of 2D gradient echo. AJNR; 34: 2092-2097

Institute for Diagnostic and Interventional Neuroradiology

Conflicts of Interest: The author has not declared any.