Acute spontaneous spinal epidural haematoma in a child

Spontaneous spinal epidural haematomas rarely occur. Patients tend to be in their sixties or seventies. Acute spontaneous spinal epidural haematomas in children without a predisposition for bleeding disorders, trauma, vascular malformations or anticoagulant therapy have seldom been described. We present a case of a 4-year-old girl with a spontaneous cervical epidural haematoma diagnosed with MR. Received: 20 January 2000; Revised: 18 April 2000; Accepted: 19 April 2000     

The role of face-to-face consultation in emergency radiology

To assess the impact of face-to-face exchange of clinical information between the radiologist and the emergency physician, a study was made of all plain films and computed tomograms (CTs) ordered in the Ohio State University Emergency Department between 11 am and 11 pm over a 4-week period. For each case, interpretations at three different times were recorded: (1) initial interpretation upon reception of the images, (2) interpretation after face-to-face discussion of clinical information with the emergency resident or attending physician, and (3) final interpretation after conferring with the attending radiologist. If any of the three interpretations differed in a given case, the case was flagged for review by a third attending radiologist to determine the clinical significance of the discrepancy. Cases which deviated from the standard model as described were excluded from the study. In total, 519 cases were ordered during the study period. Of which 231 (44.5 %) occurred in the standard order as described in the Methods section. Of these 231 cases, changes in diagnosis were observed in 31 (13 %) cases. In 12 (5.2 %) cases, the changes were judged to be clinically significant. In 19 (6 %) cases, the changes were judged to be clinically insignificant. The remaining 288 (55.5 %) cases diverged from the standard model described in our Methods section and were not included in our study. Of these, 93/519 (18 %) cases had only one interpretation because the resident radiologist, the emergency physician (resident and/or attending), and the attending radiologist were present for a single, simultaneous discussion. 77 (15 %) cases, there were only two interpretations because the initial interpretation occurred in the presence of both the resident radiologist and the emergency physician. In 69 (13 %) cases, there were two interpretations because the initial interpretation occurred in the presence of both the resident radiologist and the emergency physician. In 49 (9 %) cases, no face-to-face exchange between the resident radiologist and the emergency physician was recorded. Our data suggest that teleradiology's lack of face-to-face communication between the radiologist and the emergency physician may have an impact on the radiologic diagnosis, but only infrequently. Information contained in the images and the limited clinical history on the consult sheet appear to be sufficient for most purposes of radiologic interpretation.     

Delayed epidural hematoma after evacuation of a subdural hematoma

CT easily diagnoses epidural hematomas. The appearance of an epidural hematoma may be delayed by compression from a contralateral lesion. The possibility of a delayed epidural hematoma should be kept in mind in the presence of additional findings such as a skull fracture contralateral to the original lesion or decompression of the epidural hematoma into either the subgaleal or the subarachnoid space. We present a case in which an epidural hematoma declared itself after the evacuation of a contralateral subdural hematoma.     

Infected hepatic infarction: an unusual cause of portal venous gas

Portal venous gas is a radiologic finding that usually is due to extrahepatic abdominal processes, most commonly ischemic bowel. In this case report, we present a case of hepatic infarction and fulminant hepatic infection that resulted in extensive portal venous gas.     

Resident vs. attending interpretation of on-call studies: clinical impact

Purpose: This study was performed to determine whether significant changes to patient treatment plan or outcome result from discrepancies between on-call radiology residents and follow-up attending radiologists in their interpretation of examinations. Methods: For 70 days we recorded on-call radiology residents' readings of all computed tomography and ultrasound examinations performed in our institution and the follow-up attending radiologists' readings of these same examinations. A chart review was performed to determine whether interpretation discrepancies changed the treatment plan and clinical outcome. Results: Eight-hundred thirty-four examinations met the study guidelines. The overall discrepancy rate was 5.16 %. Of these discrepancies, 6.98 % affected the treatment plan (0.36 % of all 834 studies) and none affected the clinical outcome. Conclusion: Where there is a discrepancy between interpretation of computed tomography and ultrasound after hours by on-call radiology residents and follow-up readings by attending radiologists, this discrepancy has no significant effect on the immediate or long-term care of patients.     

Contrast enhancement pattern and frequency of previously unoperated lumbar discs on MRI

Enhancement frequency and patterns (linear intradiscal, nodular intradiscal, and peridiscal) of the 210 previously unoperated lumbar discs were evaluated using contrast-enhanced lumbar MRI. They also were compared with morphologic abnormalities (normal, bulging, protruded, and extruded) and signal alteration (high signal on T2-weighted images) of the disc. Image interpretation was decided by the consensus of two musculoskeletal radiologists. Enhancement was observed in 69 (32.6%) discs. The enhancement patterns were either intradiscal, linear of (79.7%) or nodular (10.2%), peridiscal (7.2%), or combined (2.9%). Thirty-five (77.8%) of the 45 herniated discs, 35 (77.8%) were enhanced, whereas 34 (20.6%) of 165 normal-appearing discs were enhanced. Enhancement was observed more frequently in discs with high signal intensity zones (P < 0.05). In the six patients with enhanced discs, surgery revealed granulation tissue in three extruded discs, two protruded discs, and one bulging disc. The authors believe that contrast enhancements are frequent in herniated discs and are not infrequent in normal-appearing discs. The understanding of various enhancement patterns may help in the interpretation of lumbar spine MRI.     

Swan-Ganz catheter-induced pulmonary artery pseudoaneurysm: review of radiological diagnosis and management

Iatrogenic pulmonary pseudoaneurysm is a well-known but uncommon complication related to Swan-Ganz catheters. The radiologist can play an important role in the diagnosis, prevention, and treatment of this potentially lethal complication. The authors present a case demonstrating the typical clinical and radiological presentation, and describe the role of interventional radiologic management.     

Spontaneous thrombosis of a vein of Galen aneurysmal malformation: possible effects of contrast media

A 5-year-old boy with macrocephaly and mental retardation was referred for radiologic evaluation. After cranial CT and MR imaging, the diagnosis of mural type vein of Galen aneurysmal malformation was established by angiography. Two weeks later, preembolization angiography revealed complete thrombosis of the malformation. Although it is a very rare event, vein of Galen aneurysmal malformation may spontaneously thrombose following diagnostic angiography. Possible effects of contrast media on thrombosis were discussed. Received: 27 April 1999; Revised: 29 September 1999; Accepted: 28 January 2000     

Imaging of the congenitally absent cervical pedicle: avoidance of misdiagnosis in a setting of trauma

An enlarged neuroforamen may be due to a congenitally absent cervical pedicle. We report a case initially misdiagnosed as a cervical fracture. CT and MRI are recommended for correct diagnosis and prevention of unwarranted treatment.     

Rotational injuries of the cervical spine

Injuries which are caused in part by a rotational mechanism include atlantoaxial rotatory fixation (AARF), traumatic isolation of the articular pillar (TIAP), and unilateral interfacetal dislocation (UID). The anatomy, classification, radiographic findings, and frequency of neurological signs are reviewed. Key points include: AARF: (1) Mild injury is not diagnosed radiographically since the appearance is the same as for physiologic rotation. More significant injuries demonstrate widening of the atlanto-dens interval ( > 3 mm in adults) and/or abnormal rotation ( > 45 °) of C1 on C2. (2) AARF refers to an inability to return the head to a neutral position. This may be due to trauma resulting in subluxation or dislocation or may be due to muscle spasm in torticollis. TIAP: (1) Radiographic diagnosis depends on visualizing ipsilateral pedicle and lamina fractures. (2) Rotation of an articular mass from its normal orientation on radiographs implies the presence of this injury. However, rotation of the articular mass is not always present, and the ipsilateral pedicle and lamina fractures must be seen on radiographs or CT. UID: (1) Less rotational discrepancy than expected can occur in two circumstances: articular mass fracture with dislocation or significant subluxation on the contralateral side. (2) Less anterolisthesis than expected can occur if there is articular mass fracture with dislocation.     

Thoracic computed tomography in victims of blunt trauma: changes in utilization after the introduction of helical technology

Purpose: In blunt trauma, CT of the thorax is often performed because of the limitations of plain film chest radiography in accurately depicting injuries to the thoracic aorta. The advent of helical CT has allowed the entire chest to be imaged more expeditiously. We will evaluate the extent to which this ability to rapidly obtain additional images of the thorax has increased the utilization of chest CT in blunt trauma. Methods: The radiologic records of all victims of blunt trauma who underwent total body CT (TBCT), which included abdomen, pelvis, and chest images, in the 10 months before and 10 months after our institution acquired helical CT (HCT), were retrospectively evaluated. Results: In the 10 months before we obtained HCT, 520 conventional CT of the abdomen and pelvis were completed, of which only 11 (2.1 %) included images of the full thorax. After HCT was instituted, 684 abdomen and pelvis CT were performed, of which 59 (8.6 %) were ordered with the entire chest. Of the 11 conventional TBCT, 6 (55 %) showed findings in the chest; in 2 of these cases (18 %) the abnormalities were confined exclusively to the thorax. Similarly, of the 59 helical TBCT, 31 (53 %) depicted traumatic abnormalities in the thorax, and in 23 of these cases (39 %) injuries were only within the chest. In 4 of the latter the injuries were mediastinal hematomas, 1 of which (25 %) was confirmed angiographically to represent an aortic laceration. Conclusion: The speed and ease with which additional images can be obtained by HCT has greatly increased utilization of cross-sectional imaging of the thorax after blunt trauma (approximately four-fold). Over half of these depict traumatic injuries.     

CT of untreated occipitoatlantal rotatory subluxation (OARS)

We report the first case of occipitoatlantal rotatory subluxation (OARS) in which computed tomography (CT) images at initial presentation and at follow-up when still untreated are included. In addition, we report the first images of traumatic OARS to include 3-D CT reformations. The fact that the diagnosis was missed initially underscores the need for familiarity with the radiologic findings. The clinical and radiologic imaging including 3-D CT are discussed.     

CT diagnosis of strangulation in patients with small-bowel obstruction: current status and future direction

The pathophysiology of strangulating small bowel obstruction and the limitations of known CT criteria of strangulation are reviewed. The concept of coronal mapping of the course of the superior mesenteric vessels based on axial abdominal CT in order to aid the radiologic diagnosis of intestinal strangulation is introduced.     

The role of spiral CT using the pulmonary embolus protocol: a comparison of emergency department and hospitalized populations

Purpose: To compare the incidence of CT scan-detected pulmonary embolism, the CT scan-detected alternative findings (ancillary findings and alternative radiologic diagnoses), and the patient characteristics in emergency department and in hospitalized patients. Materials and methods: 81 spiral CT scans in 79 consecutive adult emergency department patients were retrospectively identified by computer search. During the same 9-month interval, 131 consecutive adult in-patients were similarly identified. The in-house and emergency department populations were compared in respect of incidence of pulmonary embolism and alternative findings using χ ² analysis. Results: 81 emergency department scans and 135 in-patient scans were evaluated. Of these, 22 emergency department scans (27.1 %) and 37 in-patient scans (27.4 %) were positive for pulmonary embolism. Of the scans that were negative for a pulmonary embolism, 45 emergency department scans (76.2 %) and 88 in-patient scans (89.8 %) were positive for alternative findings. These included atelectasis, adenopathy, metastatic disease, and pneumomediastinum. Alternative radiologic diagnoses on scans negative for pulmonary embolism occurred in 8/59 emergency department patients (13.5 %) and in 17/98 in-patients (17.3 %). These included pneumonia, tumor, mucus plugging, pericardial effusions, and thoracic aortic dissection. Conclusions: No significant difference exists between the emergency department and in-patient populations for the incidence of pulmonary embolism on spiral CT scans. Although a significant difference did exist between these populations in respect to alternative findings, the high percentage of alternative findings in both populations shows a possible advantage of CT over the traditional test, the ventilation-perfusion scan. Therefore, spiral CT should not be limited to the in-patient population as the first-line imaging modality for the diagnostic evaluation of pulmonary embolism.     

Significance of backboard artifacts on portable trauma series chest radiographs

Purpose: To evaluate the degree to which the artifacts created by the radio-opaque components of a backboard obscure the findings on the initial trauma series chest radiograph (BBCXR). Methods: We reviewed 40 consecutive trauma victims admitted to our level I trauma center over a 3-month period who underwent a follow-up portable supine chest radiograph (PCXR) off the backboard no longer than 1 h after the initial trauma BBCXR. The original interpretation of the BBCXR was compared to the reading of the PCXR as well as to a retrospective analysis of the BBCXR performed by two radiologists blinded as to the findings on the original report. Results: The initial interpretation of the BBCXR failed to identify abnormalities reported in the follow-up PCXR of 10 individuals (25 %), most frequently bone fractures, misplaced hardware, and pleural effusions. Of these missed abnormalities, 80 % were identified retrospectively. Conversely, the initial reading of the BBCXR described a widened mediastinum in five cases, right upper lobe opacities in two, and cardiomegaly in one. Of these eight reported findings, only five (63 %) – the cardiomegaly and four of the mediastinal abnormalities – were also reported upon retrospective analysis, while the PCXR taken with the patient off the backboard confirmed widened mediastinum in only three cases. Conclusions: The initial, often hurried reading of a portable chest radiograph taken on a backboard as part of a trauma series often misses significant pathology, most of which can be detected upon more thorough examination of the original film. The backboard also tends to magnify mediastinal structures more than routine PCXR. Therefore, we advocate a more cautious analysis of the original BBCXR and routine utilization of a follow-up PCXR to confirm mediastinal abnormalities.     

Chronic nontraumatic spinal epidural hematoma of the lumbar spine: MRI diagnosis

An uncommon case of chronic nontraumatic spinal epidural hematoma of the lumbar spine in a 75-year-old woman is reported. The patient presented with a 7-month history of low back pain and bilateral sciatica. Magnetic resonance imaging enabled a correct preoperative diagnosis revealing a nodular, well-circumscribed epidural mass with peripheral enhancement and signal intensities consistent with chronic hematoma, which extended from L2 to L3. Laminectomy of L2–L3 was performed and the hematoma was totally resected. Histological examination of the surgical specimen demonstrated a chronic encapsulated hematoma. No evidence of vascular malformation was found. The patient recovered fully after surgical treatment. Received: 23 August 1999; Revised: 13 January 2000; Accepted: 4 February 2000     

Vertebral sarcoidosis: imaging findings

Sarcoidosis is a multisystemic disease of unknown aetiology characterised by noncaseating granulomatous inflammation with varying presentation and prognosis. Osseous disease reported in 1–13 % of cases commonly involves hands and feet; however, vertebral sarcoidosis is rare. This report describes the radiologic, CT, MRI and radionuclide imaging findings of vertebral involvement of a case with sarcoidosis. Received: 28 January 1999; Revised: 14 April 1999; Accepted: 25 April 1999     

CT diagnosis of cardiac and pericardial injuries

The purpose of this retrospective study was to determine the CT findings diagnostic of cardiac and pericardial injury, including signs of pericardial tamponade, in patients suffering from blunt and penetrating trauma. A search of the CT radiology database at a level I trauma center was performed to identify cases in which injury to the heart or pericardium was diagnosed, as well as to identify cases of pericardial tamponade. All cases were reviewed to ascertain the specific CT findings, and medical records were reviewed to assess the influence of CT findings on management and to assess for clinical evidence of pericardial tamponade. Eighteen patients had direct CT evidence of cardiac or pericardial injury, including nine cases of pneumopericardium, eight cases of hemopericardium, and one case of intrapericardial gastric herniation. Four of these patients were found to have direct cardiac injuries. Three additional cases with CT evidence of pericardial tamponade were identified, two secondary to cardiac compression by an anterior mediastinal hematoma and one following repair of left ventricular rupture. Of 11 patients with CT evidence of tamponade, only three were suspected clinically. Cardiac and pericardial injuries are usually diagnosed surgically and are often clinically unsuspected, particularly in blunt trauma. As CT is increasingly utilized as a general screening test for thoracic/abdominal trauma, these injuries may be first suspected on the basis of CT findings, and knowledge of the CT findings of cardiac injury or tamponade is crucial.     

CT-guided epidural/perineural injections in painful disorders of the lumbar spine: Short- and extended-term results

Purpose: Evaluation of short- and extended-term results of repeated epidural/perineural injections (EDT/PRT) of corticoids in painful afflictions of the lumbar spine. Methods: Thirty-two patients who had persistent radicular or low back pain for more than 6 weeks were treated with CT-guided injection therapy. By EDT/PRT, 40 mg of triamcinolonacetonid was injected either periradicularly or by a direct intraspinal epidural method at intervals of 3 weeks. Altogether, 140 EDT/PRT were performed in 32 patients (mean 4.4, range 2–8). In nine patients partial facet joint denervation with 1–2 ml of 50% alcohol solution was combined with EDT/PRT to reduce low back pain. Before and after treatment and at follow-up (mean 9.6 months), treatment success was evaluated on a visual analog scale and by physical examination (good = >50% improvement, moderate = 20%–50%, no improvement = <20%). Results: Short-term (end of therapy) good or moderate improvement was achieved in 91% of patients, extended-term (mean 9.6 months) in 56%. Regarding certain subgroups, those with disc herniations of the lumbar spine showed a better outcome with good or moderate improvement in 95% short-term and 69% extended-term than those with spinal stenosis who had 72% short-term and 28% long-term. Conclusion: Results indicate that CT-guided EDT/PRT in combination with partial facet joint denervation is a safe and effective outpatient treatment.     

Can chest and abdominal trauma CT eliminate the need for plain films of the spine? – Experience with 329 multiple trauma patients

Purpose: To compare the accuracy of spine plain films with chest and abdominal trauma CT in detection of spine fractures. Methods: The study prospectively enrolled 329 multiple trauma patients. Of these, 38 patients had both chest CT for trauma and thoracic spine plain films, and 87 patients had both abdominal CT for trauma and lumbar spine plain films. Results: Of the fractures visible at either chest trauma CT or thoracic spine plain film examination, all were diagnosed on CT and 62 % on plain films. Of fractures visible at either abdominal trauma CT or lumbar spine plain films, 94 % were diagnosed on CT and 67 % on plain films. The one false negative CT involved an articular process fracture, which was visible but not mentioned, in a patient with a sacral fracture. Conclusion: Evaluation of the digital scout images and bone windows when a patient has chest and abdominal trauma CT appears to be as accurate as thoracic and lumbar spine plain films in the evaluation of spinal trauma.