Emergency Head Computed Tomography Research Articles

Development and External Validation of a Deep Learning Algorithm to Identify and Localize Subarachnoid Hemorrhage on CT Scans.

In medical imaging, a limited number of trained deep learning algorithms have been externally validated and released publicly. We hypothesized that a deep learning algorithm can be trained to identify and localize subarachnoid hemorrhage (SAH) on head computed tomography (CT) scans and that the trained model performs satisfactorily when tested using external and real-world data. We used noncontrast head CT images of patients admitted to Helsinki University Hospital between 2012 and 2017. We manually segmented (i.e., delineated) SAH on 90 head CT scans and used the segmented CT scans together with 22 negative (no SAH) control CT scans in training an open-source convolutional neural network (U-Net) to identify and localize SAH. We then tested the performance of the trained algorithm by using external data sets (137 SAH and 1,242 control cases) collected in 2 foreign countries and also by creating a data set of consecutive emergency head CT scans (8 SAH and 511 control cases) performed during on-call hours in 5 different domestic hospitals in September 2021. We assessed the algorithm's capability to identify SAH by calculating patient- and slice-level performance metrics, such as sensitivity and specificity. In the external validation set of 1,379 cases, the algorithm identified 136 of 137 SAH cases correctly (sensitivity 99.3% and specificity 63.2%). Of the 49,064 axial head CT slices, the algorithm identified and localized SAH in 1845 of 2,110 slices with SAH (sensitivity 87.4% and specificity 95.3%). Of 519 consecutive emergency head CT scans imaged in September 2021, the algorithm identified all 8 SAH cases correctly (sensitivity 100.0% and specificity 75.3%). The slice-level (27,167 axial slices in total) sensitivity and specificity were 87.3% and 98.8%, respectively, as the algorithm identified and localized SAH in 58 of 77 slices with SAH. The performance of the algorithm can be tested on through a web service. We show that the shared algorithm identifies SAH cases with a high sensitivity and that the slice-level specificity is high. In addition to openly sharing a high-performing deep learning algorithm, our work presents infrequently used approaches in designing, training, testing, and reporting deep learning algorithms developed for medical imaging diagnostics. This study provides Class III evidence that a deep learning algorithm correctly identifies the presence of subarachnoid hemorrhage on CT scan.

Analysis of emergency head computed tomography in critically ill oncological patients.

BackgroundCritically ill cancer patients have an increased risk of developing acute neurological signs. The study objective was to evaluate the use and the usefulness of emergency head computed tomography (EHCT) in this category of patients.Patients and methodsThis retrospective, single-centre, cohort study included patients with EHCT performed during Intensive Care Unit (ICU) admission for a period of three years. Indications, imagistic findings, type of malignancy, and outcome were evaluated to identify diagnostic yield and correlations between abnormal findings on positive scans, malignancy type, and mortality rate.ResultsSixty-four EHCTs were performed in 54 critically ill cancer patients, with 32 scans (50%) showing previously unknown lesions and considered to be positive. The most frequent abnormal findings were ischemic (15 EHCTs, 47%) and haemorrhagic (13 EHCTs, 40%) lesions. Thirty-eight EHCTs (59%) were indicated for altered mental status, with a positivity rate of 50%. Eighteen EHCTs (48%) were performed in hematological malignancy patients: 9 (50%) of which were positive with 8/9 (89%) displaying hemorrhagic lesions. Twenty EHCTs were performed in solid tumour patients, 10 (50%) of which were positive, with 9/10 (90%) displaying ischemic lesions. Out of 54 patients, 30 (55%) died during ICU stay. The mortality rate was higher in patients with hematological malignancies and positive EHCT (78% vs. 58%).ConclusionsDiagnostic yield of EHCT in critically ill cancer patients is much higher than in other categories of ICU patients. We support the systematic use of EHCT in critically ill, mainly hemato-oncological patients with nonspecific neurological dysfunction, as it may lead to early identification of intracranial complications.

Acute cerebral venous thrombosis - still an underdiagnosed pathology in emergency computed tomography of the brain.

PurposeAcute cerebral venous thrombosis (CVT) is a rare condition that can lead to a serious clinical state; thus, prompt diagnosis and treatment are mandatory. Head computed tomography (CT) plays a crucial role in the initial prompt diagnosis in the emergency setting. The aim of the study was to retrospectively analyse emergency head CT studies and the rate of incorrect diagnoses and main sources of pitfalls.Material and methodsRetrospective analysis of 31 emergency CT studies (22 without contrast, 19F/12M, age range: 4-94 years) of patients with confirmed acute CVT.ResultsThrombosed dural sinuses were found in 24/31 (77.4%) cases, thrombosed veins in 7/31 (22.6%) cases, no lesions within vessels in 2/31 (6.5%) cases. Haemorrhagic brain lesions were found in 9/31 (29%) cases, hypodense oedema in 6/31 (19.6%) cases, brain swelling in 1/31 (3.2%) cases, and no parenchymal lesions were revealed in 15/31 (48.4%) cases. Correct diagnosis of CVT was established in 15 cases (48.4%); however, it was incorrect in 16 cases (51.6%). Incorrect cases consist of 4 groups: 1 – with both vascular and parenchymal lesions that were overlooked (50%), 2 – with vascular lesions only, which were either overlooked, misinterpreted, or covered by artefacts (31.3%,), 3 – with parenchymal lesions only, which were misinterpreted (12.5%), and 4 – with no lesions present in the emergency head CT (6.2%).ConclusionsThe high rate of incorrect diagnoses of acute CVT based on emergency head CT requires constant training of radiologists and their close cooperation with clinicians because a delayed diagnosis may be lethal to the patient.

Pattern of Emergent Head Computed Tomography Findings in a Tertiary Care Hospital during off Working Hours: Retrospective Analysis.

Introduction:Emergency head computed tomography (CT) is rising exponentially during off working hours due to evidence-based medicine, patient's expectation and desires, easy availability and apprehension of medico-legal cases, thereby raising health-care cost. There is huge gap in demand and supply of radiologist, especially during off working hours. There is need to know the pattern of emergency head findings.Materials and Methods:A retrospective analysis of all emergent noncontrast CT head during off working hours in the Department of Radiodiagnosis of a Tertiary Care Hospital, Mumbai, India, which were performed from June 2017 to May 2018. CT findings of 308 patients were analyzed.Results:About 63.6% of total head CT showed no significant abnormality. The most common abnormality was intracranial hemorrhage which was just 9.1% followed by acute infarct which was 6.2%. Extradural hemorrhage, subdural hemorrhage, and subarachnoid hemorrhage was only 1% each of total head CT findings. No significant abnormality was detected in 74.65%, 70.21%, 89.13%, 31.37%, 100%, and 69.09% in cases of head injury, seizure, giddiness/dizziness/syncope, cerebrovascular accident, transient ischemic attack, and altered sensorium, respectively.Conclusion:Pattern analysis of emergent head CT reveals that most of the emergent CT head shows no significant abnormality. There is a need for stringent guidelines for emergent head CT, training of emergency physician as well as CT technician for common findings to bridge the radiologist demand-supply gap for providing effective health care in peripheral hospitals.

Use of Head Computed Tomography (CT) in the Pediatric Emergency Department in Evaluation of Children With New-Onset Afebrile Seizure.

Cranial computed tomography (CT) is not recommended in the routine evaluation of children with first afebrile seizure due to its low yield. The objective was to assess the current practice in our pediatric emergency department regarding the use of head CT in children with first afebrile seizure and to identify the factors that lead to ordering a head CT. Medical records of patients between 1 month and 18 years old evaluated at our emergency department for presentation of first afebrile seizure between 2010 and 2014 were retrospectively reviewed. Data extracted include age, gender, seizure type, single or multiple seizures at presentation, seizure duration, predisposing conditions to seizures (ie, history of developmental delay), and whether a head CT was performed. Contingency tables with chi-square analyses were used to determine which variables were associated with increased use of head CT. Of 155 patients (88M/67F) included in the study, 72 (46.5%) underwent head CT and only 3 had clinically significant findings that did not require acute management. There were no differences in CT use by age, sex, seizure type, seizure number, seizure risk factors, or findings on physical examination. Head CT was performed more frequently in cases with seizures ≥5 minutes and unknown seizure duration ( P = .04). Despite existing evidence, the emergent head CT rate was high in our cohort. Children with seizure duration of ≥5 minutes or of unknown duration were more likely to undergo head CT in our emergency department.

Imaging in Neurocritical Care Practice.

The use of neuroimaging in conjunction with serial neurological examinations is a core component of modern neurocritical care practice. Although there is a growing role for other neuromonitoring techniques, the ability to quickly and accurately interpret images in the context of a patient's clinical status arguably remains the indispensable skill for neurocritical care practitioners. Due to its rapid acquisition time and excellent ability to detect intracerebral hemorrhage (ICH), cerebral edema, and signs of elevated intracranial pressure, computed tomography (CT) remains the most useful neuroimaging technique for intensive care unit (ICU) patients. An emergent head CT is obtained to inform most time-sensitive decisions that arise in the neurological ICU (NICU). CT features also figure prominently in prognostic scores for common NICU conditions such as traumatic brain injury (TBI), ICH, and subarachnoid hemorrhage (SAH). Among patients who are sufficiently stable to leave the ICU and lie flat for an extended period, magnetic resonance imaging provides much more detailed, high-contrast images which can aid in the detection of ischemia, diffuse axonal injury, and neuroprognostication. Though primarily used in neurocritical care research, nuclear medicine imaging techniques have some clinical applications, particularly in ancillary testing for brain death. Finally, as in the field of critical care as a whole, formal and point-of-care ultrasound studies are increasingly utilized in the NICU, and are an important tool in the neurointensivist's armamentarium. We review here the common applications of imaging in the neurocritical care setting. As ICU patients are frequently unstable and their risk of clinical decompensation increases substantially during transport away from the ICU, guidelines and recommendations for maximizing patient safety during transport to radiology studies are also explored.

Death due to delayed onset intracranial bleeding in a patient with an acute subdural hematoma

We report a fatal case of acute subdural hematoma (ASDH) involving the delayed onset of subdural bleeding after a head trauma. There were no central nervous symptoms after the head trauma and an initial head computed tomography (CT) scan did not exhibit any abnormal findings. On the second day, the patient suddenly complained of a headache and vomiting, and she lost consciousness. An emergency head CT scan detected a right-sided ASDH and severe cerebral herniation. As the patient was taking triple anticoagulant therapy, surgery could not be performed. She died 2 weeks after the onset of the ASDH. Even in cases of head trauma that do not exhibit abnormal findings on head CT, it is necessary to follow up the patient’s clinical condition. Specifically, controlling coagulability and performing repeated head CT examinations are considered to be beneficial for preventing abundant subdural bleeding and facilitating the early detection of delayed onset ASDH.

Can the Hype of “Endoscope” Become a Reality for Colloid Cyst Surgery?

Early postoperative patient surveillance after removal of intraventricular tumors is often hindered by delayed awakening and prolonged somnolence. The objective of this study was to analyze the incidence of early critical postoperative events after elective craniotomy for intraventricular tumors in adults compared with extraventricular lesions.An observational comparative study was conducted on adult patients who had undergone first-time elective craniotomy between November 2011 and August 2016. Patients were stratified into extraventricular lesions (group 1) and intraventricular tumors (group 2). The rates of late extubation, early postoperative seizures, emergency head computed tomography (CT) scans, and urgent surgical intervention within 48 hours and mortality within 30 days of surgery were analyzed from a prospective database.A total of 977 elective craniotomies were analyzed, including 951 (97.3%) in group 1 and 26 (2.7%) in group 2. Emergency CT scans were ordered significantly more frequently in group 2 (34.6% vs. 8.4%; odds ratio, 5.76; 95% confidence interval [CI], 2.49–13.35; P = 0.0002), and the incidence of urgent surgical intervention was significantly higher in group 2 (11.5% vs. 0.8%; odds ratio, 15.38; 95% CI, 3.83–61.72; P = 0.002). The main reason for urgent surgical intervention in group 2 was acute obstructive hydrocephalus. Overall surgical mortality after 30 days was 0.3% (3 cases in group 1, no cases in group 2).Intraventricular tumors are at significantly higher risk for early emergency head CT and urgent surgical intervention. This patient cohort might benefit from routine intraoperative and early postoperative imaging, as well as intraoperative extraventricular drain placement.

Dose assessment of the patient and the helper in emergency head computed tomography

Abstract Computed Tomography (CT) becomes more and more important and is frequently used in modern diagnostic techniques. CT offers an effective diagnosis on lesion and pathology; however, it also delivers a radiation dose to patients. Besides, in some special emergency cases, the patient may require someone to help him in the examination room to perform the head CT, due to the patient lost intellectual and operational capacity. This study evaluated the delivered radiation dose and the risk of radiation-induced cancer for the patient and for the helper after carrying out the emergent head CT examination. A Rando phantom with thermoluminescent dosimeter (TLD) chips inside relevant organs was used to simulate the patient during head CT examination. An effective dose of 2.06 ± 0.16 mSv and 1.46 ± 0.07 mSv without and with wearing the lead apron were found respectively, considering ICRP 60 recommendation. ICRP 103 recommendations the readings would have been 1.29 ± 0.15 mSv and 0.71 ± 0.04 mSv. The effective dose determined from the dose-length product (DLP) method (2.19 mSv) was similar to the value (2.06 ± 0.16 mSv) estimated by TLD method considering the ICRP 60 recommendation (without lead apron). Assuming a 5% total risk for fatal cancer per Sv in the general population, risk of radiation-induced cancer for patients were 1.03 × 10 −2 % (without lead apron) and 7.80 × 10 −3 % (with lead apron) for ICRP 60, and were 6.45 × 10 −3 % (without lead apron) and 3.55 × 10 −3 % (with lead apron) for ICRP 103. The dose received by the helper was assessed by wearing a personal badge. The helper during emergent head CT examination may receive a personal dose equivalent ( H p (10)) of 19.36 ± 5.89 μSv and 138.81 ± 101.28 μSv with and without lead apron, respectively, at distance of 0.3–1 m from the center of CT scanner. Based on the observed dose reduction of a factor of 7.17 we recommend that helpers wear lead apron in the CT examination room.

頭部ＣＴにおけるヘリカルスキャンとノンヘリカルスキャンのモーションアーチファクトに関する検討

For head computed tomography (CT), non-helical scanning has been recommended even in the widely used multi-slice CT (MSCT). Also, an acute stroke imaging standardization group has recommended the non-helical mode in Japan. However, no detailed comparison has been reported for current MSCT with more than 16 slices. In this study, we compared the non-helical and helical modes for head CT, focusing on temporal resolution and motion artifacts. The temporal resolution was evaluated by using temporal sensitivity profiles (TSPs) measured using a temporal impulse method. In both modes, the TSPs and temporal modulation transfer factors (MTFs) were measured for various pitch factors using 64-slice CT (Aquilion 64, Toshiba). Two motion phantoms were scanned to evaluate motion artifacts, and then quantitative analyses for motion artifacts and helical artifacts were performed by measuring multiple regions of interest (ROIs) in the phantom images. In addition, the rates of artifact occurrence for retrospective clinical cases were compared. The temporal resolution increased as the pitch factor was increased. Remarkable streak artifacts appeared in the non-helical images of the motion phantom, in spite of the equivalent effective temporal resolution. In clinical analysis, results consistent with the phantom studies were shown. These results indicated that the low pitch helical mode would be effective for emergency head CT with patient movement.

Keeping Patients Safe During Intrahospital Transport

Keeping Patients Safe During Intrahospital Transport

Emergency head CT scans: can neurosurgical registrars be relied upon to interpret them?

Neurosurgical registrars are frequently called upon by A&E staff and physicians to interpret emergency head CT (computed tomography) scans out of hours. This appears to reflect the reduced threshold for scanning patients and the nonavailability of a radiologist to report these scans. This study was undertaken to assess the safety of such practices. Five neurosurgical registrars, blinded to each other and to the radiology reports, interpreted 50 consecutive emergency head CT scans (both trauma and nontrauma) from the hospital's imaging system as a pilot study. These were initially graded as normal or abnormal. Abnormal scans were assessed for the presence of an intracranial bleed, pneumocephalus, skull fractures, cerebral contusions, mass effect, midline shift, ischaemia or hydrocephalus. The agreement of the observers' recordings with the report issued or approved by a consultant radiologist was evaluated using SPSS™ Version 13.0. Four of the five registrars assessed a further 150 scans in a similar manner to complete the study. There was a good general agreement between the formal reports and the neurosurgical registrars' identification of normal scans (average Kappa 0.79). The radiology reports and the registrars also agreed well on the presence or absence of intracranial blood, contusions and pneumocephalus (Kappa value > 0.70). The agreement was poorer for ischaemia, mass lesions (other than intracranial haematomas), grey white differentiation, evidence of raised intracranial pressure and midline shift (Kappa < 0.5). Neurosurgical registrars compared well with radiologists when it came to assessing emergency head CT scans as normal or detecting a surgical lesion. The agreement was poorer on subtle abnormalities. The practice of neurosurgical registrars informally ‘reporting’ on emergency head CT scans cannot be recommended.

Artefact misinterpretation on CT images of the head

In most radiology departments, axial computed tomography (CT) is the standard acquisition mode for routine head examinations. Historically, this practice involved the use of singledetector row machines, and has persisted even with the advent of helical and multidetector row CT systems. Artefacts seen using multidetector row CT systems have been shown to be substantially fewer than a single-detector row machines, and those that were identified were associated with helical scanning. However, artefacts can be seen if the axial scanning mode is used on a multidetector row CT system when one of the detector rows is faulty. It is important for radiologists, CT radiographers, and physicists to be aware of these as they can mimic disease and cause misinterpretation. We report the cases of three patients who underwent emergency head CT examinations that showed artefacts misinterpreted initially as pathological lesions by the radiologist, necessitating further investigations. Although there has been a report of an unknown substance on the collimator assembly causing hypoattenuating artefacts, prior to the present series, there have been no reports of contrast medium spill on the detector and optic assembly causing artefacts.

Emergency Department Use of Computed Tomography in Children With Epilepsy and Breakthrough Seizure Activity

Children with epilepsy are frequently evaluated in the emergency department for breakthrough seizure activity. This population is at risk for undergoing repeated computed tomography (CT) scans. The long-term health risks associated with CT scans has been the subject of multiple articles in recent years. In 2002, the National Cancer Institute released a statement that focused on reducing long-term health risks by limiting children's exposure to CT radiation. A retrospective study that included 124 children (ages 2 months-16 years) who presented to our emergency department with epilepsy and breakthrough seizure activity was performed to determine whether the National Cancer Institute statement made an impact on the number of head CT scans ordered. Physician's compliance with the American Academy of Neurology guidelines for emergent head CT scans in patients with known seizure disorder and the diagnostic yield of the head CT scans was also assessed. There was a significant increase (P = .016) in the number of CT scans performed from the years 2000-2001 to the years 2003-2004. Of the 21 children who had CT scans, 15 (71%) children did not meet American Academy of Neurology guidelines for an emergent CT scan. None of the children had an acute finding on CT, and all were discharged from the emergency department. This finding suggests that the yield of emergent CT scans in children with epilepsy and breakthrough seizure activity is low. This study demonstrates the increasing utilization of CT scans in this population at our institution and that compliance with American Academy of Neurology guidelines is not optimal.

Head computed tomography guidelines are being followed

Several guidelines have been developed to direct the ordering of head computed tomography (CT) for patients, but most are clinical presentation-specific. Recently, an integrated guideline for ordering emergent head CT for patients who present to the ED of Western Hospital, Footscray, Victoria, Australia, was developed in response to concerns raised regarding perceived over-utilization of head CT for ED patients. Our aim was to determine compliance with the guideline. This was an explicit retrospective medical record review of patients who presented to the ED of Western Hospital between 1/04/2004 and 17/6/2004 and had a head CT as part of their assessment. Clinical information for these cases was compared with guideline recommendations. Data are described by descriptive statistics. Of the 231 cases that were included in the study, 65 (28.1%, 95% confidence interval 23-35%) had abnormal CT findings. Guidelines were adhered to in 217 (93.9%, 95% confidence interval 91-97%) cases. For the cases where the guidelines were not adhered to (14; [6.1%]), there was only one abnormal scan the clinical significance of which is not clear. The study found that compliance with head CT guideline was high. This suggests that the guideline is both clinically relevant and supported by ED doctors or conversely that the guideline is concordant with existing ordering practices of the ED.

Clinical predictors of computed tomographic abnormalities following pediatric traumatic brain injury.

Children commonly seek attention in emergency departments following head injury. Head computed tomography (CT) is often used to decide subsequent disposition. Clinical criteria predicting CT abnormalities would allow effective and timely treatment and minimize unnecessary procedures depleting overburdened medical resources. We prospectively compared presenting clinical features with subsequent emergent head CT in 300 children less than 19 years old over a nine-month period. The disposition of patients following imaging was also recorded. Only suspected abuse was more than 50% positively predictive in children below age two and those above age two. Two signs were more than 67% positively predictive in both age groups: focal motor deficit and pupillary asymmetry. Patients with abnormal CTs were the only children to undergo emergent neurosurgery (30%) and were nearly five times as likely to be intensively monitored. Children with normal CTs were nearly five times as likely to be observed in a routine department or at home. We conclude that no single clinical feature can predict with certainty an abnormality on immediate head CT. However, children suspected of being abused, and those with focal motor deficits or pupillary asymmetry, should be imaged. Finally, emergent CT when judiciously ordered likely reduces unforeseen morbidity and minimizes costly intensive care observation.