Cerebral perfusion changes in traumatic diffuse brain injury; IMP SPECT studies

Generative adversarial network-created brain SPECTs of cerebral ischemia are indistinguishable to scans from real patients

Deep convolutional generative adversarial networks (GAN) allow for creating images from existing databases. We applied a modified light-weight GAN (FastGAN) algorithm to cerebral blood flow SPECTs and aimed to evaluate whether this technology can generate created images close to real patients. Investigating three anatomical levels (cerebellum, CER; basal ganglia, BG; cortex, COR), 551 normal (248 CER, 174 BG, 129 COR) and 387 pathological brain SPECTs using N-isopropyl p-I-123-iodoamphetamine (¹²³I-IMP) were included. For the latter scans, cerebral ischemic disease comprised 291 uni- (66 CER, 116 BG, 109 COR) and 96 bilateral defect patterns (44 BG, 52 COR). Our model was trained using a three-compartment anatomical input (dataset 'A'; including CER, BG, and COR), while for dataset 'B', only one anatomical region (COR) was included. Quantitative analyses provided mean counts (MC) and left/right (LR) hemisphere ratios, which were then compared to quantification from real images. For MC, 'B' was significantly different for normal and bilateral defect patterns (P < 0.0001, respectively), but not for unilateral ischemia (P = 0.77). Comparable results were recorded for LR, as normal and ischemia scans were significantly different relative to images acquired from real patients (P ≤ 0.01, respectively). Images provided by 'A', however, revealed comparable quantitative results when compared to real images, including normal (P = 0.8) and pathological scans (unilateral, P = 0.99; bilateral, P = 0.68) for MC. For LR, only uni- (P = 0.03), but not normal or bilateral defect scans (P ≥ 0.08) reached significance relative to images of real patients. With a minimum of only three anatomical compartments serving as stimuli, created cerebral SPECTs are indistinguishable to images from real patients. The applied FastGAN algorithm may allow to provide sufficient scan numbers in various clinical scenarios, e.g., for "data-hungry" deep learning technologies or in the context of orphan diseases.

Evaluating the Role of Reduced Oxygen Saturation and Vascular Damage in Traumatic Brain Injury Using Magnetic Resonance Perfusion-Weighted Imaging and Susceptibility-Weighted Imaging and Mapping

The cerebral vasculature, along with neurons and axons, is vulnerable to biomechanical insult during traumatic brain injury (TBI). Trauma-induced vascular injury is still an underinvestigated area in TBI research. Cerebral blood flow and metabolism could be important future treatment targets in neural critical care. Magnetic resonance imaging offers a number of key methods to probe vascular injury and its relationship with traumatic hemorrhage, perfusion deficits, venous blood oxygen saturation changes, and resultant tissue damage. They make it possible to image the hemodynamics of the brain, monitor regional damage, and potentially show changes induced in the brain's function not only acutely but also longitudinally following treatment. These methods have recently been used to show that even mild TBI (mTBI) subjects can have vascular abnormalities, and thus they provide a major step forward in better diagnosing mTBI patients.

Imaging of cerebral blood flow in patients with severe traumatic brain injury in the neurointensive care

Ischemia is a common and deleterious secondary injury following traumatic brain injury (TBI). A great challenge for the treatment of TBI patients in the neurointensive care unit (NICU) is to detect early signs of ischemia in order to prevent further advancement and deterioration of the brain tissue. Today, several imaging techniques are available to monitor cerebral blood flow (CBF) in the injured brain such as positron emission tomography (PET), single-photon emission computed tomography, xenon computed tomography (Xenon-CT), perfusion-weighted magnetic resonance imaging (MRI), and CT perfusion scan. An ideal imaging technique would enable continuous non-invasive measurement of blood flow and metabolism across the whole brain. Unfortunately, no current imaging method meets all these criteria. These techniques offer snapshots of the CBF. MRI may also provide some information about the metabolic state of the brain. PET provides images with high resolution and quantitative measurements of CBF and metabolism; however, it is a complex and costly method limited to few TBI centers. All of these methods except mobile Xenon-CT require transfer of TBI patients to the radiological department. Mobile Xenon-CT emerges as a feasible technique to monitor CBF in the NICU, with lower risk of adverse effects. Promising results have been demonstrated with Xenon-CT in predicting outcome in TBI patients. This review covers available imaging methods used to monitor CBF in patients with severe TBI.

Clinical utility of SPECT neuroimaging in the diagnosis and treatment of traumatic brain injury: a systematic review

Purpose

This systematic review evaluated the clinical utility of single photon emission computed tomography (SPECT) in traumatic brain injury (TBI).

Methods

After defining a PICO Statement (Population, Intervention, Comparison and Outcome Statement), PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) criteria were applied to identify 1600 articles. After screening, 374 articles were eligible for review. Inclusion for review was focus on SPECT in the setting of mild, moderate, or severe TBI with cerebral lobar specificity of SPECT findings. Other inclusion criteria were comparison modalities in the same subjects and articles in English. Foreign language articles, SPECT studies that did not include comparison modalities, and case reports were not included for review.

Results

We identified 19 longitudinal and 52 cross-sectional studies meeting inclusion criteria. Three longitudinal studies examined diagnostic predictive value. The first showed positive predictive value increases from initial SPECT scan shortly after trauma to one year follow up scans, from 59% to 95%. Subsequent work replicated these results in a larger cohort. Longitudinal and cross sectional studies demonstrated SPECT lesion localization not detected by CT or MRI. The most commonly abnormal regions revealed by SPECT in cross-sectional studies were frontal (94%) and temporal (77%) lobes. SPECT was found to outperform both CT and MRI in both acute and chronic imaging of TBI, particularly mild TBI. It was also found to have a near 100% negative predictive value.

Conclusions

This review demonstrates Level IIA evidence (at least one non-randomized controlled trial) for the value of SPECT in TBI. Given its advantages over CT and MRI in the detection of mild TBI in numerous studies of adequate quality, and given its excellent negative predictive value, it may be an important second test in settings where CT or MRI are negative after a closed head injury with post-injury neurological or psychiatric symptoms.

Diffuse Axonal Injury: a Brief Review and Examples of the Use of Neurofunctional Imaging (Tc-99m HMPAO SPECT) in Diagnosis and Follow-up

Diffuse axonal injury (DAI) is a growing problem nowadays as its social and economic costs amount to millions of dollars. DAI is now thought to be the predominant mechanism of injury in almost half the cases of traumatic brain injury connected with loss of consciousness. Computed tomography and magnetic resonance imaging are substantial techniques to diagnose DAI but they have their limitations. Neuropsychological tests used in follow-up disclose persistent disabilities in patients with total regression of CT and MRI changes. In those situations SPECT is appropriate as it shows lesions not disclosed by other imaging techniques. This article describes two cases in which usefulness of SPECT has been proved. A brief review of DAI has been included.

Early recovery after closed traumatic head injury: Somatosensory evoked potentials and clinical findings

OBJECTIVE

To determine the ability of somatosensory evoked potentials (SEP) compared with clinical findings to monitor and predict recovery in patients suffering from closed head injury with predominantly diffuse axonal injury (DAI).

DESIGN

Prospective cohort study.

SETTING

Neurologic intensive care unit (ICU) of a university hospital.

PATIENTS

Serial SEP recordings were obtained from 31 consecutive patients with closed head injury. The first SEP was recorded within 48 hrs after trauma, followed by recordings after another 2 days, after which the time interval for each consecutive recording was doubled. Clinical examinations were performed every 6 hrs during the ICU stay and daily after transfer to a general neurologic ward.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Twenty-three of 31 patients demonstrated pathologic SEP findings at initial examination. Of these patients, 11 recovered clinically, two remained vegetative, and ten died. In all 11 patients with clinical recovery, SEP also recovered. In 8 of 31 patients, initial SEPs were normal and remained normal until discharge, all eight had a good outcome. Initial SEP findings were related with outcome at 6 months (p = .02), and follow-up studies increased the predictive value of SEP studies (p = .009). Other factors related to outcome included age, severity of DAI, and length of ICU/hospital stay. In the 11 patients with SEP and clinical recovery, early (day 2) and late (>or=2 months) recovery was documented. Early and reliable SEP indicators of improvement included N20-P25-Amplitudes (mean recovery, 8.5 days) and central conduction time (9.6 days). Pupillary light reaction (6.4 days), Babinski reflex (12.4 days), and Glasgow Coma Score (9.6 days) were the most valuable clinical findings. Recovery of the Glasgow Coma Score frequently coincided with reduction of sedatives. In most patients, recovery was detected with SEP before clinical recovery (7/11 patients, time difference 1 wk).

CONCLUSIONS

Initial SEP findings correlate with long-term outcome in patients with closed head injury with DAI. Initial bilaterally absent cortical responses in the SEP reliably predicted death, whereas completely normal SEP findings predicted good long-term outcome. Early recovery after DAI can be detected with serial SEP recordings despite sedative medications. Electrophysiologic recovery frequently precedes clinical recovery.