Incorporating algorithmic uncertainty into a clinical machine deep learning algorithm for urgent head CTs.

Machine learning (ML) algorithms to detect critical findings on head CTs may expedite patient management. Most ML algorithms for diagnostic imaging analysis utilize dichotomous classifications to determine whether a specific abnormality is present. However, imaging findings may be indeterminate, and algorithmic inferences may have substantial uncertainty. We incorporated awareness of uncertainty into an ML algorithm that detects intracranial hemorrhage or other urgent intracranial abnormalities and evaluated prospectively identified, 1000 consecutive noncontrast head CTs assigned to Emergency Department Neuroradiology for interpretation. The algorithm classified the scans into high (IC+) and low (IC-) probabilities for intracranial hemorrhage or other urgent abnormalities. All other cases were designated as No Prediction (NP) by the algorithm. The positive predictive value for IC+ cases (N = 103) was 0.91 (CI: 0.84-0.96), and the negative predictive value for IC- cases (N = 729) was 0.94 (0.91-0.96). Admission, neurosurgical intervention, and 30-day mortality rates for IC+ was 75% (63-84), 35% (24-47), and 10% (4-20), compared to 43% (40-47), 4% (3-6), and 3% (2-5) for IC-. There were 168 NP cases, of which 32% had intracranial hemorrhage or other urgent abnormalities, 31% had artifacts and postoperative changes, and 29% had no abnormalities. An ML algorithm incorporating uncertainty classified most head CTs into clinically relevant groups with high predictive values and may help accelerate the management of patients with intracranial hemorrhage or other urgent intracranial abnormalities.

Diffusion MR Imaging of Large Vessel Occlusion Ischemic Stroke for Treatment Selection.

The role of MR imaging in the evaluation and management of ischemic stroke patients is large, and to cover it all is far beyond the scope of one article. Thus, the focus will be on the role of MR imaging in the great leap forward in stroke therapy: endovascular thrombectomy of large vessel occlusions (LVOs). Diffusion MR imaging has played a key role in the research leading to the current standard of care for LVO stroke because it is the most sensitive and reliable method for the early delineation of the ischemic core.

Ethical Issues Posed by Field Research Using Highly Portable and Cloud-Enabled Neuroimaging.

Highly portable, cloud-enabled neuroimaging technologies will fundamentally change neuroimaging research. Instead of participants traveling to the scanner, the scanner will now come to them. Field-based brain imaging research, including populations underrepresented in neuroscience research to date, will enlarge and diversify databases and pave the way for clinical and direct-to-consumer (DTC) applications. Yet these technological developments urgently require analysis of their ethical, legal, and social implications (ELSI). No consensus ethical frameworks for mobile neuroimaging exist, and existing policies for traditional MRI research are inadequate. Based on literature review and ethics analysis of neurotechnology development efforts, Shen et al. identify seven foundational, yet unresolved, ELSI issues posed by portable neuroimaging: (1) informed consent; (2) privacy; (3) capacity to accurately communicate neuroimaging results to remote participants; (4) extensive reliance on cloud-based artificial intelligence (AI) for data analysis; (5) potential bias of interpretive algorithms in diverse populations; (6) return of research results and incidental (or secondary) findings to research participants; and (7) responding to participant requests for access to their data. The article proposes a path forward to address these urgent issues.

Magnetic resonance imaging of neuroinflammation in chronic pain: a role for astrogliosis?

Noninvasive measures of neuroinflammatory processes in humans could substantially aid diagnosis and therapeutic development for many disorders, including chronic pain. Several proton magnetic resonance spectroscopy (H-MRS) metabolites have been linked with glial activity (ie, choline and myo-inositol) and found to be altered in chronic pain patients, but their role in the neuroinflammatory cascade is not well known. Our multimodal study evaluated resting functional magnetic resonance imaging connectivity and H-MRS metabolite concentration in insula cortex in 43 patients suffering from fibromyalgia, a chronic centralized pain disorder previously demonstrated to include a neuroinflammatory component, and 16 healthy controls. Patients demonstrated elevated choline (but not myo-inositol) in anterior insula (aIns) (P = 0.03), with greater choline levels linked with worse pain interference (r = 0.41, P = 0.01). In addition, reduced resting functional connectivity between aIns and putamen was associated with both pain interference (whole brain analysis, pcorrected < 0.01) and elevated aIns choline (r = -0.37, P = 0.03). In fact, aIns/putamen connectivity statistically mediated the link between aIns choline and pain interference (P < 0.01), highlighting the pathway by which neuroinflammation can impact clinical pain dysfunction. To further elucidate the molecular substrates of the effects observed, we investigated how putative neuroinflammatory H-MRS metabolites are linked with ex vivo tissue inflammatory markers in a nonhuman primate model of neuroinflammation. Results demonstrated that cortical choline levels were correlated with glial fibrillary acidic protein, a known marker for astrogliosis (Spearman r = 0.49, P = 0.03). Choline, a putative neuroinflammatory H-MRS-assessed metabolite elevated in fibromyalgia and associated with pain interference, may be linked with astrogliosis in these patients.

Simian immunodeficiency virus transiently increases brain temperature in rhesus monkeys: detection with magnetic resonance spectroscopy thermometry.

PURPOSE: To evaluate brain temperature effects of early simian immunodeficiency virus (SIV) infection in rhesus macaques using proton magnetic resonance spectroscopy (MRS) thermometry (MRSt) and to determine whether temperature correlates with brain choline or myo-inositol levels. METHODS: Brain temperature was retrospectively determined in serial MRS scans that had been acquired at baseline and at 2 and 4 weeks post-SIV infection (wpi) in 16 monkeys by calculating the chemical shift difference between N-acetylaspartate (NAA) and water peaks in sequentially acquired water-suppressed and unsuppressed point-resolved spectroscopy (PRESS) spectra. Frontal and parietal cortex, basal ganglia, and white matter spectra were analyzed. RESULTS: At 2 wpi, brain and rectal temperatures increased relative to baseline and normalized at 4 wpi. Brain temperatures correlated with choline levels in several brain areas, but not with myo-inositol levels. CONCLUSION: These data indicate that SIV transiently increases brain temperature soon after infection and that temperature is correlated with transient changes in choline levels. Given that choline levels are associated with brain inflammation in SIV-infected monkeys, our findings suggest that the SIV-induced temperature increase reflects brain inflammation. We conclude that MRSt may be informative in human immunodeficiency virus models and may be useful for assessing effects of treatments that reduce inflammation. This study also illustrates that existing MRS data sets containing unsuppressed water spectra can be used to measure tissue temperature, an important physiological parameter.

Patient selection for mechanical thrombectomy in posterior circulation emergent large-vessel occlusion.

Triage of posterior circulation stroke from emergent large-vessel occlusion (pc-ELVO) is challenging owing to the stuttering clinical course and potential for rapid decline. Growing clinical data support the use of mechanical thrombectomy in pc-ELVO, but there are limited data addressing the clinical and imaging criteria for patient selection. We present our triage algorithm used to select patients for endovascular therapy (EVT) in the setting of pc-ELVOS. We use a consecutive retrospective database from 2004 to 2016 to describe the practice patterns and prognostic factors for pc-ELVO patients treated using both medical and EVT. Patients with moderate to severe deficits (NIHSS > 10) did better when they received EVT ( p < 0.03), whereas patients with stable, mild deficits (NIHSS ≤ 10) did well (90% favorable outcome) regardless of treatment type. Roughly one-third of patients presenting with mild deficits deteriorated to moderate to severe deficits (NIHSS > 10), most of whom subsequently received EVT (9 of 12), with 56% favorable outcomes. Cerebellar and brainstem infarct volumes were independent imaging predictors of outcome. These results can be used to define triage criteria for use of EVT in pc-ELVO in future practice and clinical trials.

Clinical proton MR spectroscopy in central nervous system disorders.

A large body of published work shows that proton (hydrogen 1 [(1)H]) magnetic resonance (MR) spectroscopy has evolved from a research tool into a clinical neuroimaging modality. Herein, the authors present a summary of brain disorders in which MR spectroscopy has an impact on patient management, together with a critical consideration of common data acquisition and processing procedures. The article documents the impact of (1)H MR spectroscopy in the clinical evaluation of disorders of the central nervous system. The clinical usefulness of (1)H MR spectroscopy has been established for brain neoplasms, neonatal and pediatric disorders (hypoxia-ischemia, inherited metabolic diseases, and traumatic brain injury), demyelinating disorders, and infectious brain lesions. The growing list of disorders for which (1)H MR spectroscopy may contribute to patient management extends to neurodegenerative diseases, epilepsy, and stroke. To facilitate expanded clinical acceptance and standardization of MR spectroscopy methodology, guidelines are provided for data acquisition and analysis, quality assessment, and interpretation. Finally, the authors offer recommendations to expedite the use of robust MR spectroscopy methodology in the clinical setting, including incorporation of technical advances on clinical units.

The Massachusetts General Hospital acute stroke imaging algorithm: an experience and evidence based approach.

The Massachusetts General Hospital Neuroradiology Division employed an experience and evidence based approach to develop a neuroimaging algorithm to best select patients with severe ischemic strokes caused by anterior circulation occlusions (ACOs) for intravenous tissue plasminogen activator and endovascular treatment. Methods found to be of value included the National Institutes of Health Stroke Scale (NIHSS), non-contrast CT, CT angiography (CTA) and diffusion MRI. Perfusion imaging by CT and MRI were found to be unnecessary for safe and effective triage of patients with severe ACOs. An algorithm was adopted that includes: non-contrast CT to identify hemorrhage and large hypodensity followed by CTA to identify the ACO; diffusion MRI to estimate the core infarct; and NIHSS in conjunction with diffusion data to estimate the clinical penumbra.

Rapid identification of a major diffusion/perfusion mismatch in distal internal carotid artery or middle cerebral artery ischemic stroke.

BACKGROUND: We tested the hypothesis that in patients with occlusion of the terminal internal carotid artery and/or the proximal middle cerebral artery, a diffusion abnormality of 70 ml or less is accompanied by a diffusion/perfusion mismatch of at least 100%. METHODS: Sixty-eight consecutive patients with terminal ICA and/or proximal MCA occlusions and who underwent diffusion/perfusion MRI within 24 hours of stroke onset were retrospectively identified. DWI and mean transit time (MTT) volumes were measured. Prospectively, 48 consecutive patients were identified with the same inclusion criteria. DWI and time to peak (TTP) lesion volumes were measured. A large mismatch volume was defined as an MTT or TTP abnormality at least twice the DWI lesion volume. RESULTS: In the retrospective study, 49 of 68 patients had a DWI lesion volume ≤ 70 ml (mean 20.2 ml; SEM 2.9 ml). A DWI/MTT mismatch of > 100% was observed in all 49 patients (P < .0001). In the prospective study, there were 35/48 patients with DWI volumes ≤ 70 ml (mean 18.7 ml; SEM 3.0 ml). A mismatch > 100% was present in all 35 (P < .0001). CONCLUSIONS: Acute stroke patients with major anterior circulation artery occlusion are exceedingly likely to have a major diffusion/perfusion mismatch if the diffusion lesion volume is 70 ml or less. This suggests that physiology-based patient assessments may be made using only vessel imaging and diffusion MRI as a simple alternative to perfusion imaging.

Endovascular treatment of acute ischemic stroke: current indications.

Endovascular stroke therapy is an effective means of achieving reperfusion in stroke patients with proximal cerebral artery occlusions. However, current guideline recommendations express uncertainty regarding the clinical efficacy of catheter-based treatments, given the lack of supportive trial data. A critical problem is that it remains unclear which patients will benefit from endovascular therapy. As such, patient selection is likely highly variable in clinical practice. This article will review the existing data to discuss the clinical and imaging factors that are relevant to patient outcomes, and which may be used to guide endovascular treatment decisions. Anterior circulation strokes represent the primary focus of this review.

Quantitative assessment of core/penumbra mismatch in acute stroke: CT and MR perfusion imaging are strongly correlated when sufficient brain volume is imaged.

BACKGROUND AND PURPOSE: Our purpose was to determine (1) the correlation between quantitative CT and MR measurements of infarct core, penumbra, and mismatch; and (2) whether the difference between these measurements would alter patient selection for stroke clinical trials. METHODS: We studied 45 patients with acute middle cerebral artery stroke imaged a mean of 3.8 hours after onset (range, 0.48 to 8.35 hours) who underwent CT perfusion and MR diffusion (DWI)/perfusion imaging within 3 hours of each other. The DWI and MR-mean transit time (MTT) abnormalities were visually segmented using a semiautomated commercial analysis program. The CT-cerebral blood volume) and CT-MTT lesions were automatically segmented using a relative cerebral blood volume threshold of 0.56 and a relative MTT threshold of 1.50 on commercially available software. Percent mismatch was defined as [(MTT-DWI)/DWI volume]x100. Pearson correlation coefficients were calculated. RESULTS: There were significant correlations for DWI versus CT-cerebral blood volume lesion volumes (r2=0.88, P<0.001), for MR-MTT versus CT-MTT lesion volumes(r2=0.86, P<0.001), and for MR-MTT/DWI versus CT-MTT/CT-cerebral blood volume mismatch lesion volumes(r2=0.81, P<0.001). MR perfusion and CT perfusion agreed for determining: (1) infarct core < versus >or=100 mL in 41 of 45 (91.1%); (2) MTT lesion size < versus >2 cm diameter in 42 of 45 (93.3%); (3) mismatch < versus >20% in 41 of 45 (91.1%); and (4) inclusion versus exclusion from trial enrollment in 38 of 45 (84.4%) patients. Six of 7 disagreements were due to inadequate CT coverage. CONCLUSIONS: Advanced MR and CT perfusion imaging measurements of core/penumbra mismatch for patient selection in stroke trials are highly correlated when CT perfusion coverage is sufficient to include most of the ischemic region. Although MR is currently the preferred imaging method for determining core and penumbra, CT perfusion is comparable and potentially more available.