Objective
While the use of biomarkers for the detection of early and preclinical Alzheimer's Disease has become essential, the need to wait for over an hour after injection to obtain sufficient image ...quality can be challenging for patients with suspected dementia and their caregivers. This study aimed to develop an image-based deep-learning technique to generate delayed uptake patterns of amyloid positron emission tomography (PET) images using only early-phase images obtained from 0–20 min after radiotracer injection.
Methods
We prepared pairs of early and delayed
11
CPiB dynamic images from 253 patients (cognitively normal
n
= 32, fronto-temporal dementia
n
= 39, mild cognitive impairment
n
= 19, Alzheimer’s disease
n
= 163) as a training dataset. The neural network was trained with the early images as the input, and the output was the corresponding delayed image. A U-net convolutional neural network (CNN) and a conditional generative adversarial network (C-GAN) were used for the deep-learning architecture and the data augmentation methods, respectively. Then, an independent test data set consisting of early-phase amyloid PET images (
n
= 19) was used to generate corresponding delayed images using the trained network. Two nuclear medicine physicians interpreted the actual delayed images and predicted delayed images for amyloid positivity. In addition, the concordance of the actual delayed and predicted delayed images was assessed statistically.
Results
The concordance of amyloid positivity between the actual versus AI-predicted delayed images was 79%(
κ
= 0.60) and 79% (
κ
= 0.59) for each physician, respectively. In addition, the physicians’ agreement rate was at 89% (
κ
= 0.79) when the same image was interpreted. And, the actual versus AI-predicted delayed images were not readily distinguishable (correct answer rate, 55% and 47% for each physician, respectively). The statistical comparison of the actual versus the predicted delated images indicated that the peak signal-to-noise ratio (PSNR) was 21.8 dB ± 2.2 dB, and the structural similarity index (SSIM) was 0.45 ± 0.04.
Conclusion
This study demonstrates the feasibility of an image-based deep-learning framework to predict delayed patterns of Amyloid PET uptake using only the early phase images. This AI-based image generation method has the potential to reduce scan time for amyloid PET and increase the patient throughput, without sacrificing diagnostic accuracy for amyloid positivity.
BACKGROUND AND PURPOSE—Although studies have attempted to differentiate intracranial vascular disease using vessel wall magnetic resonance imaging (VWI), none have incorporated multicontrast imaging. ...This study uses T1- and T2-weighted VWI to differentiate intracranial vasculopathies.
METHODS—We retrospectively reviewed patients with clinically defined intracranial vasculopathies causing luminal stenosis/irregularity who underwent VWI studies. Two blinded experts evaluated T1 precontrast and postcontrast and T2-weighted VWI characteristics, including the pattern of wall thickening; presence, pattern, and intensity of postcontrast enhancement; and T2 signal characteristics.
RESULTS—Twenty-one cases of atherosclerosis (intracranial atherosclerotic disease ICAD), 4 of reversible cerebral vasoconstriction syndrome, and 4 of vasculitis were identified, with a total of 118 stenotic lesions (81 ICAD, 22 reversible cerebral vasoconstriction syndrome, and 15 vasculitic lesions). There was substantial to excellent inter-reader agreement for the assessment of lesional T2 hyperintensity (κ=0.80), pattern of wall thickening (κ=0.87), presence (κ=0.90), pattern (κ=0.73), and intensity (κ=0.77) of enhancement. ICAD lesions were significantly more likely to have eccentric wall involvement (90.1%) than reversible cerebral vasoconstriction syndrome (8.2%; P<0.001) and vasculitic lesions (6.7%; P<0.001) and were also more likely to have T2 hyperintensity present than the other 2 vasculopathies (79% versus 0%; P<0.001). There were also significant differences in the presence, intensity, and pattern of enhancement between all lesion types. Combining T1 and T2 VWI increased the sensitivity of VWI in differentiating ICAD from other vasculopathies from 90.1% to 96.3%.
CONCLUSIONS—Multicontrast VWI can be a complementary tool for intracranial vasculopathy differentiation, which often leads to more invasive workups when reversible cerebral vasoconstriction syndrome and vasculitis are in the differential diagnosis.
Orbital neoplasms in adults may be categorized on the basis of location and histologic type. Imaging features of these lesions often reflect their tissue composition. Cavernous malformations (also ...known as cavernous hemangiomas), although not true neoplasms, are the most common benign adult orbital tumor. They typically appear as a well-circumscribed, ovoid intraconal mass on cross-sectional images. Lymphoma, which may be primary or secondary to systemic disease, is the most prevalent orbital neoplasm in older adults (≥60 years of age). Choroidal melanoma is the most common primary adult ocular malignancy. Melanin has intrinsic T1 and T2 shortening effects, classically manifesting with hyperintense signal on T1-weighted magnetic resonance (MR) images and with hypointense signal on T2-weighted images. However, amelanotic or mildly pigmented lesions of melanoma do not demonstrate these characteristic MR imaging features. Breast cancer is the most common malignancy to metastasize to the orbit, followed by prostate cancer, melanoma, and lung cancer. In women with bilateral enophthalmos, metastatic scirrhous breast cancer should be considered in the differential diagnosis. Neoplasms that arise from the optic nerve or its sheath include glioma and meningioma. At imaging, gliomas often cause fusiform expansion of the optic nerve, in which the nerve itself cannot be delineated from the lesion. In contrast, meningiomas classically have a "tram-track" configuration, whereby the contrast-enhancing tumor is seen alongside the optic nerve. Neoplasms that derive from peripheral nerves include schwannoma and neurofibroma, the latter of which is associated with neurofibromatosis type 1. MR imaging is particularly valuable for evaluation of orbital neoplasms, as it provides critical anatomic information about ocular structures involved, perineural spread, and intracranial extension.
There is increasing scrutiny from healthcare organizations towards the utility and associated costs of imaging. MRI has traditionally been used as a high‐end modality, and although shown extremely ...important for many types of clinical scenarios, it has been suggested as too expensive by some. This editorial will try and explain how value should be addressed and gives some insights and practical examples of how value of MRI can be increased. It requires a global effort to increase accessibility, value for money, and impact on patient management. We hope this editorial sheds some light and gives some indications of where the field may wish to address some of its research to proactively demonstrate the value of MRI.
Level of Evidence: 5
Technical Efficacy: Stage 5
J. Magn. Reson. Imaging 2019;49:e14–e25.
To assess the prevalence, severity, and cost estimates associated with motion artifacts identified on clinical MR examinations, with a focus on the neuroaxis.
A retrospective review of 1 randomly ...selected full calendar week of MR examinations (April 2014) was conducted for the detection of significant motion artifacts in examinations performed at a single institution on 3 different MR scanners. A base-case cost estimate was computed from recently available institutional data, and correlated with sequence time and severity of motion artifacts.
A total of 192 completed clinical examinations were reviewed. Significant motion artifacts were identified on sequences in 7.5% of outpatient and 29.4% of inpatient and/or emergency department MR examinations. The prevalence of repeat sequences was 19.8% of total MRI examinations. The base-case cost estimate yielded a potential cost to the hospital of $592 per hour in lost revenue due to motion artifacts. Potential institutional average costs borne (revenue forgone) of approximately $115,000 per scanner per year may affect hospitals, owing to motion artifacts (univariate sensitivity analysis suggested a lower bound of $92,600, and an upper bound of $139,000).
Motion artifacts represent a frequent cause of MR image degradation, particularly for inpatient and emergency department patients, resulting in substantial costs to the radiology department. Greater attention and resources should be directed toward providing practical solutions to this dilemma.
Purpose
Extracranial ICA imaging has largely focused on the degree of luminal stenosis, but recent advances suggest specific plaque features are crucial in stroke risk assessment. We evaluated the ...current state of reporting carotid plaque features on neck CTAs at an academic institution.
Methods
In this retrospective observational study, we included neck CTAs performed on patients over age 50 with any reported carotid plaque. We evaluated reports for mention of the following: degree of luminal stenosis, soft plaque, calcified plaque, plaque thickness, quantification of soft and calcified plaque, plaque ulceration, and increased risk associated with specific features. We used Fisher’s exact test to compare how often each feature was mentioned.
Results
We included a total of 651 reports from unique patients (mean age, 68.1 ± 13.3 years). A total of 639 reports (98.1%) explicitly mentioned degree of stenosis per NASCET criteria. Specific plaque features were less frequently characterized: soft plaque in 116 (17.8%); calcified plaque in 166 (25.5%); quantification of the amount of soft plaque and calcified plaque in 24 (3.7%) and 16 (2.5%) reports, respectively; plaque thickness in 12 (1.8%); plaque ulceration in 476 (73.1%); and increased risk associated with plaque in 2 (0.3%). Degree of stenosis was statistically more likely to be mentioned than any other plaque feature (
p
< 0.001).
Conclusion
Currently, nearly all reports mention the degree of luminal stenosis on neck CTAs while a significant minority mention specific plaque features. Despite mounting evidence of the importance of carotid plaque features in stroke risk assessment, radiology reports do not routinely report these findings.
This article reviews the normal anatomy and variants of the anterior skull base and sinonasal cavities that are relevant to endoscopic sinus and skull base surgery. Radiologists should be aware of ...sinonasal anatomy that can be impediments to surgical access and increase risk of vascular or cranial nerve injury during surgery. Imaging features of the paranasal sinuses and anterior skull base pathologies are also discussed.
Imaging plays an essential role in identifying intracranial injury in patients with traumatic brain injury (TBI). The goals of imaging include (1) detecting injuries that may require immediate ...surgical or procedural intervention, (2) detecting injuries that may benefit from early medical therapy or vigilant neurologic supervision, and (3) determining the prognosis of patients to tailor rehabilitative therapy or help with family counseling and discharge planning. In this article, the authors perform a review of the evidence on the utility of various imaging techniques in patients presenting with TBI to provide guidance for evidence-based, clinical imaging protocols. The intent of this article is to suggest practical imaging recommendations for patients presenting with TBI across different practice settings and to simultaneously provide the rationale and background evidence supporting their use. These recommendations should ultimately assist referring physicians faced with the task of ordering appropriate imaging tests in particular patients with TBI for whom they are providing care. These recommendations should also help radiologists advise their clinical colleagues on appropriate imaging utilization for patients with TBI.
PDF
