The Swiss Approach - feasibility of a national low-dose CT lung cancer screening program.

BACKGROUND: Lung cancer is the leading cause of cancer-related deaths in Switzerland. Despite this, there is no lung cancer screening program in the country. In the United States, low-dose computed tomography (LDCT) lung cancer screening is partially established and endorsed by guidelines. Moreover, evidence is growing that screening reduces lung cancer-related mortality and this was recently shown in a large European randomized controlled trial. Implementation of a lung cancer screening program, however, is challenging and depends on many country-specific factors. The goal of this article is to outline a potential Swiss lung cancer screening program. FRAMEWORK: An exhaustive literature review on international screening models as well as interviews and site visits with international experts were initiated. Furthermore, workshops and interviews with national experts and stakeholders were conducted to share experiences and to establish the basis for a national Swiss lung cancer screening program. SCREENING APPROACH: General practitioners, pulmonologists and the media should be part of the recruitment process. Decentralisation of the screening might lead to a higher adherence rate. To reduce stigmatisation, the screening should be integrated in a "lung health check". Standardisation and a common quality level are mandatory. The PLCOm2012 risk calculation model with a threshold of 1.5% risk for developing cancer in the next six years should be used in addition to established inclusion criteria. Biennial screening is preferred. LUNG RADS and NELSON+ are applied as classification models for lung nodules. CONCLUSION: Based on data from recent studies, literature research, a health technology assessment, the information gained from this project and a pilot study the Swiss Interest Group for lung cancer screening (CH-LSIG) recommends the timely introduction of a systematic lung cancer screening program in Switzerland. The final decision is for the Swiss Cancer Screening Committee to make.

Performance of Spectral Photon-Counting Coronary CT Angiography and Comparison with Energy-Integrating-Detector CT: Objective Assessment with Model Observer.

AIMS: To evaluate spectral photon-counting CT's (SPCCT) objective image quality characteristics in vitro, compared with standard-of-care energy-integrating-detector (EID) CT. METHODS: We scanned a thorax phantom with a coronary artery module at 10 mGy on a prototype SPCCT and a clinical dual-layer EID-CT under various conditions of simulated patient size (small, medium, and large). We used filtered back-projection with a soft-tissue kernel. We assessed noise and contrast-dependent spatial resolution with noise power spectra (NPS) and target transfer functions (TTF), respectively. Detectability indices (d') of simulated non-calcified and lipid-rich atherosclerotic plaques were computed using the non-pre-whitening with eye filter model observer. RESULTS: SPCCT provided lower noise magnitude (9-38% lower NPS amplitude) and higher noise frequency peaks (sharper noise texture). Furthermore, SPCCT provided consistently higher spatial resolution (30-33% better TTF10). In the detectability analysis, SPCCT outperformed EID-CT in all investigated conditions, providing superior d'. SPCCT reached almost perfect detectability (AUC ≈ 95%) for simulated 0.5-mm-thick non-calcified plaques (for large-sized patients), whereas EID-CT had lower d' (AUC ≈ 75%). For lipid-rich atherosclerotic plaques, SPCCT achieved 85% AUC vs. 77.5% with EID-CT. CONCLUSIONS: SPCCT outperformed EID-CT in detecting simulated coronary atherosclerosis and might enhance diagnostic accuracy by providing lower noise magnitude, markedly improved spatial resolution, and superior lipid core detectability.

CVSnet: A machine learning approach for automated central vein sign assessment in multiple sclerosis.

The central vein sign (CVS) is an efficient imaging biomarker for multiple sclerosis (MS) diagnosis, but its application in clinical routine is limited by inter-rater variability and the expenditure of time associated with manual assessment. We describe a deep learning-based prototype for automated assessment of the CVS in white matter MS lesions using data from three different imaging centers. We retrospectively analyzed data from 3 T magnetic resonance images acquired on four scanners from two different vendors, including adults with MS (n = 42), MS mimics (n = 33, encompassing 12 distinct neurological diseases mimicking MS) and uncertain diagnosis (n = 5). Brain white matter lesions were manually segmented on FLAIR* images. Perivenular assessment was performed according to consensus guidelines and used as ground truth, yielding 539 CVS-positive (CVS+ ) and 448 CVS-negative (CVS- ) lesions. A 3D convolutional neural network ("CVSnet") was designed and trained on 47 datasets, keeping 33 for testing. FLAIR* lesion patches of CVS+ /CVS- lesions were used for training and validation (n = 375/298) and for testing (n = 164/150). Performance was evaluated lesion-wise and subject-wise and compared with a state-of-the-art vesselness filtering approach through McNemar's test. The proposed CVSnet approached human performance, with lesion-wise median balanced accuracy of 81%, and subject-wise balanced accuracy of 89% on the validation set, and 91% on the test set. The process of CVS assessment, in previously manually segmented lesions, was ~ 600-fold faster using the proposed CVSnet compared with human visual assessment (test set: 4 seconds vs. 40 minutes). On the validation and test sets, the lesion-wise performance outperformed the vesselness filter method (P < 0.001). The proposed deep learning prototype shows promising performance in differentiating MS from its mimics. Our approach was evaluated using data from different hospitals, enabling larger multicenter trials to evaluate the benefit of introducing the CVS marker into MS diagnostic criteria.

"Dark-blood" dual-energy computed tomography angiography for thoracic aortic wall imaging.

OBJECTIVES: To assess the capability of a newly developed material decomposition method from contrast-enhanced dual-energy CT images, aiming to better visualize the aortic wall and aortic intramural hematoma (IMH), compared with true non-contrast (TNC) CT. MATERIALS AND METHODS: Twenty-two patients (11 women; mean age, 61 ± 20 years) with acute chest pain underwent 25 dual-layer non-contrast and contrast-enhanced CT. CT-angiography images were retrospectively processed using two-material decomposition analysis, where we defined the first material as the content of a region of interest placed in the ascending aorta for each patient, and the second material as water. Two independent radiologists assessed the images from the second material termed "dark-blood" images and the TNC images regarding contrast-to-noise ratio (CNR) between the wall and the lumen, diagnostic quality regarding the presence of aortic wall thickening, and the inner/outer vessel wall conspicuity. RESULTS: Diagnostic quality scores in normal aortic segments were 0.9 ± 0.3 and 2.7 ± 0.6 (p < 0.001) and wall conspicuity scores were 0.7 ± 0.5 and 1.8 ± 0.3 (p < 0.001) on TNC and dark-blood images, respectively. In aortic segments with IMH, diagnostic quality scores were 1.7 ± 0.5 and 2.4 ± 0.6 (p < 0.001) and wall conspicuity scores were 0.7 ± 0.7 and 1.8 ± 0.3 (p < 0.001) on TNC and dark-blood images, respectively. In normal aortic segments, CNRs were 0.3 ± 0.2 and 2.8 ± 0.9 on TNC and dark-blood images, respectively (p < 0.001). In aortic segments with IMH, CNRs were 0.3 ± 0.2 and 4.0 ± 1.0 on TNC and dark-blood images, respectively (p < 0.001). CONCLUSIONS: Compared with true non-contrast CT, dark-blood material decomposition maps enhance quantitative and qualitative image quality for the assessment of normal aortic wall and IMH. KEY POINTS: • Current dual-energy CT-angiography provides virtual non-contrast and bright-blood images. • Dark-blood images represent a new way to assess the vascular wall structure with dual-energy CT and can improve the lumen-to-wall contrast compared with true non-contrast CT. • This dual-energy CT material decomposition method is likely to improve contrast resolution in other applications as well, taking advantage of the high spatial resolution of CT.

Partial volume-aware assessment of multiple sclerosis lesions.

White-matter lesion count and volume estimation are key to the diagnosis and monitoring of multiple sclerosis (MS). Automated MS lesion segmentation methods that have been proposed in the past 20 years reach their limits when applied to patients in early disease stages characterized by low lesion load and small lesions. We propose an algorithm to automatically assess MS lesion load (number and volume) while taking into account the mixing of healthy and lesional tissue in the image voxels due to partial volume effects. The proposed method works on 3D MPRAGE and 3D FLAIR images as obtained from current routine MS clinical protocols. The method was evaluated and compared with manual segmentation on a cohort of 39 early-stage MS patients with low disability, and showed higher Dice similarity coefficients (median DSC = 0.55) and higher detection rate (median DR = 61%) than two widely used methods (median DSC = 0.50, median DR < 45%) for automated MS lesion segmentation. We argue that this is due to the higher performance in segmentation of small lesions, which are inherently prone to partial volume effects.

Diffusion-weighted magnetic resonance imaging in metastatic gastrointestinal stromal tumor (GIST): a pilot study on the assessment of treatment response in comparison with 18F-FDG PET/CT.

BACKGROUND: Diffusion-weighted magnetic resonance imaging (MRI) is increasingly being used for assessing the treatment succes in oncology, but the real clinical value needs to evaluated by comparison with other, already established, metabolic imaging techniques. PURPOSE: To prospectively evaluate the clinical potential of diffusion-weighted MRI with apparent diffusion coefficient (ADC) mapping for gastrointestinal stromal tumor (GIST) response to targeted therapy compared with 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT). MATERIAL AND METHODS: Eight patients (mean age, 56 ± 11 years) known to have metastatic GIST underwent 18F-FDG PET/CT and MRI (T1Gd, DWI [b = 50,300,600], ADC mapping) simultaneously, before and after change in targeted therapy. MR and PET/CT examinations were first analyzed blindly. Second, PET/CT images were co-registered with T1Gd-MR images for lesion detection. Only 18F-FDG avid lesions were considered. Maximum standardized uptake value (SUVmax) and the corresponding minimum ADCmin were measured for the six largest lesions per patient, if any, on baseline and follow-up examinations. The relationship between changes in SUVmax and ADCmin was analyzed (Spearman's correlation). RESULTS: Twenty-four metastases (12 hepatic, 12 extra-hepatic) were compared on PET/CT and MR images. SUVmax decreased from 7.7 ± 8.1 g/mL to 5.5 ± 5.4 g/mL (P = 0.20), while ADCmin increased from 1.2 ± 0.3 × 10(-3)mm(2)/s to 1.5 ± 0.3 × 10(-3)mm(2)/s (P = 0.0002). There was a significant association between changes in SUVmax and ADCmin (rho = - 0.62, P = 0.0014), but not between changes in lesions size (P = 0.40). CONCLUSION: Changes in ADCmin correlated with the response of 18F-FDG avid GIST to targeted therapy. Thus, diffusion-weighted MRI may represent a radiation-free alternative for follow-up treatment for metastatic GIST patients.

Adaptive filtering methods for identifying cross-frequency couplings in human EEG.

Oscillations have been increasingly recognized as a core property of neural responses that contribute to spontaneous, induced, and evoked activities within and between individual neurons and neural ensembles. They are considered as a prominent mechanism for information processing within and communication between brain areas. More recently, it has been proposed that interactions between periodic components at different frequencies, known as cross-frequency couplings, may support the integration of neuronal oscillations at different temporal and spatial scales. The present study details methods based on an adaptive frequency tracking approach that improve the quantification and statistical analysis of oscillatory components and cross-frequency couplings. This approach allows for time-varying instantaneous frequency, which is particularly important when measuring phase interactions between components. We compared this adaptive approach to traditional band-pass filters in their measurement of phase-amplitude and phase-phase cross-frequency couplings. Evaluations were performed with synthetic signals and EEG data recorded from healthy humans performing an illusory contour discrimination task. First, the synthetic signals in conjunction with Monte Carlo simulations highlighted two desirable features of the proposed algorithm vs. classical filter-bank approaches: resilience to broad-band noise and oscillatory interference. Second, the analyses with real EEG signals revealed statistically more robust effects (i.e. improved sensitivity) when using an adaptive frequency tracking framework, particularly when identifying phase-amplitude couplings. This was further confirmed after generating surrogate signals from the real EEG data. Adaptive frequency tracking appears to improve the measurements of cross-frequency couplings through precise extraction of neuronal oscillations.

Assessment of liver tumor response by high-field (3 T) MRI after radiofrequency ablation: short- and mid-term evolution of diffusion parameters within the ablation zone.

PURPOSE: To compare the apparent diffusion coefficient (ADC) values of malignant liver lesions on diffusion-weighted MRI (DWI) before and after successful radiofrequency ablation (RF ablation). MATERIALS AND METHODS: Thirty-two patients with 43 malignant liver lesions (23/20: metastases/hepatocellular carcinomas (HCC)) underwent liver MRI (3.0 T) before (<1 month) and after RF ablation (at 1, 3 and 6 months) using T2-, gadolinium-enhanced T1- and DWI-weighted MR sequences. Jointly, two radiologists prospectively measured ADCs for each lesion by means of two different regions of interest (ROIs), first including the whole lesion and secondly the area with the visibly most restricted diffusion (MRDA) on ADC map. Changes of ADCs were evaluated with ANOVA and Dunnett tests. RESULTS: Thirty-one patients were successfully treated, while one patient was excluded due to focal recurrence. In metastases (n=22), the ADC in the whole lesion and in MRDA showed an up-and-down evolution. In HCC (n=20), the evolution of ADC was more complex, but with significantly higher values (p=0.013) at 1 and 6 months after RF ablation. CONCLUSION: The ADC values of malignant liver lesions successfully treated by RF ablation show a predictable evolution and may help radiologists to monitor tumor response after treatment.

Automatic quality assessment in structural brain magnetic resonance imaging.

MRI has evolved into an important diagnostic technique in medical imaging. However, reliability of the derived diagnosis can be degraded by artifacts, which challenge both radiologists and automatic computer-aided diagnosis. This work proposes a fully-automatic method for measuring image quality of three-dimensional (3D) structural MRI. Quality measures are derived by analyzing the air background of magnitude images and are capable of detecting image degradation from several sources, including bulk motion, residual magnetization from incomplete spoiling, blurring, and ghosting. The method has been validated on 749 3D T(1)-weighted 1.5T and 3T head scans acquired at 36 Alzheimer's Disease Neuroimaging Initiative (ADNI) study sites operating with various software and hardware combinations. Results are compared against qualitative grades assigned by the ADNI quality control center (taken as the reference standard). The derived quality indices are independent of the MRI system used and agree with the reference standard quality ratings with high sensitivity and specificity (>85%). The proposed procedures for quality assessment could be of great value for both research and routine clinical imaging. It could greatly improve workflow through its ability to rule out the need for a repeat scan while the patient is still in the magnet bore.

Fibertract segmentation in position orientation space from high angular resolution diffusion MRI.

In diffusion MRI, standard approaches for fibertract identification are based on algorithms that generate lines of coherent diffusion, currently known as tractography. A tract is then identified as a set of such lines selected on some criteria. In the present study, we investigate whether fibertract identification can be formulated as a segmentation task that recognizes a fibertract as a region where diffusion is intense and coherent. Indeed, we show that it is possible to segment efficiently well-known fibertracts with classical image processing methods provided that the problem is formulated in a five-dimensional space of position and orientation. As an example, we choose to adapt to this newly defined high-dimensional non-Euclidean space, called position orientation space, an algorithm based on the hidden Markov random field framework. Structures such as the cerebellar peduncles, corticospinal tract, association bundles can be identified and represented in three dimensions by a back projection technique similar to maximum intensity projection. Potential advantages and drawbacks as compared to classical tractography are discussed; for example, it appears that our formulation handles naturally crossing tracts and is not biased by human intervention.

Perfusion-CT assessment of infarct core and penumbra: receiver operating characteristic curve analysis in 130 patients suspected of acute hemispheric stroke.

BACKGROUND AND PURPOSE: Different definitions have been proposed to define the ischemic penumbra from perfusion-CT (PCT) data, based on parameters and thresholds tested only in small pilot studies. The purpose of this study was to perform a systematic evaluation of all PCT parameters (cerebral blood flow, volume [CBV], mean transit time [MTT], time-to-peak) in a large series of acute stroke patients, to determine which (combination of) parameters most accurately predicts infarct and penumbra. METHODS: One hundred and thirty patients with symptoms suggesting hemispheric stroke < or =12 hours from onset were enrolled in a prospective multicenter trial. They all underwent admission PCT and follow-up diffusion-weighted imaging/fluid-attenuated inversion recovery (DWI/FLAIR); 25 patients also underwent admission DWI/FLAIR. PCT maps were assessed for absolute and relative reduced CBV, reduced cerebral blood flow, increased MTT, and increased time-to-peak. Receiver-operating characteristic curve analysis was performed to determine the most accurate PCT parameter, and the optimal threshold for each parameter, using DWI/FLAIR as the gold standard. RESULTS: The PCT parameter that most accurately describes the tissue at risk of infarction in case of persistent arterial occlusion is the relative MTT (area under the curve=0.962), with an optimal threshold of 145%. The PCT parameter that most accurately describes the infarct core on admission is the absolute CBV (area under the curve=0.927), with an optimal threshold at 2.0 ml x 100 g(-1). CONCLUSIONS: In a large series of 130 patients, the optimal approach to define the infarct and the penumbra is a combined approach using 2 PCT parameters: relative MTT and absolute CBV, with dedicated thresholds.