Explore the vast range of titles available.

Accurate classification of focal liver lesions is an important part of liver disease diagnostics. In clinical practice, the lesion type is often determined from the abdominal MR examination, which includes T2-weighted and dynamic contrast enhanced (DCE) MR images. To date, only T2-weighted images are exploited for automatic classification of focal liver lesions. In this study additional MR sequences and risk factors are used for automatic classification to improve the results and to make a step forward to a clinically useful aid for radiologists. Clinical MRI data sets of 95 patients with in total 125 benign lesions (40 adenomas, 29 cysts and 56 hemangiomas) and 88 malignant lesions (30 hepatocellular carcinomas (HCC) and 58 metastases) were included in this study. Contrast curve, gray level histogram, and gray level co-occurrence matrix texture features were extracted from the DCE-MR and T2-weighted images. In addition, risk factors including the presence of steatosis, cirrhosis, and a known primary tumor were used as features. Fifty features with the highest ANOVA F-score were selected and fed to an extremely randomized trees classifier. The classifier evaluation was performed using the leave-one-out principle and receiver operating characteristic (ROC) curve analysis. The overall accuracy for the classification of the five major focal liver lesion types is 0.77. The sensitivity/specificity is 0.80/0.78, 0.93/0.93, 0.84/0.82, 0.73/0.56, and 0.62/0.77 for adenoma, cyst, hemangioma, HCC, and metastasis, respectively. The proposed classification system using features derived from clinical DCE-MR and T2-weighted images, with additional risk factors is able to differentiate five common types of lesions and is a step forward to a clinically useful aid for focal liver lesion diagnosis.

The performance of current machine learning methods to detect heterogeneous pathology is limited by the quantity and quality of pathology in medical images. A possible solution is anomaly detection; an approach that can detect all abnormalities by learning how ‘normal’ tissue looks like. In this work, we propose an anomaly detection method using a neural network architecture for the detection of chronic brain infarcts on brain MR images. The neural network was trained to learn the visual appearance of normal appearing brains of 697 patients. We evaluated its performance on the detection of chronic brain infarcts in 225 patients, which were previously labeled. Our proposed method detected 374 chronic brain infarcts (68% of the total amount of brain infarcts) which represented 97.5% of the total infarct volume. Additionally, 26 new brain infarcts were identified that were originally missed by the radiologist during radiological reading. Our proposed method also detected white matter hyperintensities, anomalous calcifications, and imaging artefacts. This work shows that anomaly detection is a powerful approach for the detection of multiple brain abnormalities, and can potentially be used to improve the radiological workflow efficiency by guiding radiologists to brain anomalies which otherwise remain unnoticed.

Lacunar lesions (LLs) and white matter lesions (WMLs) affect cognition. We assessed whether lesions located in specific white matter tracts were associated with cognitive performance taking into account total lesion burden. Within the Second Manifestations of ARTerial disease Magnetic Resonance (SMART-MR) study, cross-sectional analyses were performed on 516 patients with manifest arterial disease. We applied an assumption-free voxel-based lesion-symptom mapping approach to investigate the relation between LL and WML locations on 1.5 Tesla brain MRI and compound scores of executive functioning, memory and processing speed. Secondly, a multivariable linear regression model was used to relate the regional volume of LLs and WMLs within specific white matter tracts to cognitive functioning. Voxel-based lesion-symptom mapping identified several clusters of voxels with a significant correlation between WMLs and executive functioning, mostly located within the superior longitudinal fasciculus and anterior thalamic radiation. In the multivariable linear regression model, a statistically significant association was found between regional LL volume within the superior longitudinal fasciculus and anterior thalamic radiation and executive functioning after adjustment for total LL and WML burden. These findings identify the superior longitudinal fasciculus and anterior thalamic radiation as key anatomical structures in executive functioning and emphasize the role of strategically located vascular lesions in vascular cognitive impairment.

Diffuse idiopathic skeletal hyperostosis (DISH) is characterized by anterior ossification of the spine and can lead to dysphagia and airway obstruction. The morphology of the newly formed bone in the cervical spine is different compared to the thoracic spine, possibly due to dissimilarities in local vascular anatomy. In this study the spatial relationship of the new bone with the arterial system, trachea and esophagus was analyzed and compared between subjects with and without DISH. Cervical computed tomography (CT) scans were obtained from five patients with dysphagia and DISH and ten control subjects. The location of the vertebral and carotid arteries, surface area of the hyperostosis and distance between the vertebral body and the trachea and esophagus was assessed in the axial view. The surface area of the newly formed bone was located symmetrically anterior to the vertebral body. The ossifications were non-flowing in the sagittal view and no segmental vessels were observed. Substantial displacement of the trachea/esophagus was present in the group with DISH compared to the controls. The hyperostosis at the cervical level was symmetrically distributed anterior to the vertebral bodies without a flowing pattern, in contrast to the asymmetrical flowing pattern typically found in the thoracic spine. The hypothesis that the vascular system acts as a natural barrier against new bone formation in DISH could be further supported with these findings. The significant ventral displacement of the trachea and esophagus may explain the mechanism of dysphagia and airway obstruction in DISH.

Studies on the impact of small vessel disease (SVD) on cognition generally focus on white matter hyperintensity (WMH) volume. The extent to which WMH location relates to cognitive performance has received less attention, but is likely to be functionally important. We examined the relation between WMH location and cognition in a memory clinic cohort of patients with sporadic SVD. A total of 167 patients with SVD were recruited from memory clinics. Assumption-free region of interest-based analyses based on major white matter tracts and voxel-wise analyses were used to determine the association between WMH location and executive functioning, visuomotor speed and memory. Region of interest-based analyses showed that WMHs located particularly within the anterior thalamic radiation and forceps minor were inversely associated with both executive functioning and visuomotor speed, independent of total WMH volume. Memory was significantly associated with WMH volume in the forceps minor, independent of total WMH volume. An independent assumption-free voxel-wise analysis identified strategic voxels in these same tracts. Region of interest-based analyses showed that WMH volume within the anterior thalamic radiation explained 6.8% of variance in executive functioning, compared to 3.9% for total WMH volume; WMH volume within the forceps minor explained 4.6% of variance in visuomotor speed and 4.2% of variance in memory, compared to 1.8% and 1.3% respectively for total WMH volume. Our findings identify the anterior thalamic radiation and forceps minor as strategic white matter tracts in which WMHs are most strongly associated with cognitive impairment in memory clinic patients with SVD. WMH volumes in individual tracts explained more variance in cognition than total WMH burden, emphasizing the importance of lesion location when addressing the functional consequences of WMHs.

AI provides tremendous opportunities for improving patient care, but at present there is little evidence of real-world uptake. An important barrier is the lack of well-designed, vendor-neutral and future-proof infrastructures for deployment. Because current AI algorithms are very narrow in scope, it is expected that a typical hospital will deploy many algorithms concurrently. Managing stand-alone point solutions for all of these algorithms will be unmanageable. A solution to this problem is a dedicated platform for deployment of AI. Here we describe a blueprint for such a platform and the high-level design and implementation considerations of such a system that can be used clinically as well as for research and development. Close collaboration between radiologists, data scientists, software developers and experts in hospital IT as well as involvement of patients is crucial in order to successfully bring AI to the clinic.

The circle of Willis (CoW) is a circular arrangement of arteries in the human brain, exhibiting significant anatomical variability. The CoW is extensively studied in relation to neurovascular pathologies, with certain anatomical variants previously linked to ischemic stroke and intracranial aneurysms. In an individual CoW, arteries might be absent (aplasia) or underdeveloped (hypoplasia, diameter < 1 mm). As the assessment of such variations is time-consuming and susceptible to subjectivity, robust automatic extraction of personalized CoW topology from time-of-flight magnetic resonance angiography (TOF-MRA) images would highly benefit large-scale clinical investigations. Previous work has sought to extract CoW topology from voxel-based semantic segmentation masks. However, hypoplastic arteries are challenging to recover in voxel-based segmentation. Instead, we propose using a complete CoW as an anatomical prior for extracting all possible CoW arteries as shortest paths between automatically identified anatomical landmarks, guided by automatically determined artery orientation vector fields. These fields are obtained using a scale-invariant and rotation-equivariant mesh-CNN-based model (SIRE). For a 3D TOF-MRA volume, a potentially overcomplete graph of the CoW is thus extracted in which each edge represents an artery. Subsequently, a binary Random Forest classifier labels each artery as normal or hypo-/aplastic. The model was optimized and validated using a data set of 351 3D TOF-MRA scans in a cross-validation setup. We showed that using a shortest path algorithm with a cost function based on local artery orientations results in continuous artery paths, even in hypoplastic cases. We tracked the correct path in the posterior communicating arteries in 70–74% of the cases, an artery that is known to pose challenges in voxel-based segmentation models. Our downstream artery path classifier obtained an average F1 score of 0.91, demonstrating the potential of our proposed framework to extract personalized CoW topology automatically.

Objectives Several intracranial vessel wall sequences have been described in recent literature, with either 3-T or 7-T magnetic resonance imaging (MRI). In the current study, we compared 3-T and 7-T MRI in visualising both the intracranial arterial vessel wall and vessel wall lesions. Methods Twenty-one elderly asymptomatic volunteers were scanned by 3-T and 7-T MRI with an intracranial vessel wall sequence, both before and after contrast administration. Two raters scored image quality, and presence and characteristics of vessel wall lesions. Results Vessel wall visibility was equal or significantly better at 7 T for the studied arterial segments, even though there were more artefacts hampering assessment. The better visualisation of the vessel wall at 7 T was most prominent in the proximal anterior cerebral circulation and the posterior cerebral artery. In the studied elderly asymptomatic population, 48 vessel-wall lesions were identified at 3 T, of which 7 showed enhancement. At 7 T, 79 lesions were identified, of which 29 showed enhancement. Seventy-one percent of all 3-T lesions and 59 % of all 7-T lesions were also seen at the other field strength. Conclusions Despite the large variability in detected lesions at both field strengths, we believe 7-T MRI has the highest potential to identify the total burden of intracranial vessel wall lesions. Key Points • Intracranial vessel wall visibility was equal or significantly better at 7-T MRI • Most vessel wall lesions in the cerebral arteries were found at 7-T MRI • Many intracranial vessel wall lesions showed enhancement after contrast administration • Large variability in detected intracranial vessel wall lesions at both field strengths • Seven-tesla MRI has the highest potential to identify total burden of intracranial atherosclerosis

Studies in patients with brain lesions play a fundamental role in unraveling the brain's functional anatomy. Lesion‐symptom mapping (LSM) techniques can relate lesion location to cognitive performance. However, a limitation of current LSM approaches is that they can only evaluate one cognitive outcome at a time, without considering interdependencies between different cognitive tests. To overcome this challenge, we implemented canonical correlation analysis (CCA) as combined multivariable and multioutcome LSM approach. We performed a proof‐of‐concept study on 1075 patients with acute ischemic stroke to explore whether addition of CCA to a multivariable single‐outcome LSM approach (support vector regression) could identify infarct locations associated with deficits in three well‐defined verbal memory functions (encoding, consolidation, retrieval) based on four verbal memory subscores derived from the Seoul Verbal Learning Test (immediate recall, delayed recall, recognition, learning ability). We evaluated whether CCA could extract cognitive score patterns that matched prior knowledge of these verbal memory functions, and if these patterns could be linked to more specific infarct locations than through single‐outcome LSM alone. Two of the canonical modes identified with CCA showed distinct cognitive patterns that matched prior knowledge on encoding and consolidation. In addition, CCA revealed that each canonical mode was linked to a distinct infarct pattern, while with multivariable single‐outcome LSM individual verbal memory subscores were associated with largely overlapping patterns. In conclusion, our findings demonstrate that CCA can complement single‐outcome LSM techniques to help disentangle cognitive functions and their neuroanatomical correlates.

Background Observational studies suggest that loss of skeletal muscle mass (SMM) is associated with chemotherapy‐related toxicity, poor quality of life, and poor survival in metastatic colorectal cancer (mCRC) patients. Little is known about the evolution of SMM during palliative systemic therapy. We investigated changes in SMM during various consecutive palliative systemic treatment regimens using repeated abdominal computed tomography scans of mCRC patients who participated in the randomized phase 3 CAIRO3 study. Methods In the CAIRO3 study, mCRC patients with stable disease or better after 6 cycles of first‐line treatment with capecitabine + oxaliplatin + bevacizumab (CAPOX‐B) were randomized between maintenance treatment with capecitabine + bevacizumab (CAP‐B) or observation. Upon first disease progression, in both groups, CAPOX‐B or other treatment was reintroduced until the second disease progression, which was the primary study endpoint. We analysed 1355 computed tomography scans of 450 (81%) CAIRO3 patients (64 ± 9.0 years, CAP‐B n = 223; observation n = 227) for SMM at four time points (i.e. prior to the start of pre‐randomization initial treatment, at randomization, and at first and at second disease progression) using the Slice‐o‐matic software and single slice evaluation at the lumbar 3 level. By using accepted and widely used formulas, whole body SMM was calculated. A linear mixed effects model, adjusted for relevant confounders, was used to assess SMM changes for the total group and within and between study arms. Results During 6 cycles of initial treatment with CAPOX‐B prior to randomization, SMM decreased significantly in all patients [CAP‐B arm: −0.53 kg (95% CI −1.12; −0.07) and observation arm: −0.85 kg (−1.45; −0.25)]. After randomization, SMM recovered during CAP‐B treatment by 1.32 kg (0.73; 1.90) and observation by 1.20 kg (0.63; 1.78) (median time from randomization to first disease progression 8.6 and 4.1 months for CAP‐B arm and observation arm, respectively). After first progression and during reintroduction treatment with CAPOX‐B or other treatment, SMM again decreased significantly and comparable in both arms, CAP‐B: −2.71 kg (−3.37; −2.03), and observation: −2.01 kg (−2.64; −1.41) (median time from first progression until second progression CAP‐B arm: 4.7 months and observation arm: 6.6 months). Conclusions This longitudinal study provides a unique insight in SMM changes in mCRC patients during palliative systemic treatment regimens, including observation. Our data show that muscle loss is reversible and may be influenced by the intensity of systemic regimens. Although studies have shown prognostic capacity for SMM, the effects of subsequent changes in SMM are unknown and may be clues for new future therapeutic interventions.