Many techniques are proposed for the quantification of tumor heterogeneity as an imaging biomarker for differentiation between tumor types, tumor grading, response monitoring and outcome prediction. ...However, in clinical practice these methods are barely used. This study evaluates the reported performance of the described methods and identifies barriers to their implementation in clinical practice.
The Ovid, Embase, and Cochrane Central databases were searched up to 20 September 2013. Heterogeneity analysis methods were classified into four categories, i.e., non-spatial methods (NSM), spatial grey level methods (SGLM), fractal analysis (FA) methods, and filters and transforms (F&T). The performance of the different methods was compared.
Of the 7351 potentially relevant publications, 209 were included. Of these studies, 58% reported the use of NSM, 49% SGLM, 10% FA, and 28% F&T. Differentiation between tumor types, tumor grading and/or outcome prediction was the goal in 87% of the studies. Overall, the reported area under the curve (AUC) ranged from 0.5 to 1 (median 0.87). No relation was found between the performance and the quantification methods used, or between the performance and the imaging modality. A negative correlation was found between the tumor-feature ratio and the AUC, which is presumably caused by overfitting in small datasets. Cross-validation was reported in 63% of the classification studies. Retrospective analyses were conducted in 57% of the studies without a clear description.
In a research setting, heterogeneity quantification methods can differentiate between tumor types, grade tumors, and predict outcome and monitor treatment effects. To translate these methods to clinical practice, more prospective studies are required that use external datasets for validation: these datasets should be made available to the community to facilitate the development of new and improved methods.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Artificial intelligence (AI) has the potential to significantly disrupt the way radiology will be practiced in the near future, but several issues need to be resolved before AI can be widely ...implemented in daily practice. These include the role of the different stakeholders in the development of AI for imaging, the ethical development and use of AI in healthcare, the appropriate validation of each developed AI algorithm, the development of effective data sharing mechanisms, regulatory hurdles for the clearance of AI algorithms, and the development of AI educational resources for both practicing radiologists and radiology trainees. This paper details these issues and presents possible solutions based on discussions held at the 2019 meeting of the International Society for Strategic Studies in Radiology.
Key Points
• Radiologists should be aware of the different types of bias commonly encountered in AI studies, and understand their possible effects.
• Methods for effective data sharing to train, validate, and test AI algorithms need to be developed.
• It is essential for all radiologists to gain an understanding of the basic principles, potentials, and limits of AI.
The gap between predicted brain age using magnetic resonance imaging (MRI) and chronological age may serve as a biomarker for early-stage neurodegeneration. However, owing to the lack of large ...longitudinal studies, it has been challenging to validate this link. We aimed to investigate the utility of such a gap as a risk biomarker for incident dementia using a deep learning approach for predicting brain age based on MRI-derived gray matter (GM). We built a convolutional neural network (CNN) model to predict brain age trained on 3,688 dementia-free participants of the Rotterdam Study (mean age 66 ± 11 y, 55% women). Logistic regressions and Cox proportional hazards were used to assess the association of the age gap with incident dementia, adjusted for age, sex, intracranial volume, GM volume, hippocampal volume, white matter hyperintensities, years of education, and APOE ε4 allele carriership. Additionally, we computed the attention maps, which shows which regions are important for age prediction. Logistic regression and Cox proportional hazard models showed that the age gap was significantly related to incident dementia (odds ratio OR = 1.11 and 95% confidence intervals CI = 1.05–1.16; hazard ratio HR = 1.11, and 95% CI = 1.06–1.15, respectively). Attention maps indicated that GM density around the amygdala and hippocampi primarily drove the age estimation. We showed that the gap between predicted and chronological brain age is a biomarker, complimentary to those that are known, associated with risk of dementia, and could possibly be used for early-stage dementia risk screening.
With aging, the brain undergoes several structural changes. These changes reflect the normal aging process and are therefore not necessarily pathologic. In fact, better understanding of these normal ...changes is an important cornerstone to also disentangle pathologic changes. Several studies have investigated normal brain aging, both cross-sectional and longitudinal, and focused on a broad range of magnetic resonance imaging (MRI) markers. This study aims to comprise the different aspects in brain aging, by performing a comprehensive longitudinal assessment of brain aging, providing trajectories of volumetric (global and lobar; subcortical and cortical), microstructural, and focal (presence of microbleeds, lacunar or cortical infarcts) brain imaging markers in aging and the sequence in which these markers change in aging. Trajectories were calculated on 10,755 MRI scans that were acquired between 2005 and 2016 among 5286 persons aged 45 years and older from the population-based Rotterdam Study. The average number of MRI scans per participant was 2 scans (ranging from 1 to 4 scans), with a mean interval between MRI scans of 3.3 years (ranging from 0.2 to 9.5 years) and an average follow-up time of 5.2 years (ranging from 0.3 to 9.8 years). We found that trajectories of the different volumetric, microstructural, and focal markers show nonlinear curves, with accelerating change with advancing age. We found earlier acceleration of change in global and lobar volumetric and microstructural markers in men compared with women. For subcortical and cortical volumes, results show a mix of more linear and nonlinear trajectories, either increasing, decreasing, or stable over age for the subcortical and cortical volume and thickness. Differences between men and women are visible in several parcellations; however, the direction of these differences is mixed. The presence of focal markers show a nonlinear increase with age, with men having a higher probability for cortical or lacunar infarcts. The data presented in this study provide insight into the normal aging process in the brain, and its variability.
Computer-aided diagnosis of dementia using a support vector machine (SVM) can be improved with feature selection. The relevance of individual features can be quantified from the SVM weights as a ...significance map (p-map). Although these p-maps previously showed clusters of relevant voxels in dementia-related brain regions, they have not yet been used for feature selection. Therefore, we introduce two novel feature selection methods based on p-maps using a direct approach (filter) and an iterative approach (wrapper). To evaluate these p-map feature selection methods, we compared them with methods based on the SVM weight vector directly, t-statistics, and expert knowledge. We used MRI data from the Alzheimer's disease neuroimaging initiative classifying Alzheimer's disease (AD) patients, mild cognitive impairment (MCI) patients who converted to AD (MCIc), MCI patients who did not convert to AD (MCInc), and cognitively normal controls (CN). Features for each voxel were derived from gray matter morphometry. Feature selection based on the SVM weights gave better results than t-statistics and expert knowledge. The p-map methods performed slightly better than those using the weight vector. The wrapper method scored better than the filter method. Recursive feature elimination based on the p-map improved most for AD-CN: the area under the receiver-operating-characteristic curve (AUC) significantly increased from 90.3% without feature selection to 92.0% when selecting 1.5%-3% of the features. This feature selection method also improved the other classifications: AD-MCI 0.1% improvement in AUC (not significant), MCI-CN 0.7%, and MCIc-MCInc 0.1% (not significant). Although the performance improvement due to feature selection was limited, the methods based on the p-map generally had the best performance, and were therefore better in estimating the relevance of individual features.
•In this position paper, we provide a collection of views on the role of AI in the COVID-19 pandemic, from the clinical needs to the design of AI-based systems, to the translation of the developed ...tools to the clinic.•We highlight key factors in designing system solutions - per specific task; as well as design issues in managing the disease on the national level.•We focus on three specific use-cases for which AI systems can be built: from the early disease detection, the management of the disease in a hospital setting, and building patient-specific predictive models that require the combination of imaging with additional clinical features.•Infrastructure considerations and population modeling in two European countries will be described.•This pandemic has made the practical and scientific challenges of making AI solutions very explicit. A discussion concludes this paper, with a list of challenges facing the community in the AI road ahead.
In this position paper, we provide a collection of views on the role of AI in the COVID-19 pandemic, from clinical requirements to the design of AI-based systems, to the translation of the developed tools to the clinic. We highlight key factors in designing system solutions - per specific task; as well as design issues in managing the disease at the national level. We focus on three specific use-cases for which AI systems can be built: early disease detection, management in a hospital setting, and building patient-specific predictive models that require the combination of imaging with additional clinical data. Infrastructure considerations and population modeling in two European countries will be described. This pandemic has made the practical and scientific challenges of making AI solutions very explicit. A discussion concludes this paper, with a list of challenges facing the community in the AI road ahead.
Imaging plays an essential role in research on neurological diseases in the elderly. The Rotterdam Scan Study was initiated as part of the ongoing Rotterdam Study with the aim to elucidate the causes ...of neurological disease by performing imaging of the brain in a prospective population-based setting. Initially, in 1995 and 1999, random subsamples of participants from the Rotterdam Study underwent neuroimaging, whereas from 2005 onwards MRI has been implemented into the core protocol of the Rotterdam Study. In this paper, we discuss the background and rationale of the Rotterdam Scan Study. Moreover, we describe the imaging protocol, image post-processing techniques, and the main findings to date. Finally, we provide recommendations for future research, which will also be topics of investigation in the Rotterdam Scan Study.
Structural brain markers are studied extensively in the field of neurodegeneration, but are thought to occur rather late in the process. Functional measures such as functional connectivity are ...gaining interest as potentially more subtle markers of neurodegeneration. However, brain structure and function are also affected by ‘normal’ brain ageing. More information is needed on how functional connectivity relates to aging, particularly in the absence of overt neurodegenerative disease. We investigated the association of age with resting-state functional connectivity in 2878 non-demented persons between 50 and 95 years of age (54.1% women) from the population-based Rotterdam Study. We obtained nine well-known resting state networks using data-driven methodology. Within the anterior default mode network, ventral attention network, and sensorimotor network, functional connectivity was significantly lower with older age. In contrast, functional connectivity was higher with older age within the visual network. Between resting state networks, we found patterns of both increases and decreases in connectivity in approximate equal proportions. Our results reinforce the notion that the aging brain undergoes a reorganization process, and serves as a solid basis for exploring functional connectivity as a preclinical marker of neurodegenerative disease.
•The aging brain undergoes a complex functional reorganization process.•Age is related to decreases in within-network functional connectivity and to widespread increases and decreases in (anti-)correlations between different networks.•This study forms a basis for exploring functional connectivity as a preclinical marker of neurodegenerative disease.
Objectives The purpose of this study was to investigate whether the use of a semiautomated plaque quantification algorithm (reporting volumetric and geometric plaque properties) provides additional ...prognostic value for the development of acute coronary syndromes (ACS) as compared with conventional reading from cardiac computed tomography angiography (CCTA). Background CCTA enables the visualization of coronary plaque characteristics, of which some have been shown to predict ACS. Methods A total of 1,650 patients underwent 64-slice CCTA and were followed up for ACS for a mean 26 ± 10 months. In 25 patients who had ACS and 101 random controls (selected from 993 patients with coronary artery disease but without coronary event), coronary artery disease was evaluated using conventional reading (calcium score, luminal stenosis, morphology), and then independently quantified using semiautomated software (plaque volume, burden area plaque area divided by vessel area times 100%, noncalcified percentage, attenuation, remodeling). Clinical risk profile was calculated with Framingham risk score (FRS). Results There were no significant differences in conventional reading parameters between controls and patients who had ACS. Semiautomated plaque quantification showed that compared to controls, ACS patients had higher total plaque volume (median: 94 mm3 vs. 29 mm3 ) and total noncalcified volume (28 mm3 vs. 4 mm3 , p ≤ 0.001 for both). In addition, per-plaque maximal volume (median: 56 mm3 vs. 24 mm3 ), noncalcified percentage (62% vs. 26%), and plaque burden (57% vs. 36%) in ACS patients were significantly higher (p < 0.01 for all). A receiver-operating characteristic model predicting for ACS incorporating FRS and conventional CCTA reading had an area under the curve of 0.64; a second model also incorporating semiautomated plaque quantification had an area under the curve of 0.79 (p < 0.05). Conclusions The semiautomated plaque quantification algorithm identified several parameters predictive for ACS and provided incremental prognostic value over clinical risk profile and conventional CT reading. The application of this tool may improve risk stratification in patients undergoing CCTA.
Automatic Collateral Scoring From 3D CTA Images Su, Jiahang; Wolff, Lennard; van Es, Adriaan C. G. M ...
IEEE transactions on medical imaging,
06/2020, Letnik:
39, Številka:
6
Journal Article
Odprti dostop
The collateral score is an important biomarker in decision making for endovascular treatment (EVT) of patients with ischemic stroke. The existing collateral grading systems are based on visual ...inspection and prone to subjective interpretation and interobserver variation. The purpose of our work is the development of an automatic collateral scoring method. In this work, we present a method that is inspired by human collateral scoring. Firstly, we define an anatomical region by atlas-based registration and extract vessel structures using a deep convolutional neural network. From this, high-level features based on the ratios of vessel length and volume of the occluded and the contralateral side are defined. Multi-class classification models are used to map the feature space to a four-grade collateral score and a quantitative score. The dataset used for training, validation and testing is from a registry of images acquired in clinical routine at multiple medical centers. The model performance is tested on 269 subjects, achieving an accuracy of 0.8. The dichotomized collateral score accuracy is 0.9. The error is comparable to the interobserver variation, the results are comparable to the performance of two radiologists with 10 to 30 years of experience.
PDF
