•This review focuses different aspects of deep learning applications in radiology.•This paper covers evolution of deep learning, its potentials, risk and safety issues.•This review covers some deep ...learning techniques already applied.•It gives an overall view of impact of deep learning in the medical imaging industry.
The advent of Deep Learning (DL) is poised to dramatically change the delivery of healthcare in the near future. Not only has DL profoundly affected the healthcare industry it has also influenced global businesses. Within a span of very few years, advances such as self-driving cars, robots performing jobs that are hazardous to human, and chat bots talking with human operators have proved that DL has already made large impact on our lives. The open source nature of DL and decreasing prices of computer hardware will further propel such changes. In healthcare, the potential is immense due to the need to automate the processes and evolve error free paradigms. The sheer quantum of DL publications in healthcare has surpassed other domains growing at a very fast pace, particular in radiology. It is therefore imperative for the radiologists to learn about DL and how it differs from other approaches of Artificial Intelligence (AI). The next generation of radiology will see a significant role of DL and will likely serve as the base for augmented radiology (AR). Better clinical judgement by AR will help in improving the quality of life and help in life saving decisions, while lowering healthcare costs.
A comprehensive review of DL as well as its implications upon the healthcare is presented in this review. We had analysed 150 articles of DL in healthcare domain from PubMed, Google Scholar, and IEEE EXPLORE focused in medical imagery only. We have further examined the ethic, moral and legal issues surrounding the use of DL in medical imaging.
Intracerebral hemorrhage (ICH) is a complex and heterogeneous disease, and there is no effective treatment. Spontaneous ICH represents the final manifestation of different types of cerebral small ...vessel disease, usually categorized as: lobar (mostly related to cerebral amyloid angiopathy) and nonlobar (hypertension-related vasculopathy) ICH. Accurate phenotyping aims to reflect these biological differences in the underlying mechanisms and has been demonstrated to be crucial to the success of genetic studies in this field. This review summarizes how current knowledge on genetics and epigenetics of this devastating stroke subtype are contributing to improve the understanding of ICH pathophysiology and their potential role in developing therapeutic strategies.
Little information is available about sex-related differences in intracerebral hemorrhage (ICH). This is a prospective observational study to describe the sex differences in demographics, vascular ...risk factors, stroke care, and outcomes in primary ICH.
BasicMar is a hospital-based registry of all stroke patients admitted to a single public hospital that covers a population of 330,000. From 2005 to 2015, there were 515 consecutive acute primary ICH patients. Outcome data were obtained at 3 months.
More men than women had ICH (52.4% vs 47.6%); the women were older and had worse previous functional status than men, who were more likely to drink alcohol and smoke and to have ischemic heart disease and peripheral arterial disease. There were no sex differences in etiology, severity, or hemorrhage volume. ICH score was greater in women than in men (p = 0.018). Women had more lobar ICH than men (odds ratio adjusted by age was 1.75 95% confidence interval 1.18-2.58, p = 0.005). The quality of stroke care was similar in both sexes. Mortality at 3 months was 44.1% in women and 41.1% in men (p = 0.656), and 3-month poor outcome among survivors (modified Rankin Scale mRS score 3-5) 58.4% in women and 45.3% in men (p = 0.027). After adjustment for previous mRS and ICH score, there were no differences in 3-month mortality or poor outcome at 3 months between sexes.
Patients with ICH showed sex-related differences in demographic characteristics, ICH location, and vascular risk factors, but not in stroke care, 3-month mortality, or adjusted poor outcome.
Described here is a detailed novel pilot study on whether the SYNTAX (Synergy between percutaneous coronary intervention with TAXUS and cardiac surgery) score, a measure of coronary artery disease ...complexity, could be better predicted with carotid intima-media thickness (cIMT) measures using automated IMT all along the common carotid and bulb plaque compared with manual IMT determined by sonographers. Three hundred seventy consecutive patients who underwent carotid ultrasound and coronary angiography were analyzed. SYNTAX score was determined from coronary angiograms by two experienced interventional cardiologists. Unlike most methods of cIMT measurement commonly used by sonographers, our method involves a computerized automated cIMT measurement all along the carotid artery that includes the bulb region and the region proximal to the bulb (under the class of AtheroEdge systems from AtheroPoint, Roseville, CA, USA). In this study, the correlation between automated cIMT that includes bulb plaque and SYNTAX score was found to be 0.467 (p < 0.0001), compared with 0.391 (p < 0.0001) for the correlation between the sonographer's IMT reading and SYNTAX score. The correlation between the automated cIMT and the sonographer's IMT was 0.882. When compared against the radiologist's manual tracings, automated cIMT system performance had a lumen-intima error of 0.007818 ± 0.0071 mm, media-adventitia error of 0.0179 ± 0.0125 mm and automated cIMT error of 0.0099 ± 0.00988 mm. The precision of automated cIMT against the manual radiologist's reading was 98.86%. This current automated algorithm revealed a significantly stronger correlation between cIMT and coronary SYNTAX score as compared with the sonographer's cIMT measurements with multiple cardiovascular risk factors. We benchmarked our correlation between the automated cIMT that includes bulb plaque and SYNTAX score against a previously published (Ikeda et al. 2013) AtheroEdgeLink (AtheroPoint) correlation between the automated cIMT that does not include bulb plaque and SYNTAX score and had an improvement of 44.58%. By sampling cIMT in the bulb region, the automated cIMT technique improves the degree of correlation between coronary artery disease lesion complexity and carotid atherosclerosis characteristics.
A number of environmental risk factors of acute ischemic stroke have been identified, but few studies have evaluated the influence of the outdoor environment on stroke severity. We assessed the ...association of residential ambient fine particulate matter air pollution (PM2.5), noise, and surrounding greenspace with initial stroke severity.
We obtained data on patients hospitalized with acute ischemic stroke from a hospital-based prospective stroke register (2005–2014) in Barcelona. We estimated residential PM2.5 based on an established land use regression model, greenspace as the average satellite-based Normalized Difference Vegetation Index (NDVI) within a 300 m buffer of the residence, and daily (Lday), evening (Levening), night (Lnight) and average noise (Lden) level at the street nearest to the residential address using municipal noise models. Stroke severity was assessed at the time of hospital presentation using the National Institute of Health Stroke Scale (NIHSS).We used logistic regression and binomial models to evaluate the associations of PM2.5, greenspace, and noise with initial stroke severity adjusting for potential confounders.
Among 2761 patients, higher residential surrounding greenspace was associated with lower risk of severe stroke (OR for NIHSS>5, 0.75; 95% CI: 0.60–0.95), while, living in areas with higher Lden was associated with a higher risk of severe stroke (OR, 1.30; 95% CI: 1.02–1.65). PM2.5 was not associated with initial stroke severity.
In an urban setting, surrounding greenspace and traffic noise at home are associated with initial stroke severity, suggesting an important influence of the built environment on the global burden of ischemic stroke.
•Higher residential surrounding greenspace is related with less severe stroke.•Living in areas with higher annual average noise values is associated with a more severe stroke.•There is an influence of the built environment on the global burden of stroke.
Age and stroke severity are the main mortality predictors after ischemic stroke. However, chronological age and biological age are not exactly concordant. Age-related changes in DNA methylation in ...multiple CpG sites across the genome can be used to estimate biological age, which is influenced by lifestyle, environmental factors, and genetic variation. We analyzed the impact of biological age on 3-month mortality in ischemic stroke. We assessed 594 patients with acute ischemic stroke in a cohort from Hospital del Mar (Barcelona) and validated the results in an independent cohort. Demographic and clinical data, including chronological age, vascular risk factors, initial stroke severity (NIHSS score), recanalization treatment, and previous modified Rankin scale were registered. Biological age was estimated with an algorithm based on DNA methylation in 71 CpGs. Biological age was predictive of 3-month mortality (p = 0.041; OR = 1.05, 95% CI 1.00-1.10), independently of NIHSS score, chronological age, TOAST, vascular risk factors, and blood cell composition. Stratified by TOAST classification, biological age was associated with mortality only in large-artery atherosclerosis etiology (p = 0.004; OR = 1.14, 95% CI 1.04-1.25). As estimated by DNA methylation, biological age is an independent predictor of 3-month mortality in ischemic stroke regardless of chronological age, NIHSS, previous modified Rankin scale, and vascular risk factors.
Background
Stroke recurrence (SR) after an ischemic stroke is an important cause of death and disability. We conducted a hospital-based study to evaluate the role of biological age (b-Age: ...age-related DNA-methylation changes) as a risk factor for SR.
Methods
We included 587 patients in the acute phase of stroke, assessed at one tertiary stroke center (Hospital del Mar: Barcelona, Spain). B-Age was estimated with 5 different methods based on DNA methylation, and Hannum’s method was the one that better performed. We analyzed the relationships between b-Age, chronological age, sex, vascular risk factors, coronary and peripheral arterial disease, atrial fibrillation, initial neurological severity assessed by National Institutes of Health Stroke Scale (NIHSS), transient ischemic attack (TIA) in the 7 days preceding the index stroke, and symptomatic atherosclerosis. Stroke recurrence definition include: new symptoms that suggest a new ischemic event had occurred within 3 months after stroke onset and worsening by four points in the initial neurological severity (measured by National Institutes of Health Stroke Scale (NIHSS) score).
Results
Logistic regression analysis associated b-Age with SR
p
= 0.003; OR = 1.06 (95% CI: 1.02–1.09), independently of chronological age
p
= 0.022; OR = 0.96 (95% CI 0.94–1.00), symptomatic atherosclerosis (stenosis > 50% in the symptomatic territory), transient ischemic attack (TIA) in the 7 days preceding the index stroke, and initial NIHSS. The b-Age of patients with SR was 2.7 years older than patients without SR.
Conclusions
Patients with SR were biologically older than those without SR. B-Age was independently associated with high risk of developing SR.
Objectives
The spectrum of distribution of white matter hyperintensities (WMH) may reflect different functional, histopathological, and etiological features. We examined the relationships between ...cerebrovascular risk factors (CVRF) and different patterns of WMH in MRI using a qualitative visual scale in ischemic stroke (IS) patients.
Methods
We assembled clinical data and imaging findings from patients of two independent cohorts with recent IS. MRI scans were evaluated using a modified visual scale from
Fazekas
,
Wahlund
, and
Van Swieten
. WMH distributions were analyzed separately in periventricular (PV-WMH) and deep (D-WMH) white matter, basal ganglia (BG-WMH), and brainstem (B-WMH). Presence of confluence of PV-WMH and D-WMH and anterior-versus-posterior WMH predominance were also evaluated. Statistical analysis was performed with SPSS software.
Results
We included 618 patients, with a mean age of 72 years (standard deviation SD 11 years). The most frequent WMH pattern was D-WMH (73%). In a multivariable analysis, hypertension was associated with PV-WMH (odds ratio OR 1.79, 95% confidence interval CI 1.29–2.50,
p
= 0.001) and BG-WMH (OR 2.13, 95% CI 1.19–3.83,
p
= 0.012). Diabetes mellitus was significantly related to PV-WMH (OR 1.69, 95% CI 1.24–2.30,
p
= 0.001), D-WMH (OR 1.46, 95% CI 1.07–1.49,
p
= 0.017), and confluence patterns of D-WMH and PV-WMH (OR 1.62, 95% CI 1.07–2.47,
p
= 0.024). Hyperlipidemia was found to be independently related to brainstem distribution (OR 1.70, 95% CI 1.08–2.69,
p
= 0.022).
Conclusions
Different CVRF profiles were significantly related to specific WMH spatial distribution patterns in a large IS cohort.
Key Points
• An observational study of WMH in a large IS cohort was assessed by a modified visual evaluation.
• Different CVRF profiles were significantly related to specific WMH spatial distribution patterns.
• Distinct WMH anatomical patterns could be related to different pathophysiological mechanisms.
Objectives
Coronary calcification plays an important role in diagnostic classification of lesion subsets. According to histopathologic studies, vulnerable atherosclerotic plaque contains calcified ...deposits, and there can be considerable variation in the extent and degree of calcification. Intravascular ultrasound (IVUS) has demonstrated its role in imaging coronary arteries, thereby displaying calcium lesions. The aim of this work was to develop a fully automated system for detection, area and volume measurement, and characterization of the largest calcium deposits in coronary arteries. Furthermore, we demonstrate the correlation between the coronary calcium IVUS volume and the neurologic risk biomarker B‐mode carotid intima‐media thickness (IMT).
Methods
Our system automatically detects the frames with calcium, identifies the largest calcium region, and performs shape‐based volume measurements. The carotid IMT is measured by using AtheroEdge software (AtheroPoint, LLC) on B‐mode ultrasound imaging.
Results
Our database consists of low‐contrast IVUS videos and corresponding B‐mode images from 100 patients. Our experiments showed that the correlation between calcium volumes and carotid IMT was higher for the left carotid artery compared to the right carotid artery (r = 0.066 for the left carotid artery and 0.121 for the right carotid artery). We obtained 97% accuracy for automated calcium detection compared against the scoring given by our expert radiologists. Furthermore, we benchmarked shape‐based volume measurement against the conventional method, which used integration of regions and showed a correlation of 84%.
Conclusions
Since carotid IMT is an independent prognostic factor for myocardial infarction, and calcium lesions are correlated with stroke risk, we believe that this automated system for calcium volume measurement could be useful for assessing patients' cardiovascular risk.