Although not commonplace, carotid artery stenting (CAS) is occasionally accompanied by the protrusion of plaque into the stent lumen, and a variety of ischemic complications during intra- and post-operative periods. Among these complications, plaque protrusion (PP) into the stent and thrombus on the stent after CAS are some of the most worrying for clinicians.
The correction of PP is achieved by additional post-dilations or stent-in-stent implantation.

First noticed in mid-1990s
The first observations of PP date over two decades and provided the impetus for embolic protection devices. In a paper published in the mid-1990s, a team led by Frank Veith, M.D (from the Mayo Clinic) suggested that restoration of flow and removal of protection devices might lead to the continuing break off of protrusions and provoke some delayed strokes.
Today, apart from PP, key causes of late in-stent stenosis seen in CAS include neo-intimal proliferation due to self-expanding stents as well as restricted post-procedural stent dimensions from inefficient balloon dilation.

The covered stent
One way to prevent PP is by covering the plaque with a stent graft. Covered stents also offer advantage in usability, without the need for distal protection devices in difficult internal carotid arteries, where a protection filter may be near-impossible to place. However, covered stents are accompanied by a high rate of restenosis. In one randomized trial in the mid-2000s, researchers at Vienna Medical University reported a 38% restenosis rate in patients treated with covered stents for carotid artery stenosis, while restenosis was absent in the bare stent group. The precise causes of high restenosis rates with stent grafts require further research. One suspected factor is the buckling of a covered stent’s proximal and distal ends and the prevention of endothelization.

Both symptomatic and asymptomatic events
During stent implantation, plaque disruption and distal migration of plaque particles may cause symptomatic or asymptomatic ischemic events, in spite of protection devices. These can be viewed on diffusion-weighted (DW) magnetic resonance imaging (MRI).
In some CAS cases, physicians encounter plaque particles filling filters leading to symptomatic cerebral embolism. This is particularly true with ulcerated plaque and severe stenosis.  What is now of growing concern is that, after stent implantation, plaque protrusion into the lumen can lead to peri-procedural stent thrombosis, 30-day stroke, and late in-stent stenosis.

Stroke biggest complication
The biggest complication with carotid artery stenting is stroke. This can occur during CAS and for up to 30 days after the procedure. Although the cause of late stroke after CAS remains unknown, PP is generally suspected to be a key cause.
PP incidence is evaluated by IVUS (intra-venous ultrasound) and angiography, although as we shall see, there are variations in results based on methodology. The prognosis of PP (over a 30-day period) and the incidence of ischemic lesions (48 hours after CAS) are usually assessed by diffusion-weighted images.

The Tokai study
In recent years, one oft-cited report concerns a study by the Department of Cardiology at Tokai University School of Medicine (Isehara). The findings were published in October 2014 by the ‘Journal of Stroke and Cerebrovascular Disorders’. During their study, the Tokai researchers evaluated 77 CAS procedures, which were performed consecutively with IVUS between May 2008 and December 2012. All cases were distally protected with filter devices. The rate of PP was assessed at the end of each procedure using IVUS and angiography.
Six plaque protrusions (7.8%) through the stent struts were detected by IVUS but only two (2.6%) by angiography. One of the major predictors of PP was pre-procedural severe stenosis with flow delay.   Overall stroke rate was 2.6% (major 0%, minor 2.6%), and these occurred in the catheterization laboratory. However, no late stroke was observed at 30 days after procedure.
One of the key outcomes of the Tokai study was that IVUS seems to detect plaque protrusion better than angiography. Since the adequate management of plaque protrusion is considered as a means to reduce stroke complications, IVUS usage is worth considering.

The Yao-Nara study
In 2017, results of another, broader Japanese effort by researchers at Ishinkai Yao General Hospital (Yao) and Nara Medical University (Nara) suggested different conclusions. The study, which was published in the April 17 issue of ‘JACC: Cardiovascular Interventions’, sought to clarify the frequency and prognosis of plaque protrusions in CAS by analysing data on 328 patients treated under IVUS guidance in the period 2007-2016, using different types of stents and embolic protection devices.
At 30 days, the rate of ipsilateral ischemic stroke was 2.8% and the rate of transient ischemic attack was 2.6%. There were no patient deaths. Moreover, in most stroke cases, symptoms were observed immediately after dilatation.  New ischemic lesions were found in 35.7% of patients within 48 hours of the procedure, based on diffusion-weighted imaging (DWI).

Lack in lesion variation, but stent type matters, as does evaluation method
One of the most intriguing conclusions was the lack of difference in the incidence of new ischemic lesions, in terms of stable versus unstable plaques. Analysis by stent type, however, did indicate difference. There were more ipsilateral ischemic lesions with open-cell stents as compared to closed-cell stents.
The authors suggest the findings indicate a necessity to minimize PP “to prevent periprocedural ischemic stroke” and that the placement of open-cell stents with high radial force may disintegrate unstable plaque, causing protrusions. One strategy mentioned by the authors to manage PP is to perform IVUS to check for large-volume protrusions. The latter are then sought to be differentiated as being either ‘convex’ or ‘non-convex’. For the former, stent-in-stent placement is performed using closed-cell stents until the disappearance of the protrusion. In the case of ‘nonconvex’ protrusions, the authors recommend 5-10 minutes of observation, followed again by stent-in-stent placement should the protrusion enlarge, or clinical follow-up within 30 days after CAS in case of no enlargement.

As we observed previously, there are differences in PP incidence based on whether it is evaluated by angiography or IVUS. One of the most significant limitations of the Japanese study above was the occurrence of 27 cases of plaque protrusion on IVUS, but just nine cases on angiography. The study protocol required confirmation by both modalities.

Limitations to Yao-Nara study
In an editorial accompanying the study, William A. Gray, MD (Lankenau Heart Institute, Wynnewood, PA), cautioned that this two thirds difference “will clearly affect many of the subsequent associations and conclusions.” Gray also underlined that by treating plaque protrusion with stent-in-stent placement in approximately half of the cases, the researchers might have potentially changed the clinical and imaging outcomes. Furthermore, he cautioned, the study was not core-lab controlled, with no routine use of MRI before and after procedures, and that the assessors were not blinded. Finally, they did not mandate use of specific stents or perform independent neurological assessment of clinical outcomes.
As a result, the association between stent type and plaque protrusion is ‘likely’. However, it may not be as strong as the authors contend.

Such shortcomings are likely to be addressed when the Japanese effort is paired with emerging data showing reductions in both plaque protrusion and ischemic lesions via the use of mesh-covered stents. Gray agrees that this is strengthening the case for “improvements in stent design.” Indeed, emerging micromesh stent designs are expected to contribute greatly to prevention of plaque protrusion and may become a new standard for CAS.

SCAFFOLD trial
In the United States, the SCAFFOLD trial, led by Peter A. Schneider, MD, of Kaiser Foundation Hospital at Honolulu (Hawaii) is completing evaluation of a mesh-covered, open-cell heparin coated stent in patients at high surgical risk. The objectives are to make the first 30 days safer, with the understanding that reduced cell size equates to less plaque prolapse and fewer delayed events.
Other similar trials using different mesh technologies are also under way, and more are imminent. In Italy, for instance, University of Roma La Sapienza has begun a positive-control study to analyze and compare the rate of off-table subclinical neurological events in two groups of patients submitted to CAS with a close-cell stent, and a new mesh-covered carotid stent called C-Guard.
Overall, new parameters are coming into place, via stents with differences in pore size, flexibility etc. The drivers for such efforts range from new materials to a broad range of cardiovascular conditions. In France, for example, University Hospital Grenoble is conducting trials with Mguard, a stainless-steel closed cell stent covered with an ultra-thin polymer mesh sleeve, to prevent distal embolization during percutaneous coronary intervention in ST-segment-elevation myocardial infarction.

Case selection and stenting success
One of the observations of SCAFFOLD was that case selection for stenting was a key “to good clinical results.”
So far, patient selection criteria for CAS is largely based on surgical risk related to other co-morbidities. The morphology of the atherosclerotic plaque is given little attention, although studies have demonstrated the existence of extensive variability, which in turn confers specific risks for plaque vulnerability. Overall, the detection of unstable plaque on MR plaque imaging and the use of open cell stent are considered to be significant predictive factors of PP.

In recent years, there have been growing calls for devising best practices in peri-procedural management and follow-up, and for continuous feedback from clinicians to industry to improve stent design.
In general, achieving better outcomes of CAS is seen as the best method to solidify its place as a frontline treatment of carotid vascular disease. 

One promising approach for patient selection and identification of plaque, has been the use of virtual histology intravascular ultrasound imaging (VH IVUS). Researchers have suggested a strong correlation between VH IVUS plaque characterization and the true histological examination of plaque following endarterectomy, especially in ‘vulnerable’ plaque types.
The results of one of the earliest efforts in this area were published in October 2007 in ‘The Journal of Endovascular Therapy’. This followed a prospective, two-arm study by the Arizona Heart Hospital & Translational Research Center. The researchers enrolled 30 patients.
In the first arm of the study, 15 patients underwent VH IVUS examination of carotid plaque with a cerebral protection device. This was immediately followed by carotid endarterectomy (CEA), and the comparison of ‘virtual’ with true histology (classifying plaque type by VH IVUS and histopathology in a blinded study).
In the second arm, 15 patients undergoing CAS had a preliminary VH IVUS scan performed with cerebral protection. Debris collected from the filter following stenting was examined histologically and compared with the VH IVUS data.
The diagnostic accuracy of VH IVUS to agree with true histology in different carotid plaque types was 99.4% in thin-cap fibroatheroma, 96.1% for calcified thin-cap fibroatheroma, 85.9% in fibroatheroma, 85.5% for fibrocalcific, 83.4% in pathological intimal thickening, and 72.4% for calcified fibroatheroma.

By Callan Emery, Editor
A study published in Nature in September has caught the attention of the media and the interest of
Obs-Gyn specialists. In what is the largest study of the neonatal microbiome (gut bacteria), the researchers provide strong evidence that the way a baby is born impacts significantly on their microbiome.
The study by Lawley T., et al. (doi: 10.1038/s41586-019-1560-1) found that babies born through the vaginal canal carry different microbes from those delivered through caesarean section. Those born through c-section tended to lack strains of gut bacteria found in healthy children and adults. Additionally babies born through c-section showed a high-level of colonization by opportunistic pathogens associated with the hospital environment (including Enterococcus, Enterobacter and Klebsiella species).
Interestingly, the researchers note that it was the mother’s gut bacteria, and not vaginal bacteria, that made up much of the microbiome in the vaginally delivered babies. Previous studies had suggested that vaginal bacteria were swallowed by the baby on its way down the birth canal. This led to what is has been termed ‘vaginal seeding’ whereby babies born by c-section are swabbed with the mothers vaginal fluids in an effort to restore any missing microbes. However, a study by Stinson et al. (doi: 10.3389/fmed.2018.00135) has shown vaginal seeding to be unjustified and potentially unsafe.
Although a lack of exposure to the right microbes in early childhood has been implicated in autoimmune diseases, such as asthma, allergies and diabetes, the exact role of the baby’s gut bacteria is unclear and it isn’t known if these differences at birth will have any effect on later health.
The researchers, who analysed nearly 600 births in the United Kingdom, say the differences in gut bacteria between vaginally born and caesarean delivered babies largely evened out by 1 year old. They note that large follow-up studies are needed to determine if the early differences influence health outcomes.
Discussing her study, Stinson pointed out that microbes thrown out of balance in babies born by c-section are very similar to those thrown off balance in babies born to mothers receiving antibiotics but delivering vaginally. She surmises that routine antibiotic administration given to mothers delivering by c-section could be a cause of the bacterial difference in the neonatal microbiome.
Although this research does pose interesting questions about the potential health outcomes associated with c-section versus vaginal delivery, it should be emphasised that at this point mothers should not be deterred from c-section delivery if it is the right choice for the mother and her baby.

The study is part of larger effort, called the Baby Biome Study, which aims to follow thousands more newborns into childhood.

It is fascinating following our expanding knowledge of the workings of the brain with the use of functional MRI over the past 10-15 years. fMRI has provided an extraordinary view of brain function and enabled a wide range of remarkable discoveries. As this research proliferates, it promises many more new insights, with a multitude of applications.
Particularly interesting has been the growing understanding of memory formation and retrieval. Expanding on this knowledge and taking it to the next level, a recent study by neuroscientists and artificial intelligence researchers at DeepMind, Otto von Guericke University Magdeburg and the German Centre for Neurodegenerative Diseases shows how the human brain connects individual – or episodic – memories to solve problems and draw new insights.
The researchers proposed a novel brain mechanism that would allow retrieved memories to trigger the retrieval of other, related memories.
There have been many studies of episodic memories which advance the theory that they are stored as separate memory traces in a brain region called the hippocampus. Taking this as standard knowledge, the researchers’ new theory explores an anatomical connection that loops out of the hippocampus to the neighbouring entorhinal cortex but then passes back in to the hippocampus. It is this recurrent connection, the researchers thought, that allows memories retrieved from the hippocampus to trigger the retrieval of further, multiple linked memories.
To test the theory the researchers used a 7 Tesla fMRI to scan brain activation in 26 male and female study participants as they performed a task that required them to draw insights across separate events using a series of paired images. Their results are published in the September 2018 issue of Neuron.
Part of the study involved the development of a technique where they were able to separate out the parts of the entorhinal cortex that provide the input to the hippocampus, which allowed them to precisely measure the patterns of activation in the hippocampus to distinguish input and output separately.
Their resulting data showed that when the hippocampus retrieves a memory, it doesn’t simply pass it to the rest of the brain, but instead recirculates the activation back into the hippocampus, triggering the retrieval of other related memories.
They say their results preserve the best of both worlds – you preserve the ability to remember individual episodic experiences by keeping them separate, while at the same time allowing related memories to be combined on the fly at the point of retrieval.
In addition, they reckon this understanding could be replicated in Artificial Intelligence systems so they will have a greater capacity for rapidly solving novel problems.
What’s next?

In February last year, the US Food and Drug Administration (FDA) cleared the first medical device which uses artificial intelligence (AI) to provide clinical decision support for stroke. The Viz.AI Contact application uses an AI algorithm to identify a suspected stroke and notifies a specialist more quickly than was previously possible. Faster treatment, in turn, lessens the extent of a stroke or its progression. Subsequent FDA clearances and a recent decision to formalize regulations for such evaluations are likely to stimulate further innovation and acceptance of AI devices.

Saving time
Viz.AI Contact analyses CT images of the brain and sends a text notification by smartphone or tablet to a vascular neurologist or a neuro-interventional specialist, should a large vessel occlusion (LVO) be suspected. The algorithm automatically notifies the specialist at the same time that a review of the images is being conducted by a first-line provider. This is faster than the usual standard of care where patients wait for a radiologist to firstly review CT images and then notify a neurovascular specialist.

Retrospective study and real world data
Viz.AI, Inc., which developed the Contact application, submitted a retrospective study of 300 CT scans. This compared the performance of the image analysis algorithm and notification functionality against two trained neuro-radiologists.
Real-world evidence from a clinical study demonstrated quicker notification of a neurovascular specialist, in cases where blockage of a large vessel in the brain was suspected. In more than 95 percent of cases, the automatic notification was faster, saving an average of 52 minutes (with a range of between 6 and 206 minutes).

De Novo premarket review
The Viz.AI application was reviewed by the FDA through its De Novo premarket review process, a regulatory pathway for new types of medical devices that carry low to moderate risk, but lack a legally marketed predicate device to base a determination of equivalence. The FDA action creates a new regulatory classification, allowing other devices with the same medical imaging intended to obtain marketing authorization by 510(k) notification. One of the first areas to benefit from Viz.Ai will be AI or computer-aided triage devices, whose potential in fields such as emergency medicine is likely to be vast. Viz.AI, Inc., itself is developing Viz ICH, which uses AI to automatically detect intra-cerebral hemorrhages and triage the patient directly to the neurosurgeon on call.

Decision support for breast cancer screening
Nine months after FDA approval of Viz.AI, at the 2018 Radiological Society of North America (RSNA) annual meeting in November, Siemens Healthineers showcased the AI-based features of syngo.Breast Care, a mammography solution. syngo.Breast Care aims to provide interactive decision support for breast cancer screening.
Transpira, Siemens’ mammography reading software, is based on deep learning techniques, with training provided via over 1 million images. As a result, syngo.Breast Care’s AI-based algorithms evaluate and interpret individual lesions as well as 2-D mammograms and 3-D tomosynthesis. The system also sorts and scores cases on a 10-point scale, based on radiologist preferences of risk factors such as lesions, micro-calcifications and other abnormalities.
Siemens Healthineers aims to integrate interactive decision support into syngo.Breast Care, and reduce radiologists’ workload for the interpretation of mammograms. This has become especially challenging, given rapid growth in the use of techniques such as 3-D breast tomosynthesis.

Small firms also in play
Smaller firms have also targeted this area. ICAD’s ProFound AI, for example, also leverages AI to detect cancer in breast tomosynthesis. The software, which was FDA cleared less than a month after syngo.Breast Care was unveiled, examines every image in a tomosynthesis scan, detects malignant soft tissue densities and calcifications.
Profound AI estimates a ‘Certainty of Finding’ for each detection and, like the classification system in syngo.Breast Care, assigns Case Scores to each case to represent confidence that a detection or case is malignant. The scores are represented on a scale from 0 to 100 percent, with higher scores indicate high confidence levels in malignancy. This, in turn, is expected to improve detection, lead to fewer patient recalls and save mammographers time in reading images. This makes it geared toward screening, although it can evidently be used for diagnostic studies.

AI at inflection point
The above examples demonstrate that the use of AI is now close to an inflection point in terms of clinical decision support tools. These will provide physicians usable interactive and dynamic pathways which move beyond decision support to true evidence-based decision making, along with personalized care recommendations.
To many experts, AI seems to have been the missing link for tools that assist radiologists in improving appropriateness of follow-up recommendations for incidental findings, and thereby to enhance adherence to guidelines available at point of care. One of the consequences of such AI-assisted tools will be to reduce the variability in follow-up recommendations, as well as unnecessary imaging studies.

Diagnosis and decision support versus analysis and detection
Maximum attention to AI in imaging is currently on diagnosis and decision support. AI in areas such as quantitative analysis and assisted detection can be considered a spin-off from automation, which has been around for a longer period of time, but reinforced more recently by machine learning.
Automated quantification tools are now sufficiently mature and routinely accepted in the market. AI algorithms are used to make measurements from imaging exams and perform calculations which were previously manual and time-consuming. AI-driven quantitative analysis tools also are being used in data analytics for data mining electronic medical records, billing systems, patient scheduling and even in stand-alone scanners. Mined data range from radiation dose used by particular technologists for specific protocols to predictive analytics that pinpoint spikes in demand by day and time, and schedule back-up staff in the radiology department.
By contrast, the application of AI (and even automation) in medical fields such as computer-aided diagnosis and clinical decision support is very recent, and is likely to be some time before they become commonplace. The principal focus on AI use for image diagnosis is where timing is crucial – such as a heart attack or stroke (e.g. Viz.AI Contact). Closely related areas include tools to reduce review time for complex exams, and help triage patients needing more immediate care or other kinds of back-up.

Other new AI imaging applications
One exciting new entrant into AI in imaging is IcoMetrix, from Belgium’s IcoBrain. This FDA-cleared algorithm analyses CT scans to characterize traumatic brain injury, using deep learning to quantify the severity of such typically qualitative indicators of brain injury as hyperdense volumes, compression of the basal cisterns and midline brain shift.
Another FDA-cleared device is Cardio AIMR, which analyses MR images for cardiovascular blood flow. Its developer, Arterys, also has other AI tools to measure and track liver lesions and lung nodules, accelerate display of medical images, and interface with the common desktop Google Chrome browser to display mammograms.

The challenge of integration
Although the FDA is clearing the way for follow-on AI products, there are concerns that the process is constrained to highly specific medical imaging diagnostic reviews. Some radiologists are questioning the viability of new AI software systems, if they require scores of different contracts and integration into a hospital or enterprise imaging system – which would be a problem not only for hospital IT departments but also for legal review.
One of the ways forward is by reconfiguring approaches to enterprise imaging by streamlining workflow. Some vendors are developing bridges between different AI applications. One of the immediate goals is to have AI imaging dovetail into picture archive and communication systems (PACS) as well as vendor neutral archives. For example, Viz.ai software is designed to receive DICOM images directly from any CT scanner to a local virtual machine (VM) behind a network’s firewall.

Major firms nurture start-ups
Leading healthcare technology vendors are also starting to actively partner with smaller companies to provide a combination of in-house and third-party apps via a web-based AI app store platform. One good example of this is Siemens’ Digital Ecosystem, which offers an online menu of apps from Siemens and its partner, including some offering AI-enabled technology. Similar AI app store initiatives are also being taken by other vendors.
At RSNA 2018, where Siemens showcased syngo.Breast Care, IBM Watson said it would begin to partner with AI vendors to offer products on its new AI Marketplace, by offering standardized application programming interfaces (API) for building or integrating third party software and making it available through the IBM Cloud. Smaller vendors have seized such opportunities. French imaging agent vendor Guerbet, for instance, is working with IBM Watson Health to develop AI software to support liver cancer diagnosis and care.
IBM had initially planned to develop and launch its own AI solutions across the healthcare spectrum. However, it had to cope not only with delays in commercializing its own AI products, but small and nimbler start-ups, such as viz.AI getting ahead in obtaining FDA clearance. The biggest setback was MD Anderson ending its partnership on cancer imaging with IBM.
Other major players are also treading similar paths. GE Healthcare’s Edison platform is designed to help accelerate the development and adoption of AI and other new technologies, with clinical partners using Edison to develop and test algorithms and mate them to Edison applications and smart devices. On its part, at RSNA 2018, Philips Healthcare also launched its IntelliSpace Discovery 3.0 visualization and analysis platform to prepare patient data to train and validate deep learning algorithms. The platform is designed specifically to support imaging research.

FDA to formalize De Novo rules
Developments in AI-enabled clinical decision support, like broader AI healthcare applications, are likely to pick up after the FDA decided to formally establish regulations for the De Novo classification process in December 2018. Although the De Novo process is part of the Food and Drug Administration Modernization Act, the FDA Safety Innovation Act and the 21st Century Cures Act, it is currently not covered by any specific regulations. If finalized, the proposed rules are intended to provide clarity and transparency on the De Novo classification process.