On SPIN
By Tom Kennedy
Brain seizures occur in five per cent of all babies born alive. In the vast majority of these cases, the seizures go undetected, the babies in which they occur do not change expression and show no obvious outward signs of distress, and there is no medical intervention. This lack of detection and intervention means that lasting damage can occur to developing little brains, leading to serious problems down the line for children such as cerebral palsy, epilepsy and learning difficulties. But a researcher in Cork has found a way to detect the previously undetectable seizures and facilitate early treatment that can prevent serious brain injury in babies. In Ireland it is known that 300 babies per year are affected by significant brain injury and the incidence of unseen seizures could be higher even than many assume.
Up to 300 newborns in Ireland each year suffer significant brain injury. Many of these, researchers now believe, may be the result of undetected brain seizures. Now, however, researchers in Cork have come up with technology that facilitates the detection of over 90 per cent of such seizures. This early detection means that treatment can be introduced quickly and prevent serious mental injury occuring during a critical time in the baby's development.
We are born helpless. A gasp of air, a cry, and then back to sleep. Nothing apparently to disturb the crinkly face, yet this is a time of extremely rapid internal development. In just a few short hours, as senses begin to tune in to a new and strange environment, connections that will serve us for life, are being laid down. A baby might sense light, but it cannot yet see, sounds do not yet mean anything, and apart from that primeval cry, there is no expression.
Fortunately, that look of peaceful contentment usually shows that all is progressing well, but in an alarming number of cases, the face can turn out to be just a mask. A new born baby can undergo a seizure, lasting up to several hours, without any outward change in expression, and because of this, the most attentive and caring nursing staff may have no reason to suspect that anything has gone wrong.
Some babies are more at risk, such as those that have had a difficult birth, but as Dr Geraldine Boylan, who heads the Neonatal Brain Research Group at University College Cork, points out, seizures may actually occur in five per cent of all live births. In the vast majority of cases, 85 per cent, no one notices that a seizure is in progress, and as Dr Boylan explains, the failure to intervene quickly can have very serious consequences. The silent injuries occur at a critical time in brain development, and the evidence suggests that many of the mental problems that afflict adults, including cerebral palsy, epilepsy and learning difficulties, originate in undetected seizures that occurred shortly after birth.
In Ireland it is known that about 300 babies a year are affected by significant brain injuries, and according to Dr Boylan, the problems caused by unseen seizures could be much higher than we generally assume. "There is a real urgency in addressing this problem," she remarked, adding that it is something that continues to happen 24 hours a day, seven days a week.
Detection of a seizure is not actually all that difficult, at least not in technical terms. With EEG (electroencephalography) monitoring seizures show up immediately as erratic pulses on the graph, but, while this type of supervision is now common in intensive care units, nursing staff in general do not have access to this type of equipment, and even if they had, they would lack the expertise required to interpret the results.
"Interpretation of the recordings," explained Dr Boylan, "requires examination by a neurophysiologist who is also an expert in neonatal EEG analysis." The EEG pattern of an adult or a child is a lot different from that of a newborn baby. In a baby, patterns change from week to week, so it can take an expert to make sense of the graphs. Such expertise is quite thin on the ground, and, of course, a seizure can occur at any time of the day or night.
Fast response is critical, and, again, like detection, intervention does not necessarily require complex or expensive treatment. Lack of oxygen, often caused by constriction of the blood supply during birth, can cause serious problems, but if these are detected in time, all that may need to be done is to cool the baby down by three degrees Celsius. "This can be done with a cooling cap or jacket," said Dr Boylan, "but it has to be done within six hours of delivery."
Dr Boylan could see that while solutions to the problems exist, little could be done until the technology becomes easy enough for non-experts to use and cheap enough to deploy everywhere.
The Neonatal Unit at Cork University Hospital, with 800 newborns a year, provided the ideal setting to conduct clinical research, but coming up with workable solutions involved combining medical, engineering and software disciplines. Drawing the right kind experts together was not that difficult, and not just because the objectives were so obviously worthwhile. As Dr Boylan observed, one of the advantages that Ireland has over many other countries is that different specialists are quite happy to work together as a team. In 2003 this team began working as the multidisciplinary Neonatal Brain Research Group, and their progress since then has been spectacular.
One of the basic problems they faced concerned interpretation, and on this it became clear that two significant improvements could be made. The EEG equipment itself could be simplified, and instead of trying to upgrade interpretative skills in neonatal units, the graphs could be transmitted directly to the expert. "Several intensive units could share access to the one diagnostic service," said Dr Boylan. However, high bandwidth links between hospitals could only go so far in supporting this exchange, so the group is now working with Professor John Morrison in Computer Science and the Boole Centre for Research in Informatics to develop a system that would allow patient data to be streamed from the bedside to a remote expert.
Another improvement that the group made was to slim down the data, so that the essential pattern could be picked out instantly by intelligent software. "Being able to see a pattern is something that people only get with experience," said Dr Boylan, so her first Science Foundation Grant in 2006 went towards working with Dr Liam Marnane at UCC's Electrical Engineering Department on developing an algorithm to automatically detect seizures.
Not an easy task, for, until this research got underway, there was not enough known on how babies with problems compared to normal patterns, or indeed how signals could be related to associated factors, such as heartbeat. Looking at normal patterns is important. "You can't interpret abnormal signs without knowing what the normal ones are like," said Dr Boylan. The full 12 channel EEG, said Dr Boylan, produces a fairly complex pattern. "We needed to factor in other things that might give us a better idea of what was going on." Dr Boylan said that this research has produced excellent results in filtering out all the unwanted false signals, such as those produced by crying or physical movements. "We have a really good algorithm now," she remarked.
Ease of use is essential, and a compact low cost cotside monitor has now been produced that will, undoubtedly, bring about an enormous improvement in the care of the newborn throughout the world. "Currently this detects 93 per cent of all seizures," said Dr Boylan, a huge advance on the 53 per cent possible until then using more limited technology.The group has attracted a lot of support, an industry partner is planning to launch the monitor onto the world market, and a European project is underway to evaluate its performance. In this project the slimmed down monitors are being operated in parallel with full EEG
In many ways the beneficial spin offs from this research are likely to continue, and in the most recent development, Dr Boylan began leading a European wide project to look at the drugs being used to treat seizures in the newborn. The drugs in current use have not changed in half a century, and with medicines in general, clinical trials usually only involve adults. Better EEG monitoring means that the impact of drugs on young babies can be assessed properly and with greater accuracy, and new possibilities explored.
One such candidate is a drug called Bumetanide, widely used over the past thirty years as a diuretic in adults and children. Recent research, said Dr Boylan, suggests that this drug could be effective in treatment of seizures in newborn babies. However, up to now, ethical considerations ruled out the use of such potentially effective drugs because clinical studies had not been conducted. Because of the advances in EEG monitoring, brought about by the UCC Neonatal Brain Research Group, scientists now have the means to assess the safety and effectiveness of existing and new treatments.