A new brain mapping technique promises relief from drug-resistant epilepsy without also requiring a highly invasive brain mapping experience. A new report from Citizens United for Research in Epilepsy, or CURE, explains that a “new noninvasive method can effectively map the source and scale of seizure activity in people with epilepsy.”
Describing a recent study, the article goes on to say that this new tool may reduce surgeries required for treatment while also eliminating the need for the far more invasive brain mapping of the past. Explaining it further, the article says that clinicians have been forced to "rely on an invasive technique to try to find the area responsible for someone’s seizures by implanting electrodes deep into the outer layer of the brain.”
This preliminary surgery is then followed up by a second treatment – removal of brain tissue identified as problematic by the electrodes. However, these two surgical interventions are also the underlying cause of brain bleeding and even infection.
AI and EEG Combined
The new mapping technique is authentically cutting edge as it uses the traditional power of EEG recordings (relying on 76 electrodes positioned carefully on the scalp) and AI, or artificial intelligence. Together, they can identify the locations of seizure activity in the brain, as well as revealing the amount of brain tissue involved.
Is it as precise as electrodes within the actual tissue? The authors of the study say no. As reported, “noninvasive imaging techniques come with trade-offs. While EEG is adept at recording activity during a seizure, it cannot always hone in on a seizure’s source with precision.”
The quantity of tissue, bone, and fluid between the scalp electrodes and the brain makes this challenging, and some signals may be distorted. Traditional workarounds involved complex math and a bit of guesswork, but the introduction of AI and "new algorithm, called ‘fast spatio-temporal iteratively reweighted edge sparsity (FAST-IRES)’” may overcome the traditional hurdles.
This mathematical approach offers information relating to the EEG signal. It is capable of indicating the extent of the brain’s network involved in the EEG outcomes. Relying on machine learning, AI, it can tabulate the “thresholds at which signals are deemed significant.”
They tested their approach on 36 traditional epilepsy patients ready to use surgical intervention to address their symptoms. These patients used the traditional, invasive method with electrodes implanted into the brain tissue. However, they also participated in FAST-IRES on EEG recording experiments.
The research team discovered that their method was as effective as the implanted electrodes, where the determination of the extent of the seizure source and its location were concerned. In other words, the 36 patients could have relied on the single surgical process rather than the implant and then surgical treatment. They would have had identical surgical outcomes from a single surgery.
Epilepsy is known to be due to too much electrical activity in the brain, which then triggers a seizure. The consequences of untreated seizures are immense and destructive, and yet millions face drug-resistance or difficult to treat conditions. Some people may even die because of the damage caused by epilepsy. Little wonder that so many are willing to undergo two very risky and invasive surgical procedures to improve their quality of life.
With this new technology, the invasiveness of diagnosis may come to an end, and a very targeted area of the brain may be determined without electrode implantation. With 1/3 of all epilepsy patients facing drug-resistant seizures activity, this new tool is an authentic lifesaver and should be moved forward quickly through testing and approval.