IIT Delhi develops non-invasive tool to detect epilepsy

A team of researchers at IIT Delhi has come up with a non-invasive EEG-based Brain Source Localization (BSL) framework for epilepsy focal detection that is time efficient and patient-friendly.

An Indian Institute of Technology (IIT) Delhi's team of researchers led by Prof Lalan Kumar, Department of Electrical Engineering, has come up with a non-invasive EEG-based Brain Source Localization (BSL) framework for epilepsy focal detection that is time efficient and patient-friendly. 

The IIT Delhi's release reads, "Given the EEG data with seizure, the array processing algorithms can point the co-ordinates within minutes. In particular, the researchers have proposed novel head harmonics based algorithms for seizure localization."

“We have proposed utilization of spherical harmonics and head harmonics basis functions for seizure Localization. To the best our knowledge, this is the first attempt in non-invasive and time efficient seizure Localization,” said Prof Lalan Kumar.

“It is a breakthrough considering comfort of patients,” Prof Kumar added.

More Details

Dr Amita Giri, a Prime Minister’s Research Fellows (PMRF) in the Department of Electrical Engineering at IIT Delhi, has developed the novel epileptic region detection method as a major part of her PhD work. The other members of the research team include Prof. Tapan K. Gandhi, Electrical Engineering Dept., IIT Delhi and Dr Nilesh Kurwale, Deenanath Mangeshkar Hospital and Research Center, Pune, Maharashtra, India.

Their study titled ‘Anatomical harmonics basis based brain source localization with application to epilepsy’ was published in Nature’s Scientific Reports.

The researchers have validated the proposed source localization algorithms on clinical EEG data for epileptogenic zone localization. The proposed framework offers an effective solution to clinicians in automated and time-efficient seizure localization.

Read more:

• IIT Delhi, Koo partner to spread awareness on air pollution
• IIT Delhi launches Certificate Programme in Machine Learning and Deep Learning