Edge AI Neurostimulation Controllers Versus Continuous Therapy for Drug-Resistant Focal Epilepsy
Edge Computing Artificial Intelligence Neurostimulation Controllers Versus Standard Continuous Therapy for Drug Resistant Focal Epilepsy: A Randomized Controlled Trial
Drug-resistant focal epilepsy remains a major therapeutic challenge. Conventional neurostimulators rely on static, continuous open-loop stimulation or low-bandwidth reactive loops that suffer from classification latency or high battery consumption. This prospective, double-blind, randomized controlled trial evaluates whether hardware-integrated edge computing artificial intelligence (AI) neurostimulators demonstrate superior seizure reduction, lower processing latency, and reduced battery drain compared to standard continuous neurostimulation therapy over a 6-week intervention period.
調査の概要
状態
詳細な説明
Eighty adult patients with drug-resistant focal epilepsy undergoing intracranial lead implantation across tertiary referral centers were randomized 1:1 post-implantation into two parallel groups: Edge Computing AI Neurostimulation: Received localized on-hardware machine learning feature extraction and pre-ictal spike classification from intracranial EEG, triggering adaptive targeted high-frequency electrical impulse bursts. Standard Continuous Therapy: Received conventional fixed-frequency, open-loop electrical neurostimulation. The trial evaluated clinical seizure frequency tracked via direct intracortical telemetry logs validated by daily patient diaries, as well as operational hardware metrics including microchip classification latency, real-time inference execution loads, and daily milliampere-hour battery consumption over 6 weeks.
研究の種類
入学 (実際)
段階
- 適用できない
連絡先と場所
研究場所
-
-
Punjab Province
-
Lahore、Punjab Province、パキスタン、54570
- Tertiary Neurosurgical Referral Hospitals
-
-
参加基準
適格基準
就学可能な年齢
健康ボランティアの受け入れ
受講資格のある性別
説明
Inclusion Criteria:
- Diagnosed with refractory (drug-resistant) focal epilepsy.
- Minimum of four electroclinically documented disabling focal seizures per month despite compliance with two or more appropriately chosen antiseizure medications.
- Scheduled for clinical intracranial lead implantation (subdural grid or depth electrodes).
- Aged 18 to 65 years at the time of enrollment.
Exclusion Criteria:
- Diagnosis of primary generalized epilepsy.
- Progressive neurological disorders.
- Severe cognitive impairment.
- Active psychiatric conditions.
- Previous cranial neurostimulation implants.
- Medical contraindications to surgical implantation or continuous intracortical telemetry.
研究計画
研究はどのように設計されていますか？
デザインの詳細
- 主な目的：処理
- 割り当て：ランダム化
- 介入モデル：並列代入
- マスキング：トリプル
武器と介入
参加者グループ / アーム |
介入・治療 |
|---|---|
|
実験的：Edge Computing AI Neurostimulation
Participants received an implantable neurostimulator running local edge computing machine learning algorithms to classify pre-ictal spikes from intracortical EEG in real time, delivering adaptive targeted electrical impulse bursts. |
Hardware-integrated neural processing engine executing real-time feature extraction and pre-ictal spike classification from intracranial EEG, triggering adaptive electrical bursts. |
|
アクティブコンパレータ：Standard Continuous Therapy
Participants received continuous open-loop static neurostimulation delivering fixed-frequency electrical pulses regardless of underlying electrical activity. |
Static, fixed-frequency continuous open-loop electrical pulse stimulation delivered regardless of underlying electrical activity. |
この研究は何を測定していますか？
主要な結果の測定
結果測定 |
メジャーの説明 |
時間枠 |
|---|---|---|
|
Monthly Clinical Focal Seizure Frequency
時間枠：Baseline and Week 6 (completion of intervention)
|
Number of clinical focal seizures per month, recorded continuously via automated direct intracortical telemetry logs and verified against daily patient seizure diaries |
Baseline and Week 6 (completion of intervention)
|
二次結果の測定
結果測定 |
メジャーの説明 |
時間枠 |
|---|---|---|
|
Microchip Pre-Ictal Classification Latency
時間枠：Continuously across the 6-week intervention
|
Local on-hardware signal processing and classification response time measured in milliseconds (ms) from raw intracranial EEG pattern identification to trigger execution. |
Continuously across the 6-week intervention
|
|
Daily Battery Drain Rate
時間枠：Continuously across the 6-week intervention
|
Average daily battery consumption rate measured in milliampere-hours per day (mAh/day) from device telemetry logs. |
Continuously across the 6-week intervention
|
|
Real-Time Microchip Inference Load
時間枠：Continuously across the 6-week intervention
|
Percentage of microchip processing capacity utilized during real-time on-device machine learning execution. |
Continuously across the 6-week intervention
|
|
Number of Participants with Adverse Events
時間枠：Throughout the 8-month trial duration
|
Hardware-related, surgical, or neurological adverse events documented throughout the trial. |
Throughout the 8-month trial duration
|
協力者と研究者
捜査官
- 主任研究者：Ammar Khalil、University of Kotli
研究記録日
主要日程の研究
研究開始 (実際)
一次修了 (実際)
研究の完了 (実際)
試験登録日
最初に提出
QC基準を満たした最初の提出物
最初の投稿 (実際)
学習記録の更新
投稿された最後の更新 (実際)
QC基準を満たした最後の更新が送信されました
最終確認日
詳しくは
本研究に関する用語
その他の研究ID番号
- UoK1117A
個々の参加者データ (IPD) の計画
個々の参加者データ (IPD) を共有する予定はありますか?
医薬品およびデバイス情報、研究文書
米国FDA規制医薬品の研究
米国FDA規制機器製品の研究
この情報は、Web サイト clinicaltrials.gov から変更なしで直接取得したものです。研究の詳細を変更、削除、または更新するリクエストがある場合は、register@clinicaltrials.gov。 までご連絡ください。 clinicaltrials.gov に変更が加えられるとすぐに、ウェブサイトでも自動的に更新されます。