Jerk-locked back-averaging (JLBA) is a specialised neurophysiological technique used to identify the cortical origin of myoclonic jerks by extracting EEG signals that are time-locked to muscle activity (Riva et al. 2024; Latorre et al. 2023). It is considered part of the "gold standard" for the clinical evaluation of myoclonus (Latorre et al. 2023).

Mechanism of JLBA

• Assumption of signal survival: The procedure is an extension of EEG–EMG polygraphy based on the principle that neural activity directly responsible for a jerk will be phase-locked to that event (Latorre et al. 2023).
• Averaging process: Multiple EEG epochs (typically 50 to 200 trials) preceding the EMG bursts are averaged together (Latorre et al. 2023; Riva et al. 2024). During this process, random background EEG activity cancels itself out, while the time-locked activity (the "pre-myoclonus spike") survives and becomes visible (Latorre et al. 2023).
• Anatomical localisation: When myoclonus is cortical, the averaged EEG trace typically reveals a transient discharge in the contralateral primary motor cortex (M1), occurring just before the muscle jerk (Latorre et al. 2023; Riva et al. 2024).

Diagnostic statistics and prevalence

JLBA is a highly specific but sometimes insensitive tool. Across various examined studies, it confirmed a clinical diagnosis of cortical myoclonus in approximately 70% of cases (Latorre et al. 2023).

Positive JLBA rates by clinical condition:

• Primary generalised epileptic myoclonus: 100% (11/11) (Latorre et al. 2023).
• Benign adult familial myoclonus epilepsy (BAFME): 100% (10/10) (Latorre et al. 2023).
• Parkinson’s Disease (PD): 100% (20/20) in patients with small-amplitude myoclonus (Latorre et al. 2023).
• Dementia with Lewy Bodies (LBD): 100% (7/7) (Latorre et al. 2023).
• Huntington’s Disease (HD): 100% (3/3) in juvenile and childhood-onset cases (Latorre et al. 2023).
• Progressive myoclonus epilepsy (PME): Reported positive rates vary from 50% (7/14) to 88% (15/17) (Latorre et al. 2023).
• Alzheimer’s Disease (AD): 70% (7/10) (Latorre et al. 2023).
• Multiple System Atrophy (MSA): 82% (9/11) (Latorre et al. 2023).
• Corticobasal Syndrome (CBS): Often negative, with reported positive rates ranging from 0% to 29% (Latorre et al. 2023).

Temporal data and latency

The timing between the EEG transient and the EMG discharge provides clues about the underlying pathophysiology (Latorre et al. 2023).

• Standard hand latency: For cortical myoclonus involving hand muscles, the latency is typically around 20 ms, which is compatible with transmission along fast-conducting corticospinal fibres (Latorre et al. 2023; Riva et al. 2024).
• Arm vs. leg: Intervals range from 10–20 ms for arm muscles to approximately 30 ms for leg muscles (Riva et al. 2024).
• Disease-specific variations:
  • Creutzfeldt-Jakob Disease (CJD): Latencies are notably longer, ranging from 50 to 280 ms, suggesting transmission through slower-conducting pathways (Latorre et al. 2023).
  • Alzheimer's Disease: Reported latencies of 20–40 ms (Latorre et al. 2023).
  • Parkinson's Disease: Reported latencies of 15–40 ms (Latorre et al. 2023).
  • Shorter Latencies: Reports of latencies <15 ms are difficult to interpret and may suggest technical errors or a subcortical generator masquerading as cortical activity (Latorre et al. 2023).

Technical limitations

• Artefacts: High-frequency or high-amplitude jerks can produce artefacts that cancel out relevant signals (Latorre et al. 2023).
• Subjectivity: There is currently no standardised automatic process for EMG burst detection, and the amplitude threshold used to trigger the back-averaging is often subjective (Latorre et al. 2023; van der Veen et al. 2024).

EEG sensitivity: EEG may have intrinsic limitations in disclosing pre-myoclonic activity; one study found magnetoencephalography (MEG) to be more effective for this purpose (Latorre et al. 2023).