Intersession Consistency of Single-Trial Classification of the Prefrontal Response to Mental Arithmetic and the No-Control State by NIRS
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Near-infrared spectroscopy (NIRS) has been recently investigated for use in noninvasive brain-computer interface (BCI) technologies. Previous studies have demonstrated the ability to classify patterns of neural activation associated with different mental tasks (e.g., mental arithmetic) using NIRS signals. Though these studies represent an important step towards the realization of an NIRS-BCI, there is a paucity of literature regarding the consistency of these responses, and the ability to classify them on a single-trial basis, over multiple sessions. This is important when moving out of an experimental context toward a practical system, where performance must be maintained over longer periods. When considering response consistency across sessions, two questions arise: 1) can the hemodynamic response to the activation task be distinguished from a baseline (or other task) condition, consistently across sessions, and if so, 2) are the spatiotemporal characteristics of the response which best distinguish it from the baseline (or other task) condition consistent across sessions. The answers will have implications for the viability of an NIRS-BCI system, and the design strategies (especially in terms of classifier training protocols) adopted. In this study, we investigated the consistency of classification of a mental arithmetic task and a no-control condition over five experimental sessions. Mixed model linear regression on intrasession classification accuracies indicate that the task and baseline states remain differentiable across multiple sessions, with no significant decrease in accuracy (p = 0.67). Intersession analysis, however, revealed inconsistencies in spatiotemporal response characteristics. Based on these results, we investigated several different practical classifier training protocols, including scenarios in which the training and test data come from 1) different sessions, 2) the same session, and 3) a combination of both. Results indicate that when selecting optimal classifier training protocols for NIRS-BCI, a compromise between accuracy and convenience (e.g., in terms of duration/frequency of training data collection) must be considered.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it