NiCLIP: Neuroimaging contrastive language-image pretraining model for predicting text from brain activation images
Bibliographic record
Abstract
Predicting cognitive processes from brain activation maps has remained an open question within the neuroscience community for many years. Meta-analytic functional decoding methods aim to tackle this issue by providing a quantitative estimation of behavioral profiles associated with specific brain regions. Existing methods face intrinsic challenges in neuroimaging meta-analysis, particularly in consolidating textual information from publications, as they rely on limited metrics that do not capture the semantic context of the text. The combination of large language models (LLMs) with advanced deep contrastive learning models (e.g., CLIP) for aligning text with images has revolutionized neuroimaging meta-analysis, potentially offering solutions to functional decoding challenges. In this work, we present NiCLIP, a contrastive language-image pretrained model that predicts cognitive tasks, concepts, and domains from brain activation patterns. We leveraged over 23,000 neuroscientific articles to train a CLIP model for text-to-brain association. We demonstrated that fine-tuned LLMs (e.g., BrainGPT models) outperform their base LLM counterparts. Evaluation of NiCLIP predictions revealed that performance is optimized when using full-text articles instead of abstracts, as well as a curated cognitive ontology with precise task-concept-domain mappings. Our results indicated that NiCLIP accurately predicts cognitive tasks from group-level activation maps provided by the Human Connectome Project across multiple domains (e.g., emotion, language, motor) and precisely characterizes the functional roles of specific brain regions, including the amygdala, hippocampus, and temporoparietal junction. However, NiCLIP showed limitations with noisy subject-level activation maps. NiCLIP represents a significant advancement in quantitative functional decoding for neuroimaging, offering researchers a powerful tool for hypothesis generation and scientific discovery.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".