Development of a full TSC analysis pipeline: preliminary applications in aMCI and AD
Bibliographic record
Abstract
Sodium (23Na) has a crucial role in cellular homeostasis, pH regulation, and action potential propagation in muscle and neuronal fibers. Healthy tissues are characterized by low intracellular concentration [Na]1 and high extracellular concentration [Na]e, maintained by the sodium-potassium pump. The human brain consumes nearly 50% of its energy to maintain this gradient, highlighting its importance. In fact, abnormal sodium levels in the brain can lead to cellular apoptosis and necrosis, resulting in severe neurological impairments. Sodium magnetic resonance imaging (23Na-MRI) has recently seen significant growth due to its ability to non-invasively quantify sodium levels in the brain. However, the technique is hampered by several technical challenges because of sodium nuclear magnetic intrinsic properties: low gyromagnetic ratio, short T2* relaxation times, and low concentration in human tissues. These factors result in long acquisition times, incompatible with clinical activities, and low signal-to-noise ratio (SNR). In addition, 23Na-MRI requires specialized hardware, pulse sequences, and advanced reconstruction methods, as well as extensive post-processing steps that can be time-consuming and dependent on operator expertise. These challenges have all contributed to limiting its adoption in clinical practice, despite sodium's involvement in many neurological disorders, including multiple sclerosis (MS), Alzheimer’s (AD), stroke, epilepsy, cancer, and traumatic brain injury (TBI). This work aims to develop an acquisition setup and a fully automated TSC analysis pipeline to make the technique suitable for clinical practice, even for operators with no prior experience. For the acquisitions, we used the Fermat LOoped oRthogonally Encoded Trajectories (FLORET), which has already yielded promising results in vivo with acquisition times compatible with clinical practice at 3T, and data were reconstructed offline with MATLAB. For the calibration method, we used external sodium references positioned in phantom holders we designed to ensure patient comfort and the use of the same phantom masks for all acquisitions and post-processing steps. The post-processing pipeline was designed to perform the following tasks automatically: reconstruction, denoising, estimation of SNR and calibration curve, estimation of sodium levels in brain tissues, and coregistration of sodium maps in the Montreal Neurological Institute (MNI) space for group studies. It completed all these steps within approximately 4 minutes, facilitating the analysis and mitigating errors related to inter- or intra-operator variability or operator expertise. The acquisition setup and the analysis pipeline were tested at IRCCS Santa Lucia, where 23Na-MRI was conducted for the first time and included in a PNRR project on mild cognitive impairment (MCI) and Alzheimer's (AD) patients.
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 | 0.001 | 0.000 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.008 | 0.005 |
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".