IMG-67. Confirmatory clinical trial evaluating diagnostic performance of confocal laser endomicroscopy (cCeLL - Ex vivo) for intraoperative brain tumor diagnosis
Bibliographic record
Abstract
Abstract cCeLL-Ex vivo is a second-generation confocal laser endomicroscopy (CLE) platform that uses indocyanine green (ICG) as a fluorescent dye. This multicenter, assessor-blinded, prospective confirmatory trial (KCT0008869) comprehensively evaluated the diagnostic efficacy of CLE compared to frozen section (FS) analysis for intraoperative brain tumor diagnosis. In parallel, an early-stage AI diagnostic algorithm was developed using CLE images to support tumor classification. The trial was conducted from October 2022 to December 2024 across four tertiary medical centers in Korea and Canada. Patients aged ≥19 years undergoing surgery for newly diagnosed brain tumors were enrolled. The primary endpoint was to evaluate the diagnostic accuracy and prove non-inferiority of CLE compared to FS. Secondary endpoints included comparison of sensitivity, specificity, diagnostic accuracy by sample location, diagnostic turnaround time, interobserver variability, and tumor subtype classification. An AI model was developed using a Swin Transformer-based hierarchical framework. A total of 461 samples from 376 patients (mean age: 55.6 years; 162 males) were analyzed. Most samples were obtained from the tumor core (n=382, 82.8%). Meningioma (n=121, 26.2%) and glioma (n=106, 23.0%) were the most common tumor types. CLE demonstrated non-inferior diagnostic accuracy to FS (0.92 vs. 0.94; P=0.14), with comparable sensitivity (95% vs. 96%; P=0.404) and specificity (79% vs. 68%; P=0.307). Accuracy at the tumor margin was also comparable (CLE: 0.72 vs. FS: 0.68; P=0.831). CLE had a significantly faster diagnostic turnaround (8m12s vs. 23m38s; P<0.001). Interobserver agreement between neuropathologists was excellent (κ=0.805). The AI algorithm achieved 97% diagnostic accuracy (sensitivity: 98.9%, specificity: 84.6%) and 88.6% accuracy for tumor classification. Overall, CLE demonstrated non-inferior diagnostic efficacy with a significantly faster diagnostic turnaround than FS. Integration of AI with CLE may further assist in making faster and more accurate intraoperative decisions.
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 | 0.008 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".