Raman Spectroscopy Detects Differences Between Idiopathic Pulmonary Fibrosis and Other Interstitial Lung Disease Biopsies
Bibliographic record
Abstract
Abstract Rationale: The diagnosis of Idiopathic Pulmonary fibrosis (IPF) is inaccurate, complex, costly and time-consuming. Raman Spectroscopy (RS) is sensitive to molecular changes and able to classify different tissue states with modest analysis. We were able to detect RS differences between IPF and other interstitial lung disease (ILD) biopsies and decently predict new IPF and non-IPF samples. Methods: This study received approval of the University of Galway Research Ethics Committee. Ten IPF and 11 non-IPF ILD frozen biopsies from different patients from the IUCPQ Biobank Quebec Respiratory Health Research Network were serially sectioned for Masson's Trichrome (MT) and RS analysis. We gathered a total of 136 RS measurements using a Witec Alpha500 Raman microscope with a 785 nm laser source. Each measurement included nine individual spectra in 10 µm2 area. Spectra were processed for cosmic ray reduction, Savitzky-Golay smoothing and background correction. All spectra were normalized to the 1450 cm-1 peak. We removed outliers using the Euclidean distance (ED) to the spectral median. We trained our model by selecting the 50 most significant wavenumbers describing highly and low fibrotic areas using a Linear Mixed Effect Model (lmem). Then, we used principal component analysis (PCA) to establish a Fibrotic_PC_limit. We used the ED to this limit to classify new measurements as highly or low fibrotic and applied the same methodology to define an IPF_PC_limit. Finally, we adopted the leave-one-sample-out validation to predict new measurements as IPF or non-IPF by ED to the IPF_PC_limit. We averaged all the measurements in each sample to evaluate prediction performance. Results: We only detected 15 outliers of the total 136 RS measurements (Fig1a). The lmem identified the 857-953, 1244-1342, 1445 cm-1 peaks as the most significant for fibrosis. The ED to the Fibrotic_PC_limit classified a total of 71.43% of all the fibrotic measurements selected by MT staining. The lmem identified the 621-769, 851-1006, 1148, 1458, 1532 cm-1 peaks as being most descriptive of IPF in fibrotic areas (Fig1c). The leave-one-sample-out validation for the ED to the IPF_PC_limit resulted in IPF and non-IPF predictive values of 63.74% and 70% respectively (Fig1e). We are including more measurements per sample to further enhance prediction performance. Conclusions: RS was highly accurate at detecting fibrotic areas in ILD biopsies. We were able to detect spectral differences between IPF and non-IPF ILD biopsies and use these differences to decently predict new samples. The use of RS has potential to improve IPF diagnosis.
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 | 0.002 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 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 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".