SPAGHETTI leverages massive H&E morphological models for phase contrast microscopy images with a generative deep learning approach
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
Phase contrast microscopy (PCM) is a powerful cell imaging method, one of the few technologies to delineate and track cell structure in live cells without staining. Despite PCM’s great potential and popularity in monitoring live-cell populations, there is a lack of algorithms to extract morphological information from these images due to the lack of large training datasets. To overcome this challenge and enable advanced, high throughput quantitative analysis of PCM images, we introduce SPAGHETTI: a lightweight image translator built on a modified cycle-consistent generative adversarial network. SPAGHETTI translates PCM images into those resembling hematoxylin and eosin (H&E) pathological images which, due to the pervasive and widespread use in clinical settings, are the basis for most large-scale deep learning models for quantitative analyses. We demonstrate that by first using SPAGHETTI to translate PCM images into H&E-like images, we could achieve significantly improved performance on cell segmentation through the use of tissue and H&E-specific cell segmentation models. We also show that by passing translated PCM images across several independent datasets into H&E feature extractor models, we improve the performance of cell-type annotation, experimental media classification, and cell viability prediction. Overall, SPAGHETTI enables many quantitative analyses of PCM that were previously impossible and acts as a valuable preprocessing step to help researchers gather novel information about cell states through the downstream quantitative analysis of morphological features.
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 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| 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