Automated Data Analysis for Event-Driven Scanning Transmission Electron Microscopy (Tempo-STEM)
Notice bibliographique
Résumé
Recently [1], we introduced a method for improving the information-per-electron ratio in scanning transmission electron microscopy (STEM) based on pulse counting and event-driven beam blanking. This method, which we are calling “Tempo” (trigger-event modulated probability observation [2]) combines pulse counting, a custom logic circuit, and a fast beam blanker with arbitrary timing control (EDM or Electrostatic Dose Modulator) to listen to the STEM detector signals and blank the beam when a certain threshold is reached in each pixel. Tempo turns the usual measurement of STEM image intensity on its head. Rather than measuring the number of electrons detected in a fixed dwell time, Tempo measures the amount of time to reach a fixed number of detected electrons. When this threshold is reached it either immediately moves on to the next pixel or blanks the beam for the duration of the dwell time. This reduces dose, especially in high-scattering-rate sample regions, while also improving the information-per-electron efficiency by avoiding diminishing returns and equalizing the signal to noise ratio across the image. The process of calibrating and analyzing Tempo data is nontrivial. One must convert an analog pulse width modulation signal into a precise beam-on time in microseconds, then determine the ratio between that signal and a pulse-counted signal to determine either a detection rate (in counts per microsecond) or a mean detection interval (in microseconds per count). Various corrections must be made for time lags, statistical biases, and pixels where the threshold was not reached. This can be complex for an end user who would like to use Tempo without having to study the theory of the data analysis. To meet this need, we are introducing software that provides simple, real-time calibration and data analysis for Tempo measurements. The software integrates with the STEM control software (including planned plugins for Gatan’s Digital Micrograph[3] and JEOL’s FEMTUS) to provide a seamless interface for automatically generating output images that act like virtual detectors. Updates are done in real time, even as scans are in progress. The main interface window (Figure 1A) provides the essential features, with access to a few frequently accessed parameters and workflow controls, while an Advanced Settings window (Figure 1B) provides detailed, sophisticated low-level control when needed by expert users. This includes auto-calibration routines that determine zero offsets and scaling of the time axis. To the extent possible (depending on third-party constraints), the code will be made open source, allowing sophisticated users to fully understand the calibrations and add their own algorithms. The result is simple, directly interpretable images such as are shown in Figure 2. All input and output images, along with metadata fully documenting the calculations, are saved in standard file formats with consistent filename conventions allowing easy import into other analysis software. Example GUI windows for one version of the software. A. Main window used for general operation by all users. B. Advanced window with options that are set-once-and-forget for basic operation but enable fine control for unusual workflows. Example calibrated STEM images produced by the analysis. (Left) Beam exposure time in μs per pixel, using a maximum pixel dwell time of 12 μs. (Right) Dark field detection rate in counts per μs.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,005 | 0,010 |
| Méta-épidémiologie (sens strict) | 0,002 | 0,001 |
| Méta-épidémiologie (sens large) | 0,002 | 0,001 |
| Bibliométrie | 0,005 | 0,003 |
| Études des sciences et des technologies | 0,001 | 0,001 |
| Communication savante | 0,002 | 0,002 |
| Science ouverte | 0,003 | 0,002 |
| Intégrité de la recherche | 0,001 | 0,002 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,014 | 0,006 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».