Highlighting Anomalies in Ultrasonic Scan Data by Shannon Information Processing
Bibliographic record
Abstract
Analysis of ultrasonic nondestructive testing data can be viewed as an exercise in anomaly detection, whereby, signals/image features found to appear consistently throughout a scan are attributed to non-relevant reflections from the weld root/cap, back-wall, wedge base, etc., while isolated indications of sufficient amplitude (with respect to a calibrated reference level) are identified as potential defects requiring further characterization.Finding indications which correspond to flaws in this way can be challenging, particularly when they are low amplitude and/or are proximate to strong geometric reflectors.In this paper, a novel post-processing scheme is proposed to facilitate identifying defects in encoded ultrasonic scans.The algorithm effectively emphasizes indications which are unusual compared with an un-flawed reference scan (or region of a scan).This is accomplished by estimating the distribution of signal amplitudes in the reference scan and then using the reference distribution to compute the so-called "Shannon Information" associated with new data points.Shannon Information is a fundamental metric in Information Theory, which quantifies how surprised one should be to observe a given quantity -ultrasonic signal amplitude in this case -knowing how that quantity varies in general, as described by its probability distribution.The mathematical foundations of the technique are outlined in detail followed by a demonstration of the efficacy of Shannon Information processing on Time of Flight Diffraction (ToFD) and Phased Array (PA) scans featuring flaws which are difficult to discern by conventional means.The proposed method of presenting ultrasonic inspection data is observed to increase signal-to-noise ratio and highlight subtle perturbations in consistent, non-relevant scan features associated with presence of defects.In addition, in the absence of flaws, Shannon Information processing is shown to transform ultrasonic signals, projection views and Delay and Sum rendered images into stationary random processes with known null distributions which allows for global detection thresholds to be set according to a desired level of statistical significance.
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 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".