The research, published in BMC Medical Imaging by a team led by Lei Yang, Chuan Liu, Ting Li, Wen Li and corresponding author Wei He, was conducted at The First Affiliated Hospital of Chongqing Medical University. The investigators enrolled 81 patients with hemoptysis who, between June 2024 and June 2025, underwent both a bronchial artery CTA and a subsequent digital subtraction angiography, or DSA, examination before embolization. In total, 154 bronchial arteries could be analyzed in pairs, meaning that every vessel measured on the CT scan could be directly compared with the same vessel measured during the catheter study. This paired design is what gives the study its power, because it removes much of the variability that plagues comparisons between different groups of patients.

The technical question at the heart of the paper is deceptively simple: do the two imaging modalities agree? CTA is a three-dimensional, noninvasive technique in which X-ray images acquired during an intravenous contrast bolus are reconstructed into multiplanar views, allowing the radiologist to measure a vessel’s diameter at its origin and along its course. DSA, by contrast, is a two-dimensional fluoroscopic technique in which contrast is injected directly into the selected artery through a catheter. The two methods differ in physics, in timing, and in the physiological state of the vessel at the moment of imaging. A systematic difference between them could matter enormously, because clinical decisions about embolization often hinge on a single number: the widely used threshold of 2 millimeters, above which a bronchial artery is generally considered abnormally enlarged and a plausible bleeding source.

The answer from the data is reassuring at the group level. CTA-derived diameters averaged 2.34 millimeters, slightly larger than the 2.18 millimeters measured on DSA, yielding a mean vessel-level difference of just 0.16 millimeters. At the patient level, the mean difference shrank further to 0.14 millimeters. The two modalities correlated strongly, with a Spearman correlation coefficient of 0.83 and a P value below 0.001. Inter-observer reproducibility was good for both techniques, with intraclass correlation coefficients of 0.854 for CTA and 0.823 for DSA, indicating that different readers measuring the same images arrived at very similar numbers. In other words, the CT scanner is not systematically lying about the size of these arteries; it reads them, on average, a fraction of a millimeter larger than the catheter study does.

Agreement statistics, however, tell a subtler story than averages alone. The 95 percent limits of agreement, which describe the range within which most individual differences fall, spanned from minus 0.73 to plus 1.04 millimeters at the vessel level and from minus 0.55 to plus 0.83 millimeters at the patient level. That means a specific artery measured on CTA could plausibly appear up to a millimeter different from its DSA measurement. When the researchers tested the clinically meaningful 2-millimeter cutoff, using DSA as the reference standard, CTA at the same cutoff achieved a vessel-level sensitivity of 92.0 percent but a specificity of only 61.2 percent. At the patient level, the picture improved, with sensitivity of 86.7 percent and specificity of 80.4 percent. The pattern suggests that CTA tends to over-call borderline arteries: it rarely misses a truly enlarged vessel, but it sometimes flags vessels that DSA would measure just below the threshold.

For the interventionalist, this asymmetry has a practical interpretation. A CTA that shows a 2.1-millimeter bronchial artery should prompt careful interrogation during angiography rather than automatic embolization, because roughly four in ten such vessels may prove to be under 2 millimeters on direct measurement. Conversely, a CTA that shows a clearly enlarged artery of 3 or 4 millimeters can be trusted almost without reservation. The authors emphasize that CTA remains genuinely useful for preprocedural assessment, since it guides the angiographer to the relevant vessels, reveals anomalous origins such as arteries arising from the aortic arch or subclavian vessels, and shortens fluoroscopy time. What the new numbers add is a quantified sense of how much confidence to place in measurements that sit near the decision boundary.

The study also tackled a more speculative physiological question: does active bleeding at the time of imaging change the caliber of the bronchial arteries? The investigators stratified their analysis by whether the patient was experiencing hemoptysis during the CTA, during the DSA, during both, or during neither. Using a linear mixed-effects model with a random intercept per patient, a statistical approach that correctly accounts for the fact that multiple arteries from the same patient are not independent observations, they found that bleeding status at neither examination reached statistical significance. The coefficient for bleeding during CTA was minus 0.024 with a 95 percent confidence interval spanning minus 0.294 to 0.247 and a P value of 0.863, while the coefficient for bleeding during DSA was 0.057 with a confidence interval of minus 0.175 to 0.288 and a P value of 0.628. Neither comes close to conventional significance thresholds.

Yet the exploratory subgroup analysis contained an intriguing hint. When hemoptysis occurred only during the CTA, the median CTA-DSA difference was 0.00 millimeters, whereas when no bleeding occurred during either examination, the median difference rose to 0.20 millimeters at the vessel level and 0.33 millimeters at the patient level. The authors are careful to label this finding as hypothesis-generating rather than confirmatory, and they explicitly describe transient vasoconstriction as speculative. One possible mechanism they raise is that active bleeding may be accompanied by vasospasm or altered vascular tone that temporarily changes arterial caliber, potentially making the two examinations more similar if both capture the vessel in a constricted state. The subgroup in which bleeding occurred only during DSA contained just a single vessel and patient, precluding any statistical inference, and the overall between-group comparison excluding that group did not reach significance, with a Kruskal-Wallis P value of 0.154 and all Dunn-adjusted pairwise comparisons above 0.05.

The study’s limitations are those inherent to its design. It was single-center and retrospective, which means the CTA protocols and DSA techniques reflected one institution’s practice and the analysis depended on measurements already recorded in the picture archiving and communication system. The modest sample size, particularly within bleeding subgroups, limits the precision of the stratified estimates. The authors also used DeepSeek, an artificial intelligence tool, for language refinement and grammar checking, a disclosure they make transparently while taking full responsibility for the content. The work was supported by the Tibet Autonomous Region Natural Science Foundation as part of a group-style medical aid project, reflecting the collaboration between hospitals in Chongqing and the Tibetan plateau region. The ethics committee waived individual informed consent given the retrospective nature of the analysis, and the protocol was approved under the Declaration of Helsinki.

For clinicians managing hemoptysis, the takeaway is a calibrated one. CTA and DSA measure bronchial arteries with a small, predictable systematic offset of roughly 0.15 millimeters, and with strong correlation and good reproducibility, so the CT suite remains the right place to plan an embolization procedure. But the limits of agreement remind us that individual vessels near the 2-millimeter threshold can shift across the line between the two examinations, and the modest specificity of CTA at that cutoff argues for angiographic confirmation of borderline findings. As for whether the drama of active bleeding itself reshapes the arteries being measured, the data lean toward no, while leaving the door ajar for future prospective studies with larger cohorts to test the vasoconstriction hypothesis properly. In a procedure where millimeters decide treatment, knowing exactly how much the ruler can be trusted is itself a clinical advance.

Subject of Research: Comparison of bronchial artery diameter measurements between CT angiography and digital subtraction angiography in patients with hemoptysis

Article Title: Comparison of bronchial artery diameters on CT angiography versus digital subtraction angiography in hemoptysis: a retrospective study

Article References: Yang, L., Liu, C., Li, T., Li, W., & He, W. (2026). Comparison of bronchial artery diameters on CT angiography versus digital subtraction angiography in hemoptysis: a retrospective study. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02866-3

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02866-3

Keywords: hemoptysis, bronchial artery, CT angiography, digital subtraction angiography, embolization, vascular measurement, interventional radiology, radiology, limits of agreement, vasoconstriction, BMC Medical Imaging, retrospective study

The core problem the authors identify is deceptively simple. Classical test theory, the statistical backbone of most reliability work in psychology, assumes that an observed score is the sum of a stable true score and random error. Four assumptions underpin the resulting reliability coefficients: true score and error must be additive, independent, and random, and the true score must not change between repeated measurements. In practice, these assumptions are routinely violated by practice and habituation effects — the well-documented tendency of participants to improve on cognitive and coordinative tasks simply by repeating them, without any intervention in between. When the retest score is systematically better than the test score, the parallel-readings assumption collapses, and with it the interpretability of the reliability coefficient.

The scale of the problem is striking. Habituation effects have been documented across a wide range of standard instruments, including the Stroop test, the Eriksen flanker task, the Simon task, the Digit Symbol Substitution Test, the Wechsler Memory Scale, and the Halstead-Reitan neuropsychological battery. The largest gains typically appear between the first two testing sessions, with progressively flatter improvement curves as participants grow familiar with the task. Long-term learning can even persist for years: performance on the Wechsler Adult Intelligence Scale and the Wechsler Memory Scale has been shown to be affected by task exposure that occurred months or years earlier. In one series of experiments cited in the paper, trial-to-trial random errors in neurocognitive tests started as high as 6 to 43 percent and only stabilized below 5 percent in reaction tasks after several days of repeated testing.

What makes this especially dangerous is that the field’s favorite reliability statistics can look excellent while hiding serious problems. The intraclass correlation coefficient (ICC) and Cronbach’s alpha summarize the ratio of true-score variance to observed variance, but they cannot distinguish between different sources of measurement error. The authors point to published examples where an ICC of 0.9 or higher — conventionally rated as excellent — coexisted with a mean absolute test–retest error of 20 percent and maximum individual errors approaching 90 percent. A high correlation between two measurements says nothing about how close those measurements actually are to each other on the original scale, which is what clinicians and experimenters usually need to know.

The classical remedy for this blind spot is the Bland–Altman analysis, introduced in 1986, which plots the difference between two measurements against their mean. The average difference reveals systematic bias, while the 95 percent limits of agreement — the mean difference plus or minus 1.96 standard deviations — describe the expected range of random scatter. But the authors argue that Bland–Altman analyses, as commonly used, suffer from a critical limitation: the limits of agreement are purely descriptive. They describe where the data fell, but they say nothing about whether that amount of error is acceptable for the scientific or clinical question at hand. Bland and Altman themselves noted in the original 1986 paper that acceptable limits should ideally be defined in advance to aid interpretation — advice that has largely been ignored in practice.

There is a second, subtler flaw. Standard limits of agreement are drawn as parallel horizontal lines on the Bland–Altman plot, which implicitly assumes that the absolute size of measurement error is constant across the whole range of measured values. In many empirical domains, however, the scatter grows as the measured value grows — a phenomenon known as heteroscedasticity. Parallel limits therefore tolerate a disproportionately large percentage error for small measurement values while being overly strict for large ones. A limit of ±1 degree might be trivial for a mean of 180 degrees but catastrophic for a mean of 0.5 degrees, as an example from visual-vertical perception research in stroke patients illustrates.

The new R application addresses both problems with a concept the authors call limits of acceptance, or LoAcc. Unlike limits of agreement, which are computed from the data after the fact, limits of acceptance are defined a priori by the researcher as a proportional threshold — for example, 5 percent of each participant’s mean test–retest value. Because the acceptable band scales with the magnitude of the measurement, the resulting boundaries fan out across the Bland–Altman plot rather than running parallel, correctly accommodating heteroscedastic data. A test–retest difference is classified as acceptable only if it falls within the proportional band for that individual. The app then color-codes the results: green if the systematic bias is non-significant and at least 95 percent of observations fall within the acceptance limits, red if either criterion fails.

The software itself is a Shiny app that automates the entire test–retest evaluation workflow. Users upload an Excel file in wide format, with a participant identifier in the first column and measurement pairs in consecutive columns; the app can also be used for validity studies against a gold standard or for inter-rater objectivity analyses. It then computes the full battery of reliability metrics — all six ICC variants with 95 percent confidence intervals, Cronbach’s alpha, the standard error of measurement, the minimal detectable change, the coefficient of variation, the mean absolute error, and the mean absolute percentage error — alongside a paired-samples t-test for systematic bias and publication-ready Bland–Altman plots. Users can set the acceptance threshold per variable pair, specify measurement units, and export results tables and figures automatically. The code is openly available via the Open Science Framework.

To demonstrate the tool, the authors simulated Stroop test data across four consecutive testing days with 100 virtual participants, using a 5 percent acceptance threshold calibrated to the error level observed after four days of habituation in their earlier empirical work. The progression is instructive. Between days 1 and 2, mean performance rose significantly and 86 of 100 observations fell within the acceptance limits — both criteria failed, and the plot lit up red. Between days 2 and 3, the systematic bias had largely vanished, but random scatter remained too high, with 88 of 100 observations inside the limits. Only between days 3 and 4 did the data satisfy the criterion: no significant bias, 97 of 100 observations within the limits, and the mean absolute error shrinking from 0.828 seconds to 0.479 seconds. Applied to a real dataset from a choice reaction task tested twice daily over five days, the app showed the same convergence, and also revealed a near-threshold case at a 10 percent limit where a single data point sitting exactly on the boundary tipped the classification — a reminder, the authors note, that such judgments should consider the magnitude of the deviation rather than being treated as purely binary.

The broader message is a call for transparency that resonates far beyond cognitive psychology. The authors argue that expected intervention-induced changes from previous studies can serve as a rational anchor for setting acceptance limits: a 10 percent measurement error might be tolerable if an intervention is expected to change the outcome by 100 percent, but it is unacceptable if the expected effect is only 10 percent. They acknowledge that no universal guideline exists for what counts as acceptable, and that the 5 and 10 percent thresholds used in their examples are illustrative rather than prescriptive. Still, by forcing researchers to state their tolerance for error in advance and by separating systematic from random error — since a valid ICC interpretation requires the absence of systematic bias — the tool turns habituation from an invisible confound into a measurable, plottable, and ultimately controllable quantity. For a field increasingly worried about the reproducibility of its measurements, knowing exactly how much habituation is enough may prove to be one of the most practical questions psychology has finally learned to answer.

Subject of Research: Test–retest reliability, habituation effects, and agreement analysis in behavioral and neurocognitive research

Article Title: How much habituation is enough? An R application for ad hoc content-related limits of acceptance in behavioral research

Article References: Warneke, K., Siegel, S. D., Afonso, J., Herbsleb, M., & Wallot, S. (2026). How much habituation is enough? An R application for ad hoc content-related limits of acceptance in behavioral research. Behavior Research Methods, 58(11), Article 307. https://doi.org/10.3758/s13428-026-03170-2

Image Credits: AI Generated

DOI: 10.3758/s13428-026-03170-2

Keywords: habituation, practice effects, test–retest reliability, Bland–Altman analysis, limits of agreement, limits of acceptance, intraclass correlation coefficient, Cronbach's alpha, measurement error, R software, psychometrics, neurocognitive testing