{
  "id": 3104207,
  "title": "Sample size buys detection, not localisation: an identifiability limit for hippocampal subfield morphometry",
  "url": "https://urgent.news/2026/08/24/sample-size-buys-detection-not-localisation-an-identifiability-limit",
  "topic": "science",
  "section": "Science",
  "published": "2026-08-24T00:00:00.000Z",
  "source": {
    "name": "bioRxiv",
    "slug": "biorxiv",
    "url": "https://www.biorxiv.org/content/10.64898/2026.08.19.745596v1?rss=1"
  },
  "original_language": "en",
  "account": "Automated segmentation has made measuring hippocampal subfield volumes a common practice, with studies now reporting which subfield is linked to a particular outcome. However, these reports often disagree, and the usual explanation—insufficient statistical power due to sample size—may not be the only factor at play. In a study involving 638 participants from a large, population-derived cohort, researchers discovered an additional limit that affects the detection of subfield-specific impacts on cognitive functions.\n\nThe findings reveal that no single hippocampal subfield contributes to a general cognitive factor beyond a basic global size component. Each subfield's coefficient interval does not exclude zero, accounting for only half a percent of the outcome variance, and no model outperforms a global factor alone in predicting outcomes for new data. The challenge arises because a null result cannot differentiate between the absence of an effect and an effect that the study design is unable to detect. To test this concept, researchers artificially implanted known effects at specific locations and sizes within the designed study and in a modified version of the same data that retained the sample size, complexity, and effect size while eliminating correlations between subfields. Despite removing the correlation between subfields, collinearity still posed significant challenges, requiring a two to three times larger sample size to achieve adequate recovery rates. Moreover, increasing the resolution of the parcellation did not help in improving the recovery rate.\n\nThe study suggests that selecting the appropriate estimator can improve study design outcomes more effectively than addressing collinearity issues. Ultimately, the research provides a calibration surface that allows researchers to evaluate and choose a suitable design before beginning data collection, addressing the identifiability limits in hippocampal subfield morphometry.",
  "summary": "Automated segmentation has made hippocampal subfield volumes a routine measurement, and studies now report which subfield relates to an outcome rather than whether the hippocampus does. Those reports do not agree with one another, and the standard explanation is insufficient statistical power. We argue that a second limit operates independently of sample size. Using 638 participants from a…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}