A (Sub)field Guide to Quality Control in Hippocampal Subfield Segmentation on Highresolution T2-weighted MRI

Abstract

Inquiries into properties of brain structure and function have progressed due to developments in magnetic resonance imaging (MRI). To sustain progress in investigating and quantifying neuroanatomical details in vivo, the reliability and validity of brain measurements are paramount. Quality control (QC) is a set of procedures for mitigating errors and ensuring the validity and reliability of brain measurements. Despite its importance, there is little guidance on best QC practices and reporting procedures. The study of hippocampal subfields in vivo is a critical case for QC because of their small size, inter-dependent boundary definitions, and common artifacts in the MRI data used for subfield measurements. We addressed this gap by surveying the broader scientific community studying hippocampal subfields on their views and approaches to QC. We received responses from 37 investigators spanning 10 countries, covering different career stages, and studying both healthy and pathological development and aging. In this sample, 81% of researchers considered QC to be very important or important, and 19% viewed it as fairly important. Despite this, only 46% of researchers reported on their QC processes in prior publications. In many instances, lack of reporting appeared due to ambiguous guidance on relevant details and guidance for reporting, rather than absence of QC. Here, we provide recommendations for correcting errors to maximize reliability and minimize bias. We also summarize threats to segmentation accuracy, review common QC methods, and make recommendations for best practices and reporting in publications. Implementing the recommended QC practices will collectively improve inferences to the larger population, as well as have implications for clinical practice and public health.

Figure 1.

Illustration of the QC process and investigator-guided decision making for data quality. Green checkmarks indicate passed QC, while red cross marks indicate failed QC.

Figure 2.

Examples of the quality of T₂-weighted images according to rating categories of “Pass” (left panel), “Check” (middle panel), and “Fail” (right panel). In this example, image quality rated as “Pass” and “Check” are considered passable. However, those in the “Check” category may be at higher risk of subsequent segmentation errors. The types of imaging artifacts present in the example images are noted.

Figure 3.

Examples of scan quality ratings based on the landmark visualization of the SRLM, a critical landmark for segmentation in the head (top panel) and body (bottom panel) of the hippocampus. Additional landmarks, the alveus and uncus, are also depicted. Note: Definitions of quality may differ between investigators but there should be consistency in the application of operational definitions.

Figure 4.

Example QC approach using a 4-point error severity scale from a published and validated protocol (Canada et al., 2023). In this example protocol, errors could be 0- not present (not pictured here), 1- minor (<10% of label affected), 2- moderate (10–25% of label affected), or 3- major (>25% label affected) and were categorized by type. Only major errors are corrected in this protocol to mitigate human bias. Overestimated label (OL), underestimated boundary (UB), and dropped pixel (DP) errors are depicted.

Figure 5.

Example QC approach from https://www.youtube.com/watch?v=XHXu-AGR6pE for the PMC segmentation atlas (Yushkevich et al., 2015b) applied to data collected using the parameters reported in Daugherty et al. (2016). In this approach, QC mosaic screenshots generated by ASHS are examined by hemisphere to detect potential errors for each subject. Segmentations are subsequently fully reviewed slice-by-slice only for regions with identified errors. This protocol recommends correcting only larger errors that are present on a predetermined number of slices (3 slices is used in the linked demonstration). In the depicted example, major errors were present in the right hemisphere QC screenshot for the subiculum (pink) label. In the full review of the segmentation, errors in the right subiculum label were present in 8 slices. Following this QC approach, the depicted subject would require manual correction of these errors or be excluded from analyses.