Anatomical and Image Interface Mechanisms Driving 3.0T MRI False Positives in Superior Semicircular Canal Evaluation: A Photon-Counting Detector CT Benchmark Study

Technical Summary: Anatomical and Image Interface Mechanisms Driving 3.0T MRI False Positives in Superior Semicircular Canal Evaluation

Problem Statement
Superior semicircular canal dehiscence (SSCD) syndrome involves an osseous discontinuity over the superior semicircular canal, leading to vestibulo-auditory symptoms. While 3.0T Magnetic Resonance Imaging (MRI) offers superior soft-tissue contrast and avoids ionizing radiation, it frequently yields false-positive diagnoses when evaluating the exceedingly thin bone overlying the canal. These errors manifest as a misleading appearance of high-signal cerebrospinal fluid (CSF) "penetration" through an intact bony barrier. Conversely, while High-Resolution CT (HRCT) is standard, its spatial resolution (0.5–0.6 mm) limits its ability to definitively rule out micro-defects. There is a lack of systematic quantification regarding the diagnostic accuracy of 3.0T MRI when benchmarked against ultra-high-resolution imaging, and the specific anatomical and technical mechanisms driving MRI false positives remain unclear.

Methodology
This retrospective study analyzed 49 patients (98 ears) presenting with vestibulo-auditory symptoms who underwent concurrent 3.0T MRI and temporal bone Photon-Counting Detector CT (PCD-CT).

• Imaging Protocols: PCD-CT was performed on a Siemens NAEOTOM Alpha system with an ultra-high-resolution protocol (0.2 mm slice thickness). MRI utilized a Siemens Vida 3.0T scanner with a 3D-T2 SPACE sequence (0.6 mm slice thickness).
• Reference Standard: PCD-CT served as the reference standard for determining true osseous continuity, capable of resolving sub-millimeter structures.
• Analysis: Two blinded radiologists independently evaluated images. They assessed bone continuity, measured bone thickness at the arcuate eminence apex, and graded image quality (artifact burden, brain-bone-fluid interface clarity, SNR, anatomical visibility) and diagnostic confidence on a 5-point Likert scale.
• Statistical Approach: A Generalized Estimating Equation (GEE) model was employed to account for the within-patient clustering of paired ears. Comparisons between True-Negative (TN) and False-Positive (FP) cohorts focused on bone thickness and qualitative image metrics, with Bonferroni correction applied for multiple testing.

Key Contributions
The study establishes PCD-CT as a benchmark for evaluating semicircular canal bony integrity, revealing specific physical and anatomical drivers behind MRI misdiagnosis:

1. Quantification of False Positives: The study quantifies the false-positive rate of 3.0T MRI for SSCD at 24.72%, with a Positive Predictive Value (PPV) of only 26.67%.
2. Identification of the Anatomical Threshold: The research identifies a specific bone thickness threshold associated with false positives. Cases misdiagnosed by MRI had significantly thinner overlying bone (median 0.39 mm) compared to true negatives (median 0.82 mm).
3. Mechanism Elucidation: The paper attributes false positives primarily to the partial volume effect. When bone thickness falls below approximately 0.45 mm, the spatial resolution limits of MRI prevent the effective resolution of the low-signal bony septum between high-signal CSF and endolymph, creating an artificial appearance of a defect.
4. Image Interface Factors: The study highlights that reduced clarity of the brain-bone-fluid interface and lower diagnostic confidence scores are correlated with false-positive interpretations.

Results

• Diagnostic Performance: Of 98 ears, PCD-CT confirmed 9 true SSCD cases. MRI identified 8 true positives, 22 false positives, 67 true negatives, and 1 false negative.
• Bone Thickness Analysis: The false-positive cohort exhibited significantly thinner bone than the true-negative cohort (GEE: Wald χ² = 71.656, P < 0.001).
• Threshold Determination: ROC analysis determined an optimal bone thickness cut-off of 0.45 mm for differentiating MRI false positives (AUC = 0.918). Below this threshold, the likelihood of a false-positive MRI diagnosis increases significantly.
• Qualitative Metrics: False-positive cases showed a trend toward lower scores in brain-bone-fluid interface clarity and diagnostic confidence, though only the confidence metric remained statistically significant after rigorous correction.
• False Negative Case: A single false negative was attributed to a thick bone with a focal defect where CSF did not sufficiently inundate the gap to generate a high-signal signature.

Significance and Claims
The authors claim that the overdiagnosis of SSCD on 3.0T MRI is mechanically driven by partial volume effects over exceptionally thin bone (<0.45 mm), compounded by blurred image interfaces and diminished observer confidence.

• Clinical Implication: Positive MRI findings for SSCD should be treated as "high-suspicion" signals rather than definitive diagnoses.
• Recommendation: To avoid overtreatment and unnecessary surgical interventions, the study advocates for the routine use of high-resolution CT (specifically PCD-CT or high-end HRCT) to verify positive MRI results. This ensures accurate depiction of the true osseous microarchitecture before clinical decision-making.
• Limitations: The authors modestly note that the small number of true-positive cases (n=8) results in wide confidence intervals for sensitivity and PPV, and the single-center retrospective design may introduce selection bias. The ROC analysis is presented as a mechanistic demonstration of physical limits rather than a standalone clinical decision rule.