Your Scan Body Matched the Library File. The Scan Quality Did Not.
A retrospective cohort of 243 scanbodies at Hacettepe University finds that successful CAD library alignment does not guarantee adequate scan quality — nearly half showed rough texture representation, and original cylindrical PEEK scanbodies were over six times more likely to degrade than non-original aluminium alternatives.
Source Paper
A Retrospective Cohort Study on Scan Quality of Implant Scanbodies Matched With CAD Libraries
Digital dentistry has all but declared the implant digital impression solved. Intraoral scanner, scanbody, CAD library file, software algorithm: the implant position is captured and transferred without impression material, without a custom tray, and without sending the technician back three times because the margin vanished under the blood. The workflow works. Except that “the software matched the library file” and “the scan quality was adequate” are, it turns out, two entirely different things, and the field has spent some time not making that distinction.
“A Retrospective Cohort Study on Scan Quality of Implant Scanbodies Matched With CAD Libraries” (Demir Boz and Akça, Clinical Implant Dentistry and Related Research, 2025) examines exactly that gap. Fatmanur Demir Boz and Kıvanç Akça at Hacettepe University in Ankara did not ask whether the scanbody aligned with its library file. They asked what happened to scan quality after it did.
The Data Anchor
The study reviewed 243 scanbodies on Straumann bone-level implants (BLT/BL; RC/NC), captured over four years (January 2017 to January 2021) with a 3Shape TRIOS3 scanner using a standardised two-stage cut-out-and-rescan protocol. Two scanbody types were compared: the original one-piece PEEK cylindrical design (CARES Mono Scanbody; Institut Straumann AG), and a non-original one-piece aluminium design with more pronounced surface geometry (3Shape A/S). Digital records were matched to CAD library files in 3Shape Dental System v2020 and evaluated across five quality domains: reference area completeness, scanbody completeness, base integration, geometry representation, and texture representation; chi-squared testing and logistic regression with backward elimination followed.
Key Findings
- Library matching did not guarantee scan integrity. Despite successful CAD alignment across all 243 cases, 44.4% showed rough texture representation and 16% impaired geometry. Base integration was the one domain that held entirely: 100% sufficient.
- Original PEEK scanbodies were substantially riskier. For rough texture, OR = 6.663 (95% CI: 3.617–12.276, p < 0.001); for geometry impairment, OR = 4.462 (95% CI: 1.928–10.327, p < 0.001). The cylindrical design’s uniform cross-section gave the stitching algorithm little to anchor against.
- Mandibular sites compounded the problem. Mandibular locations showed 4.081 times the risk of rough texture (p < 0.001) and 3.378 times the risk of impaired geometry (p = 0.003) compared to maxillary sites; consistent with the documented challenges of scanning in a confined, moisture-prone environment.
- Single-tooth crowns were harder to scan completely than two-implant FDPs. Incomplete scanbody capture occurred in 83.3% of the incomplete-body group for single-tooth restorations versus 16.7% for two-implant cases (OR = 6.598, 95% CI: 1.858–23.432, p = 0.004).
- Limitations: non-homogeneous distribution across anterior and posterior sites; scanbody reuse history untracked; downstream effects on restoration fit not assessed.
💡 The Clinical Bottom Line
A green tick in the software alignment step is not a quality certificate. This study’s central finding is that nearly half of all scanbodies produced degraded texture representation despite successful library matching. Confirming alignment is necessary but not sufficient; post-scan review of geometry and texture representation needs to become routine.
The difference between original and non-original scanbodies here is not about brand loyalty but about geometry. A scanbody with pronounced surface features gives the scanning algorithm reliable reference points; a smooth cylinder, however faithfully it aligns to its library file, is harder to stitch accurately. That is a design specification question, not a procurement one.
For clinicians selecting scanbodies where compatible options exist, preferring designs with distinct geometric variation is now supported by this data — particularly for mandibular posterior single-tooth cases, which emerged as the highest-risk combination. The library match, satisfying as it is to watch, is only the beginning of the story.
Dr Samuel Rosehill is a general dentist with a prosthodontic focus, practising at Ethical Dental in Coffs Harbour, NSW. He holds a BDSc (Hons) from the University of Queensland, an MBA, an MMktg, and an MClinDent in Fixed & Removable Prosthodontics (Distinction) from King’s College London.
Clinical Relevance
Among 243 scanbodies evaluated retrospectively at Hacettepe University, 44.4% exhibited rough texture representation and 16% impaired geometry — even where the software algorithm successfully matched each scanbody to its CAD library file. Original cylindrical PEEK scanbodies were 6.7 times more likely to produce rough texture than non-original aluminium alternatives. Mandibular sites showed four times the risk of rough texture compared to maxillary sites. Clinicians should not assume library alignment confirms scan integrity; post-scan review of geometry and texture representation is essential, and scanbodies with distinct geometric features are preferable.
Disclosure: The author has no financial conflicts of interest related to the products or topics discussed in this review. This is an independent summary prepared for educational purposes.
Continue the conversation
This review is also published on Substack, where you can leave comments and join the discussion.
Read on Substack →