K Number
K232074
Device Name
Serafin®
Manufacturer
Date Cleared
2024-02-22

(225 days)

Product Code
Regulation Number
872.5470
Panel
DE
Reference & Predicate Devices
Predicate For
N/A
AI/MLSaMDIVD (In Vitro Diagnostic)TherapeuticDiagnosticis PCCP AuthorizedThirdpartyExpeditedreview
Intended Use

The Serafin is intended for the orthodontic treatment of malocclusion.

Device Description

Serafin® is composed of series of transparent removable orthodontic appliances manufactured using thermoforming technology based on a 3D model of the patient's teeth, customized according to a Dental Practitioner's specific prescription, indicated for treatment of dental malocclusion by applying continuous gentle force to realign teeth.

The time for each stage of aligner usage is set in the personal treatment plan, and this varies for each patient and the Dental Practitioner's choice. Each set of aligners needs to be worn for at least 2 weeks. Patients the total treatment time lasts about 6 to 24 months.

AI/ML Overview

The provided text is a 510(k) summary for the Serafin® orthodontic clear aligner system. It focuses on demonstrating substantial equivalence to a predicate device (Invisalign) through non-clinical testing. Crucially, this document does not contain information about clinical studies with human subjects, algorithm performance metrics (such as accuracy, sensitivity, or specificity), or the use of AI/machine learning.

Therefore, I cannot provide details on the following aspects of your request:

  • A table of acceptance criteria and reported device performance related to AI/algorithm output (as no such data is present).
  • Sample size used for a test set for AI/algorithm performance.
  • Number of experts used to establish ground truth for a test set.
  • Adjudication method for a test set.
  • Multi-reader multi-case (MRMC) comparative effectiveness study.
  • Standalone (algorithm only) performance.
  • Type of ground truth used (expert consensus, pathology, outcomes data, etc.) for AI/algorithm performance.
  • Sample size for the training set for an AI/algorithm.
  • How the ground truth for the training set was established for an AI/algorithm.

The document focuses on the physical, chemical, and biological properties of the device itself (the clear aligner), not on computational performance or diagnostic capabilities associated with AI.

Here's what the document does provide regarding acceptance criteria and testing, albeit for the physical device, not an AI algorithm:

1. Acceptance Criteria and Device Performance (for the physical device):

Acceptance Criteria (Set Test Criteria)Reported Device Performance
Physical and Chemical Tests:All test results met the preset test criteria, except as noted for Flexural Modulus.
AppearanceConformed to standard
DimensionConformed to standard
Water sorption & solubilityLow water uptake, Low water solubility (conformed to standard)
3-point flexural strength (Flexural Modulus, Flexural Strength, Tear Strength)Flexural Modulus did not meet standard requirement of 1500 MPa. The document states this requirement is not applicable as the device is designed for subtle force, not rigidity. Flexural Strength and Tear Strength conformed to standard.
HardnessConformed to standard
pHConformed to standard
Potassium permanganate reducing substances residueConformed to standard
Residue on evaporationConformed to standard
Heavy metalsConformed to standard
Ultraviolet-visible spectrumConformed to standard
Biological Evaluation:Testing demonstrated biocompatibility.
CytotoxicityConformed to standard
Skin sensitizationConformed to standard
Acute oral mucosa irritationConformed to standard
Intracutaneous reactivityConformed to standard
Acute systemic toxicityConformed to standard
IrritationConformed to standard
Biological risk assessment (ISO 10993-1, ISO 7405)Conducted
Toxicological risk assessment (for leachables/degradation products)Conducted to monitor and mitigate biological risk.

2. Sample size used for the test set and the data provenance:

  • No information on a "test set" for an AI algorithm.
  • The non-clinical testing was performed on "test articles" (implied to be samples of the Serafin® aligner material/device). The exact number of samples is not stated.
  • Data Provenance: The testing was done by the manufacturer (TNS Co., Ltd., based in South Korea) or their designated testing facilities. It's non-clinical lab testing of material properties.

3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

  • Not applicable as the testing is for physical/chemical/biological properties of the device material, not for an AI algorithm requiring expert-established ground truth. Standards (e.g., ISO, MFDS, ADA ANSI) define the test procedures and acceptance criteria.

4. Adjudication method (e.g., 2+1, 3+1, none) for the test set:

  • Not applicable. This concept applies to human reader studies or AI validation with human adjudicated ground truth, which is not described.

5. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

  • No, an MRMC study was not done, as this submission is for a physical medical device (orthodontic aligner), not an AI-powered diagnostic or assistive tool.

6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

  • Not applicable, as there is no AI algorithm being evaluated in this submission. The "3D software" mentioned is for generating a 3D model of teeth and treatment plan, which is then reviewed by a dental practitioner, not for AI-driven diagnosis or analysis.

7. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

  • For the physical device testing, the "ground truth" or reference was established by international and national standards (ISO 20795-2, ISO 10993 series, MFDS guidelines, ADA ANSI 41-2020) for material properties (e.g., flexural strength, water sorption) and biocompatibility (e.g., cytotoxicity).

8. The sample size for the training set:

  • Not applicable, as there is no AI algorithm with a distinct "training set" described in this document.

9. How the ground truth for the training set was established:

  • Not applicable.

§ 872.5470 Orthodontic plastic bracket.

(a)
Identification. An orthodontic plastic bracket is a plastic device intended to be bonded to a tooth to apply pressure to a tooth from a flexible orthodontic wire to alter its position.(b)
Classification. Class II.