(258 days)
Harioculture TL-16 Time-lapse Incubator consists of the following devices with the following indications for use:
Time-lapse Incubator provides an environment with controlled temperature and gas concentrations (CO2 and O2) or mixed gas (CO2 and other gases) for the development of embryos at or near body temperature. Use of the Time-lapse Incubator is limited to five days (120 hr) covering the time from post insemination to day five of development.
The Hariomed culture dish is intended for preparing, storing, and transferring human embryos. The Hariomed culture dish must be used together with the Time-lapse Incubator.
The Harioculture-client software is intended for displaying, comparing, storing, and transferring images generated by the Time-lapse Incubator. This software includes a user annotation function for capturing information on embryo development parameters as well as a user-defined modeling function, which allows the user to combine annotated information on embryo development parameters to aid in embryo selection. The Harioculture-client software does not control any hardware components in the Time-lapse Incubator.
The Harioculture-server software is intended to store, archive and transfer data. In addition, this software includes functions for managing models and performing calculations based on image data and embryo development parameters.
The Time-lapse Incubator, Harioculture-client software, and Harioculture-server software must be used together to export embryo images from the Time-lapse Incubator. The Harioculture-client software and Harioculture-server software must be used together to assist users to analyze the embryo images.
Harioculture TL-16 Time-lapse Incubator consists of the following components:
- Time-lapse Incubator,
- Harioculture-server software,
- Harioculture-client software,
- Hariomed culture dish.
The Time-lapse Incubator is a benchtop embryo incubator with a built-in microscope for time-lapse imaging intended to be used for the culture and monitoring of embryos used in Assisted Reproductive Technology (ART) procedures. It provides temperature control, gas control, and time-lapse microscopy at multiple focal planes. This device can hold up to 16 culture dishes (Hariomed culture dish) in the culture cabin. The device can be connected to the capacity expansion module which in turn can also accommodate 16 culture dishes. The capacity expansion module is similar to the main incubator and provides the same culture capacity as the incubator host. However, it cannot be run independently and needs to be used together with the Incubator host. The incubator host includes the software to control the temperature, gas concentration, imaging as well as image display on the incubator screen. The culture dishes are placed in the culture slots on the turntable in the Time-lapse Incubator. The culture slots provide direct heat transfer to the Hariomed culture dishes. The built-in microscope consists of an illumination unit (red LED, 630±5 nm) and an inverted microscope/camera unit. During image acquisition, turntable is rotated to position individual culture dishes on the microscopy system and image stacks are acquired for individual embryos in each culture dish.
The Time-lapse Incubator supports two gas control and culture modes:
- Self-mixed mode – Under the self-mixed gas culture mode, CO2, and N2 gases are supplied from medical grade gas cylinders. These gasses (CO2, N2 and re-flux gas) are mixed in a mixing chamber and passed through a HEPA/VOC filter prior to delivery to the culture cabin. This culture mode does not allow for humidification of the chamber (dry-culture mode).
- Pre-mixed mode – Under the pre-mixed gas culture mode, the gas is supplied from standard premixed medical gas cylinders. Pre-mixed gas passes through the HEPA/VOC filter and a humidification box to form a humidified gas which then enters the culture cabin. The humidification box is sterile, single-use component with a sterility assurance level (SAL) of 10-6 and a shelf-life of two years. The pre-mixed culture mode allows humidification of chamber (wet-culture mode).
The Hariomed culture dish is a single use, single-patient, polystyrene, radiation sterilized culture dish intended for preparing, storing, and transferring human embryos. The Hariomed culture dish is intended for use only with Harioculture TL-16 Time-lapse Incubator and includes two culture ponds. Each culture pond has eight microwells and a total of 16 embryos from a single patient can be cultured on one dish. Each dish includes four rinsing wells for rinsing and handling the embryos before or after incubation. The dishes have area for unique barcode labels that can be printed through the Harioculture-client software. The barcode labels provide patient identification information. The Hariomed culture dish has a sterility assurance level (SAL) of 10-6 and a shelf-life of two years.
The Harioculture-client software is intended for displaying, comparing, storing, and transferring images generated by the Time-lapse Incubator. The data that can be viewed using this software includes embryo images, incubation details, alarms, log files and other instrument parameters. This software also includes a user annotation function for capturing information on embryo development parameters as well as a user-defined modeling function, which allows the user to combine annotated information on embryo development parameters to aid in embryo selection. The Harioculture-client software does not control any hardware components in the Time-lapse Incubator.
The Harioculture-server software allows users to update and view common data. The server acts as the central unit, which stores data and controls the data flow to and from the connected devices. The server can be connected to multiple Time-lapse incubators and computers with the Harioculture-client software.
It's important to note that the provided FDA 510(k) clearance letter and summary do not describe a clinical study involving human patients or complex AI-driven diagnostic decisions. Instead, the document focuses on the performance and safety validation of a medical device (Time-lapse Incubator) and its associated hardware and software components. The "acceptance criteria" and "study" described in the document are primarily for bench testing, in-vitro assays (like Mouse Embryo Assay), and engineering validation, not a comparative clinical trial for an AI-assisted diagnostic tool in the traditional sense (e.g., comparing AI-assisted radiologist performance to unassisted radiologist performance).
Therefore, some of the requested information regarding "AI performance," "human readers," "effect size," "ground truth establishment" for training data, and "adjudication methods" as one might see in the validation of a diagnostic AI algorithm will not be directly applicable or available in this document.
However, I can extract and structure the information that is present in the document in a way that aligns with your request's format, while highlighting the type of studies actually performed.
Device Description: Harioculture TL-16 Time-lapse Incubator System
The Harioculture TL-16 Time-lapse Incubator system is a medical device designed for Assisted Reproductive Technology (ART) procedures. It comprises:
- Time-lapse Incubator: Provides a controlled environment (temperature, CO2/O2) for embryo development, with a built-in microscope for time-lapse imaging.
- Hariomed culture dish: A single-use, radiation-sterilized polystyrene dish for culturing human embryos within the incubator.
- Harioculture-client software: Displays, compares, stores, and transfers images from the incubator, and includes user annotation and user-defined modeling functions to aid in embryo selection. It does not control hardware.
- Harioculture-server software: Stores, archives, and transfers data, manages models, and performs calculations based on image data and embryo development parameters.
The system's client and server software assist users in analyzing embryo images, and the primary "AI-like" function is the "user-defined modeling function" which combines annotated information to aid in embryo selection. This suggests a user-configurable scoring or selection model rather than a pre-trained deep learning AI model for image interpretation.
Acceptance Criteria and Reported Device Performance
The "acceptance criteria" and "reported device performance" in this context refer to the engineering specifications and the results of non-clinical bench and in-vitro testing, rather than clinical endpoints from a human study.
Table of Acceptance Criteria and Reported Device Performance
| Parameter | Acceptance Criteria (Design Specification) | Reported Device Performance (Test Results) | Source of Information / Test Type | Note |
|---|---|---|---|---|
| Incubator Parameters | ||||
| Temperature accuracy | ± 0.1 °C | Met the specification. | Bench performance testing | Confirms precise temperature control. |
| CO2 accuracy (5.0% setting) | ± 0.1% | Met the specification. | Bench performance testing | Confirms precise gas concentration control. |
| CO2 accuracy (other settings) | ± 0.2% | Met the specification. | Bench performance testing | Confirms precise gas concentration control. |
| O2 accuracy (5.0% setting) | ± 0.1% | Met the specification. | Bench performance testing | Confirms precise gas concentration control. |
| O2 accuracy (other settings) | ± 0.2% | Met the specification. | Bench performance testing | Confirms precise gas concentration control. |
| CO2 recovery (5% ± 0.2 %, after 30s door opening) | ＜3.5 min | Met the specification. | Bench performance testing | Confirms rapid recovery of CO2 levels after brief opening. |
| O2 recovery (5% ± 0.2 %, after 30s door opening) | ＜3.5 min | Met the specification. | Bench performance testing | Confirms rapid recovery of O2 levels after brief opening. |
| Recirculation rate (Host only, self-mixed) | > 49 L/h (full purification every 2min) | Met the specification. | Bench performance testing | Ensures effective gas circulation and purification. |
| Recirculation rate (Host + Expansion, self-mixed) | > 24.5 L/h (full purification every 4min) | Met the specification. | Bench performance testing | Ensures effective gas circulation and purification with expansion module. |
| Resolution (Microscope) | >3 pixels per μm | Met the specification (compared to predicate's 3 pixels/µm, which is "Same"). | Bench performance testing (Imaging testing) | Indicates image clarity and ability to resolve fine details. |
| Total light exposure / embryo / day | Endotoxin testing | Ensures the device is non-toxic to embryos from bacterial endotoxins. | ||
| Sterility (SAL) | 10^-6 | Met the specification. | Radiation sterilization and validation testing | Confirms the culture dish is sterile. |
| Shelf-life | 2 years | Met the specification (via accelerated aging). | Accelerated aging (ASTM F1980:2021) | Ensures the device maintains its properties over its shelf life. |
| Humidification Box Parameters | ||||
| Sterility (SAL) | 10^-6 | Met the specification. | Radiation sterilization and validation testing | Confirms the humidification box is sterile. |
| Use-life | 30 days | Met the specification. | Continuous humidification testing | Confirms functionality over its intended use period. |
| Shelf-life | 2 years | Met the specification (via accelerated aging). | Accelerated aging (ASTM F1980:2021) | Ensures the device maintains its properties over its shelf life. |
| Software Parameters | ||||
| Software function/performance | N/A (General V&V) | Verified and validated | Software verification and validation per FDA Guidance (2023) | Confirms software functions as intended and safely. |
| Cybersecurity | N/A (General Evaluation) | Evaluated | Cybersecurity evaluation per FDA Guidance (2023) | Addresses potential cybersecurity risks. |
| General Safety | ||||
| Electrical Safety | Compliant with IEC 60601-1:2005/AMD2:2020 | Met the standard. | Electrical safety testing | Ensures the device is electrically safe. |
| Electromagnetic Compatibility (EMC) | Compliant with IEC 60601-1-2:2014+A1:2020, IEC TS 60601-4-2: 2024, and 2022 FDA Guidance | Met the standards. | EMC testing | Ensures the device does not interfere with or is not affected by other electronic devices. |
Study Details (Based on the provided non-clinical information)
1. Sample sizes used for the test set and the data provenance:
- Sample Sizes: The document does not specify exact sample sizes for each bench test (e.g., how many incubators were tested for temperature accuracy, or how many culture dishes for MEA). However, it implies that sufficient samples were tested to meet the specified standards and guidances.
- Data Provenance: The tests are non-clinical, primarily conducted as bench performance testing and in-vitro assays. Therefore, "country of origin of the data" generally refers to the location of the testing facility, not patient data origin. The document does not specify the testing facility's location, but the applicant company is in China. The data (results) are retrospective relative to the 510(k) submission date.
2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- Not Applicable in the traditional sense for this device. Given that the studies are primarily bench performance testing, in-vitro biological assays (MEA, Endotoxin), and engineering validations (electrical safety, EMC, software V&V), "ground truth" is established by adherence to recognized international standards (e.g., ISO, IEC, ASTM, USP) and FDA guidances. The "experts" would be the qualified personnel performing these standardized tests and the internal experts validating compliance to design specifications. Their qualifications would be in engineering, microbiology, quality assurance, etc., as appropriate for each specific test. There is no mention of a ground truth established by medical experts (e.g., radiologists) for image interpretation.
3. Adjudication method (e.g. 2+1, 3+1, none) for the test set:
- Not Applicable. Adjudication methods like 2+1 or 3+1 are typically used in clinical studies where multiple human readers assess medical images or data, and their disagreements need to be resolved to establish a definitive "ground truth" (e.g., presence or absence of a disease). Since this document describes non-clinical engineering and in-vitro performance testing of a device hardware and software, such adjudication methods are not relevant.
4. 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 such study was done or described. The device includes a "user-defined modeling function" to aid in embryo selection, but there is no mention of a clinical MRMC study evaluating the effectiveness of human embryologists assisted by this software feature compared to unassisted embryologists. The validation focuses on the safety and performance of the incubator's physical parameters, image acquisition, and software functionalities (display, storage, basic user annotation/modeling), not a clinical outcome related to increased human reader performance.
5. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:
- No "standalone algorithm performance" in the typical diagnostic AI sense. The document describes software functions for displaying, storing, and users annotating/modeling, and performing calculations based on image data. This implies a tool to aid the user, not a standalone AI algorithm that provides a diagnosis or makes a definitive embryo selection recommendation without human intervention. The performance testing for the software was "Software verification and validation per the 2023 FDA Guidance Document, 'Content of Premarket Submissions for Device Software Functions'," which assures the software functions as intended and is safe, but not a standalone evaluative study of an AI's diagnostic accuracy.
6. The type of ground truth used:
- For the physical parameters of the Incubator (temperature, gas, imaging): The "ground truth" is based on engineering specifications and measurements conforming to established industry standards (e.g., accuracy, stability, recovery times).
- For the Hariomed culture dish and humidification box: The "ground truth" is established through biological assays (MEA, Endotoxin) and sterilization validation according to a priori acceptance criteria defined by regulatory standards (e.g., ≥80% blastocyst development, ≤0.5 EU/device endotoxin, SAL 10^-6).
- For software: The "ground truth" is established through software verification and validation testing against its design specifications and functional requirements to ensure it performs as intended and is safe.
7. The sample size for the training set:
- Not Applicable. This document does not describe the development or training of a machine learning or deep learning model. The "user-defined modeling function" implies that users define criteria/models based on their expert knowledge and annotated data, rather than a system trained on a large dataset. Therefore, there is no "training set" in the context of an AI algorithm described here.
8. How the ground truth for the training set was established:
- Not Applicable (as there is no described AI training set). If the "user-defined modeling function" were to incorporate any form of machine learning in its development (which is not explicitly stated in the summary), the ground truth for such hypothetical training would typically come from expert embryologist annotations of embryo development parameters and correlation with clinical outcomes. However, the provided document does not delve into this level of detail. The emphasis is on the software being a tool for user annotation and modeling, not an internally generated AI recommendation.
§ 884.6120 Assisted reproduction accessories.
(a)
Identification. Assisted reproduction accessories are a group of devices used during assisted reproduction procedures, in conjunction with assisted reproduction needles and/or assisted reproduction catheters, to aspirate, incubate, infuse, and/or maintain temperature. This generic type of device may include:(1) Powered aspiration pumps used to provide low flow, intermittent vacuum for the aspiration of eggs (ova).
(2) Syringe pumps (powered or manual) used to activate a syringe to infuse or aspirate small volumes of fluid during assisted reproduction procedures.
(3) Collection tube warmers, used to maintain the temperature of egg (oocyte) collection tubes at or near body temperature. A dish/plate/microscope stage warmer is a device used to maintain the temperature of the egg (oocyte) during manipulation.
(4) Embryo incubators, used to store and preserve gametes and/or embryos at or near body temperature.
(5) Cryopreservation instrumentation and devices, used to contain, freeze, and maintain gametes and/or embryos at an appropriate freezing temperature.
(b)
Classification. Class II (special controls) (design specifications, labeling requirements, and clinical testing). The device, when it is a simple embryo incubator with only temperature, gas, and humidity control; a syringe pump; a collection tube warmer; a dish/plate/microscope stage warmer; a controlled-rate cryopreservation freezer; or an assisted reproduction laminar flow workstation is exempt from the premarket notification procedures in subpart E of part 807 of this chapter subject to the limitations in § 884.9.