(91 days)
The SPARK Scan is intended to acquire, display, and store the electrical activity of a patient's brain obtained by placing two or more electrodes on the head to aid in diagnosis.
The SPARK Scan is intended for the acquisition, display, and storage of electroencephalogram (EEG) data. The SPARK Scan is intended to be used by EEG technicians, or appropriately trained Nurses and Medical Assistants practicing in any medical setting where EEG data collection may be required.
The SPARK Scan is an SaMD product that consists of the Edge Software and Data Storage and Communication Platform.
The Edge software runs locally on the user's device and consists of a User interface and the EEG hardware interface, including the electroconductive gel (K1117). The User Interface is a desktop application that (a) manages interaction with the EEG hardware interface, (b) facilitates set up and recording of EEG data, and (c) provides limited access to review of patient records according to the user's permissions.
The Data Storage and Communication Platform (Cloud Software): The Cloud software runs on a server managed by SPARK Neuro and contains no user interface. The Cloud Software is responsible only for managing (a) authorization and authentication of users and (b) storage, validation, and access to all data collected on the system.
The SPARK Scan is compatible with three 3rd-party accessory devices: an FDA-cleared EEG hardware system, a standard off-the-shelf-laptop, and FDA-cleared EEG Recording Viewing Platforms.
The provided text is a 510(k) summary for the SPARK Scan device. It primarily focuses on demonstrating substantial equivalence to a predicate device rather than providing a detailed study proving the device meets specific acceptance criteria based on performance metrics. The device is an electroencephalograph (EEG) intended to acquire, display, and store brain electrical activity to aid in diagnosis.
Here's an analysis of the provided information concerning acceptance criteria and study details:
1. A table of acceptance criteria and the reported device performance
The document does not explicitly state acceptance criteria in terms of performance metrics (e.g., accuracy, sensitivity, specificity) for the SPARK Scan's ability to "aid in diagnosis." Instead, the "Performance Testing Summary" section indicates:
- Non-Clinical Testing:
- Software verification and validation testing
- Cybersecurity Risk Analysis and Testing
- Human Factors Use Related Risk Analysis
- Biocompatibility: Not applicable (no patient contact).
- Electrical Safety and Electromagnetic Compatibility: Not applicable (no hardware).
- Animal Testing: Not required.
- Clinical Testing: Not required.
The comparison table between the subject device (SPARK Scan) and the predicate device (Mitsar-EEG) highlights technological characteristics rather than performance metrics against acceptance criteria.
| Feature | Acceptance Criteria (Implied from Predicate Comparison) | Reported Device Performance (SPARK Scan) |
|---|---|---|
| Intended Use | Acquire, display, store brain electrical activity to aid in diagnosis. | Acquire, display, store brain electrical activity to aid in diagnosis. (Identical to predicate) |
| Intended User | Medical Staff | Medical Staff (Identical to predicate) |
| Target Population | Adults | Adults (Identical to predicate) |
| Use Environment | Healthcare Facilities | Healthcare Facilities (Identical to predicate) |
| Signal Acquisition | Yes, EEG | Yes, EEG (Identical to predicate) |
| Number of Recording Channels | Up to 21 (Predicate) | Up to 32 (Similar to predicate, subject device has more) |
| Impedance Test | Yes | Yes (Identical to predicate) |
| Sampling Rate | 500Hz | 500Hz (Identical to predicate) |
| Interface | PC | PC (Identical to predicate) |
| Software V&V | Successful completion of V&V testing. | Completed software verification and validation testing. |
| Cybersecurity | Successful completion of risk analysis and testing. | Completed Cybersecurity Risk Analysis and Testing. |
| Human Factors | Successful completion of use-related risk analysis. | Completed Human Factors Use Related Risk Analysis. |
2. Sample size used for the test set and the data provenance (e.g. country of origin of the data, retrospective or prospective)
The document states, "Clinical testing is not required to support substantial equivalence." This implies there was no clinical test set of patient data used for performance claims. The "test set" for the reported performance appears to be related to software verification and validation, cybersecurity, and human factors, which are not described using patient data.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts (e.g. radiologist with 10 years of experience)
Since no clinical testing on patient data with diagnostic outcomes was performed or required, there is no mention of experts establishing a ground truth for a test set in the context of diagnostic aid. The ground truth for software testing would be defined by specifications and requirements, not expert interpretation of medical data.
4. Adjudication method (e.g. 2+1, 3+1, none) for the test set
No adjudication method is described because no clinical test set requiring expert consensus for ground truth was used.
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 MRMC study was conducted or is mentioned. The device is described as software for acquiring, displaying, and storing EEG data, not as an AI-assisted diagnostic tool that aids human readers in interpretation. Clinical testing was deemed "not required to support substantial equivalence."
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
The SPARK Scan is described as an SaMD (Software as a Medical Device) product comprising Edge Software and a Data Storage and Communication Platform, compatible with third-party EEG hardware. Its function is to acquire, display, and store EEG data. It is not presented as an "algorithm only" device that provides a diagnostic output in a standalone capacity without human interpretation. It aids in diagnosis by providing raw EEG data, requiring "Medical Staff" (Intended User) for interpretation. Clinical testing was not required for its clearance.
7. The type of ground truth used (expert concensus, pathology, outcomes data, etc)
For the non-clinical testing conducted (software V&V, cybersecurity, human factors), the ground truth would be defined by engineering specifications, security protocols, and human factors engineering principles, respectively. No medical ground truth (e.g., expert consensus on diagnosis, pathology, or outcomes data) was used or required for this submission.
8. The sample size for the training set
The document does not mention a training set for machine learning or AI algorithms, as the device's function is described as data acquisition, display, and storage, not as an AI diagnostic tool. Therefore, sample size for a training set is not applicable or provided.
9. How the ground truth for the training set was established
Not applicable, as no training set for machine learning or AI is mentioned.
§ 882.1400 Electroencephalograph.
(a)
Identification. An electroencephalograph is a device used to measure and record the electrical activity of the patient's brain obtained by placing two or more electrodes on the head.(b)
Classification. Class II (performance standards).