(166 days)
The Microlife Digital Infrared Forehead Thermometer, Model FR1MN1-1 is intended for the intermittent measurement and monitoring of human body temperature. The device is indicated for use by people of all ages in the home.
The Microlife Digital Infrared Forehead Thermometer, Model FR1MN1-1 is an electronic thermometer using an infrared sensor (thermopile) to detect body temperature from the forehead. Microlife FR1MN1-1 especially enables the user to take measurements and be alerted for possible elevated temperature readings accordingly.
The Microlife Digital Infrared Forehead Thermometer, Model FR1MN1-1, consists of mainly seven parts:
a) Thermopile Sensor
b) Application-Specific Intergrated Circuit
c) Erasable Programmable Read-Only Memory Integrated Circuit
d) Lens
e) LCD and Backlight
f) 2 Keys (Start key, O/I key)
g) 2 batteries AAA (LR03) 1.5V
The new Model FR1MN1-1 has the same intended use and temperature measurement fundamental algorithm as the 510(k) cleared devices.
Auto-Display memory The last reading is automatically displayed for 2 seconds when the unit is switched ON.
The provided document is a 510(k) premarket notification for a medical device, specifically the Microlife Digital Infrared Forehead Thermometer, Model FR1MN1-1. The primary purpose of this document is to demonstrate "substantial equivalence" to a legally marketed predicate device, not necessarily to detail comprehensive acceptance criteria and a deep dive into an "AI" algorithm's performance as would be expected for a novel AI/ML-based diagnostic device.
Therefore, the information available in this document is limited regarding the acceptance criteria and study details as they relate to AI/ML devices. The device described is a simple digital thermometer, not an AI-powered diagnostic tool. The "software algorithm" mentioned primarily refers to clinical offset tables for temperature conversion, not an AI model.
However, I will extract the relevant information as accurately as possible based on the provided text, interpreting "acceptance criteria" and "study" in the context of a thermometer's performance, particularly its accuracy.
Here's the breakdown:
Acceptance Criteria and Reported Device Performance
The core performance metric for a thermometer is its accuracy. The document refers to "Accuracy (Clinical Mode)" as a key characteristic.
1. Table of Acceptance Criteria and Reported Device Performance (as inferred for a thermometer)
| Metric | Acceptance Criteria (from predicate) | Reported Device Performance (for FR1MN1-1) |
|---|---|---|
| Accuracy (Clinical Mode) | ±0.3 °C, 34.0 ~ 42.2 °C (±0.5 °F, 93.2 ~ 108.0 °F) (Predicate) | 0.4 °F, 96.8 ~ 102.2 °F (FR1MN1-1) |
| Clinical Bias | Not explicitly stated as a numerical criterion, but evaluated | Evaluated (details not in summary) |
| Clinical Uncertainty | Not explicitly stated as a numerical criterion, but evaluated | Evaluated (details not in summary) |
| Clinical Repeatability | Not explicitly stated as a numerical criterion, but evaluated | Evaluated (details not in summary) |
| Compliance with Standards (e.g., ASTM E 1965-98 (2009), IEC 80601-2-56) | Pass/Fail Criteria Outlined in Standard | Passed All Testing Requirements (Implicitly meets the standard's accuracy and other criteria) |
Note on Accuracy: The reported accuracy for the new device (0.4 °F) compared to the predicate (±0.5 °F) suggests an improvement or at least an equivalent level of performance within a specific range. However, the predicate's accuracy is given as a range (e.g., 34.0 ~ 42.2 °C), while the new device's accuracy is given as a single value (0.4 °F) for a specific range (96.8 ~ 102.2 °F). This difference in reporting style makes direct comparison of the exact numerical criteria challenging from this summary section alone. The "Passed All Testing Requirements" for ASTM E 1965-98 (2009) and IEC 80601-2-56 is the key here, as these standards define the specific accuracy acceptance criteria for clinical thermometers.
Study Details
2. Sample size used for the test set and the data provenance
- Sample Size: The document states that "Controlled human clinical studies were conducted," but it does NOT specify the exact sample size (number of subjects/patients) used for the clinical validation.
- Data Provenance: Not explicitly stated, but clinical studies are generally conducted in a prospective manner. The country of origin of the data is not specified. Given the submitter's identification (Microlife Intellectual Property GmbH, Switzerland), the studies could have been conducted in Europe or elsewhere. It is described as "Controlled human clinical studies," which implies prospective data collection for the purpose of the study.
3. Number of experts used to establish the ground truth for the test set and the qualifications of those experts
- This question is not applicable in the context of a thermometer. For a thermometer, "ground truth" for body temperature is typically established by comparative measurements against a highly accurate reference thermometer (e.g., a calibrated rectal thermometer in a clinical setting), not by expert opinion. The "experts" would be the clinicians performing the comparative measurements, and their qualifications would be standard medical practice, not specialized expert consensus like for image interpretation.
4. Adjudication method for the test set
- This question is not applicable to a thermometer's performance evaluation. Adjudication methods (like 2+1, 3+1) are used for resolving disagreements among human readers/experts in subjective tasks (e.g., image interpretation for AI ground truth). For a thermometer, comparative measurements against a reference standard are objective.
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.
- This question is not applicable. The device is a thermometer, not an AI-assisted diagnostic tool for human readers. There is no human-in-the-loop AI assistance scenario for a digital thermometer.
6. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done
- Yes, a standalone performance evaluation was implicitly done. The "software algorithm" refers to the thermometer's internal processing of infrared data into a temperature reading. The clinical study evaluated the direct output of the device (the algorithm's output) compared to reference methods. The study presented "clinical bias, clinical uncertainty and clinical repeatability per clinical validation for Microlife FR1MN1-1," which are direct measures of the device's accuracy and precision, independent of human interpretation of its output. The "algorithm" here is simply the conversion of raw sensor data to a temperature display.
7. The type of ground truth used (expert consensus, pathology, outcomes data, etc)
- The ground truth for temperature measurement would be established by comparative measurements against a gold standard or reference thermometer, typically invasive methods like rectal temperature or highly accurate oral thermometers, in a controlled clinical setting. The document explicitly references "Controlled human clinical studies were conducted in accordance with ASTM E1965-98, IEC 80601-2-56 Test Report." These standards define the methodology for establishing accurate body temperature ground truth for thermometer validation.
8. The sample size for the training set
- The document implies that the "software algorithm" (specifically, the "clinical offset tables") was "adjusted to improve accuracy." While this hints at an iterative development process that might involve data, the document does NOT specify a distinct "training set" or its sample size. For a conventional device like this, the "training" might refer to engineers calibrating the offset tables based on empirical testing during development, rather than an explicit machine learning training process.
9. How the ground truth for the training set was established
- Given that the document does not mention a distinct "training set" in the machine learning sense, this question is not directly applicable. If "training" refers to the development and calibration of the offset tables, the ground truth would have been established through laboratory calibration and preliminary human clinical trials, similar to how the validation ground truth was established (i.e., comparison to reference thermometers).
§ 880.2910 Clinical electronic thermometer.
(a)
Identification. A clinical electronic thermometer is a device used to measure the body temperature of a patient by means of a transducer coupled with an electronic signal amplification, conditioning, and display unit. The transducer may be in a detachable probe with or without a disposable cover.(b)
Classification. Class II (performance standards). The device is exempt from the premarket notification procedures in part 807, subpart E of this chapter, subject to the limitations in § 880.9 and the following conditions for exemption:(1) Device is not a clinical thermometer with telethermographic functions;
(2) Device is not a clinical thermometer with continuous temperature measurement functions; and
(3) Appropriate analysis and testing (such as that outlined in the currently FDA-recognized editions, as appropriate, of ISO 80601-2-56, “Medical electrical equipment—Part 2-56: Particular requirements for basic safety and essential performance of clinical thermometers for body temperature measurement,” or ASTM E1965, “Standard Specification for Infrared Thermometers for Intermittent Determination of Patient Temperature,” or ASTM E1112, “Standard Specification for Electronic Thermometer for Intermittent Determination of Patient Temperature,” or ASTM E1104, “Standard Specification for Clinical Thermometer Probe Covers and Sheaths”) must validate specifications and performance of the device.