Search Results

The Ava Fertility Tracker is intended to measure and display physiological parameters (body temperature, resting pulse rate, heart rate variability, and breathing rate) as an aid in ovulation prediction to facilitate conception (not to be used for contraception).

Device Description

The Ava Fertility Tracker aids women in ovulation prediction and to facilitate conception. The Ava Fertility Tracker is a non-invasive device made up of the following components:
The Ava bracelet incorporating hardware with electronics and embedded software
A mobile application running on smartphones (iOS and Android)
A backend server software, including algorithms running on remote servers
The Ava bracelet is worn by women at night and tracks the following physiological parameters with built-in sensors:
Temperature sensor - provides information on skin temperature
Photoplethysmography (PPG) sensor – measures the interbeat interval (IBI) providing information on pulse rate (HR), heart rate variability (HRV), breathing rate (BR) and perfusion. Only HR, HRV, and BR are used by the algorithms for prediction of the fertile window. Perfusion information is used to assess the quality of the PPG sensor data.
Accelerometer provides information on movement and sleep duration and phases
The user synchronizes the bracelet to the mobile app each morning. The mobile app receives the raw datasets via Bluetooth Low Energy (BLE) and transfers this data to the backend server. The algorithm is run on the physiological parameters collected and the prediction of ovulation and the fertile window is determined. This information is transferred and displayed in the mobile app. The mobile app will also display a single mean value for HR, temperature, HRV ratio, BR, and sleep duration from the prior night.

The mobile application has three modes:
Trying to Conceive Mode - The application shows the user the predicted day of ovulation plus five predicted fertile days leading up to the predicted day of ovulation. Users receive the prior night summary of measured parameters displayed in graphs. Users can track additional information on their cycle, such as sexual intercourse or cervical fluid (this information does not feed into the algorithm).
Cycle Tracking Mode– For users not wanting to conceive but wanting to track their cycle. These users receive the same information provided to Trying to Conceive Mode users.
Pregnancy Mode - Users receive the prior night summary of measured parameters displayed on graphs. They can also track additional information as relates to their pregnancy, such as their weight or their daily water intake. For each week of their pregnancy, a different piece of informational content gets displayed in the application, such as typical height and weight of a baby at this stage of their pregnancy.

AI/ML Overview

Here's a breakdown of the acceptance criteria and study information for the Ava Fertility Tracker, based on the provided document:

Acceptance Criteria and Device Performance

Test Name	Specification	Reported Device Performance
Bench Testing
Transportation test	D.U.T. functional and no visual damage	Pass
Connector mating test	>1460 cycles	Pass
Battery Test 1	Battery retains >50% capacity after 730 charging/discharging cycles	Pass
Battery Test 2	Valid physiological data transferred to app after one night of data collection (at end of life with aged D.U.T.)	Pass
Cyclic strap elongation	Maximum 10% of straps break completely after 730 cycles of elongation and relaxation by 15%	Pass
Drop Test	D.U.T. functional and no visual damage after dropping from 1.5 m onto a hard surface (10 drops)	Pass
Shock test	D.U.T. functional and no visual damage after 15g acceleration applied (54 shocks)	Pass
Vibration test	D.U.T. functional and no visual damage after broad band random vibration excitation for 4.5h	Pass
Load Test	D.U.T. functional and no visual damage after 250N mechanical load for 1 hour	Pass
Perfume / insect repellent test	D.U.T. functional and no visual damage after 30-day exposure to perfume and insect repellent housing	Partial pass - Labeling to avoid exposure to ethanol-containing products.
Cosmetic cream test	D.U.T. functional and no visual damage after 30-day exposure to cosmetic cream housing	Pass
Sweat test	D.U.T. functional and no visual damage after 30-day exposure to artificial sweat housing	Pass
Temperature sensor validation	Precision: ± 0.2°C, Accuracy: ± 0.8°C (accelerated aging test)	Pass
Accelerometer sensor validation	Accuracy < ± 0.02 g, Precision < ± 0.001g (accelerated aging test)	Partial Pass – Precision met at all timepoints, while accuracy passed in aged devices. Justification provided in support of precision as the more important factor as the device only uses relative acceleration changes that are used in the sleep state computation.
PPG sensor Validation	PPG sensor signal ± 3% error for IBI and HR, and < 2.5% signal error for BR and HRV	Pass
Clinical Performance (Primary Endpoint)
Ovulation Day Prediction Error	Equivalence bounds of ±2 days compared to urinary LH tests as a reference.	Mean error of 0.315 days (95% Cl -0.029 to 0.698) (p-value for equivalence <0.001). This shows the device can predict the time of ovulation within the specified limits.
Clinical Performance (Secondary Endpoint)
Detection of Post-Ovulation Temperature Shifts	Comparison of temperature shifts detected by Ava Fertility Tracker (wrist skin temperature) vs. Lady-Comp (basal-body temperature).	The Ava Fertility Tracker detected significantly more post-ovulation temperature shifts compared to Lady-Comp (p<0.001).

Study Details for Clinical Performance:

Sample size used for the test set and the data provenance:
- Test Set Sample Size: The final analysis was performed on data from 61 women and 154 cycles.
- Data Provenance: The study was a single center, prospective clinical study. The country of origin is not explicitly stated, but Ava AG is based in Zürich, Switzerland, and Ava Science Inc. in San Francisco, USA. The study design ("single center, prospective") indicates primary data collection for this purpose.
Number of experts used to establish the ground truth for the test set and the qualifications of those experts:
- The document does not specify the number or qualifications of experts used to establish ground truth.
Adjudication method (e.g., 2+1, 3+1, none) for the test set:
- The document does not specify an adjudication method.
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 comparative effectiveness study involving human readers (as typically seen in image-based diagnostics) was not performed.
- The study compared the device's performance against a reference method (urinary LH tests) and a predicate device (Lady-Comp). It was not structured to assess human improvement with or without AI assistance in interpretation.
If a standalone (i.e., algorithm only without human-in-the-loop performance) was done:
- Yes, the clinical study essentially assessed the standalone performance of the Ava Fertility Tracker algorithm in predicting ovulation day, using "urinary LH tests as a reference." The device runs its algorithms on collected physiological parameters to determine predictions which are then displayed to the user.
The type of ground truth used (expert consensus, pathology, outcomes data, etc.):
- The primary ground truth for ovulation prediction was urinary LH tests.
- For the secondary endpoint (temperature shifts), the comparison was against the Lady-Comp device's basal-body temperature readings.
The sample size for the training set:
- The document does not provide information regarding the sample size for the training set used for the Ava Fertility Tracker's algorithms.
How the ground truth for the training set was established:
- The document does not provide information on how the ground truth for the training set was established.

Ask a Question

Ask a specific question about this device

Page 1 of 1