Notifications
Clear all
Hardware Forum
3
Posts
2
Users
0
Likes
49
Views
Mar 30, 2023 9:40 pm
Hi,
I am testing an fNIRS prototype (made with ADS1299, red 735 nm LED, infrared 850 nm LED, and Vishay Semiconductor's VBPW34SR photodiodes). The data below is taken with a red LED that is continuously on for 2 minutes while the PD is recording. The LED and PD were on the forehead with a separation of 0.5cm.
In the raw data, the heart rate is not clearly visible, and the data appears fairly distorted, From the fft, the majority frequency component is less than 1 Hz.
However, when I highpass filter the data with a cutoff at 1Hz, the distortion decreases by a lot, and heart rate becomes somewhat visible. The smaller the cutoff is (0.01, 0.1 Hz) the more distorted it is.
All plots can be seen in the attached PDF.
Does this indicate an issue in the hardware, or can it be taken care of with digital signal processing?
What further tests should I try for debugging?
Any thoughts would be much appreciated!!
2 Replies
Apr 01, 2023 2:17 pm
At that separation you should have an obvious heart rate signal in your recorded signal. What is the optical power you have from the LED?
Apr 02, 2023 2:31 am
Thank you for taking the time to respond to my post!
I am sorry the figures are not visible, I have added them to a document that is viewable here: https://docs.google.com/document/d/1d_MUGyrnxGscwmeS_NNY4RnBa_t7kwH5AsOUywFjy5w/edit?usp=sharing
The plots in this are from a 735 nm LED that is continuously on.
The PD reading varies depending on the filter cutoff. I am using an Everlight Electronics's 67-21S/NFR2C-P2050A3B21522Z6/2T
which as radiometric power varying from 15-50 mW.