(G-242) Imaging Photoplethysmography to Determine Cutaneous Blood Volume Changes in Response to Tactile Forces

Tactile feedback is essential for movement control, particularly for dexterous movement of the hands where skin mechanoreceptors are at their highest density. Restoring functional movements after paralysis through engineered connections between the brain and body (i.e. neuroprostheses) similarly requires a mechanism to convey a sense of touch. This is achieved by stimulating somatosensory brain areas in response to tactile signals acquired by sensors on the desired effector, such as robotic hands [1]. We previously developed an implantable artificial mechanoreceptor (IAM) to acquire tactile signals [2]. The IAM was a wireless hermetically sealed device, capacitively sensing forces acting on the skin from a location under the skin [3]. However, sensing mechanical deformations transmitted to the subcutaneous space, well below the location of natural mechanoreceptors, may limit the achievable sensitivity. As an alternative, it is readily observable that forces transiently affect skin color due to blood volume changes in the compressed skin capillaries. This effect has been used previously to sense fingertip forces from optical measurements on the fingernail [4]. We hypothesize that a miniature subcutaneous device could optically interrogate the blood volume of the overlying dermis to yield a sensitive tactile sensor for neuroprosthetic applications. For rational design of this sensor, it is critical to understand the spatiotemporal properties of blood volume changes in response to tactile forces. Imaging photoplethysmography (iPPG) is a technique traditionally used to obtain high field-of-view (FOV) images of the underlying circulation [5]. Here, we report the design of an iPPG system to make these measurements.

The major components of our iPPG system are the 3.2 Megapixel RGB camera (Blackfly BFS-U3-32S4C-C, Teledyne FLIR) with resolution 2048x1536, a 3.5mm Lens (Max aperture F2, Edmund Optics), and an LED ring (RL5604 WHITE, Advanced Illumination).The camera can receive a spectral band of 400-700nm. The lens is specified for a working distance (WD) of 0- ∞ mm with a horizontal FOV of 102.4° and vertical of 82.3°. A 3D-printed clear force probe with a hemispherical tip (dia = 5mm) was mounted to the end of the lens and measured 16 mm from the tip to the convexity of the lens. 

The camera system with the probe was positioned at direct contact with the skin surface. The subject’s arm was placed underneath the camera and the ring LED was turned on prior to acquisition. A laptop was connected to the camera and the Spinnaker SDK (SpinView GUI) was used to acquire a 20 second recording of the human subject’s hand. The recording showed the probe starting from the center of the palm, moving medially to the level of the 5th digit, then translating distally to the MCP joint. The video was then processed with spatial and temporal normalization, the Plane-Orthogonal-to-Skin (POS) method, and a frequency filter with a bandwidth of 0.0095-0.145Hz, encompassing physiologic bands of myogenic, neurogenic, and endothelial activity [5-6]. The output of this process produced a normalized cutaneous blood volume (CBV) signal, indicative of the presence of microcirculatory blood in a given area.

Results:

Discussion:

Prior studies showed that iPPG can detect AC signals at physiological heart rate (0.67-4Hz), but the results of this study showed that iPPG can also resolve high-resolution images of microcirculation at near-DC bandwidths without significant noise.  The results highlighted that the application of normal forces, as low as < 10mmHg sustained over 1-2 seconds, were responsible for initial movement of blood away from the point of compression (POC), followed by increased perfusion. The rush in perfusion occurred over 1 second intervals from the surrounding tissue microcirculation. This was consistent across all time points and demonstrated pressure-induced vasodilation (PIV), a result of myogenic activity at compressed precapillary sphincters [7]. Under myogenic control, the distal arterioles increase blood flow to capillaries following compression.

In shear, a “blue void” can be seen trailing the POC indicating reduced CBV as the adaptive vasodilatory mechanisms lag behind the moving compression wave. This lag is expected as the myogenic control and endothelial metabolic activity will be slower than the probe traveling at ~1cm/sec along the palm. Notably, this compression wave was not followed by increased CBV to the “void” tissues. Instead the surrounding microcirculation prioritized supplying the tissue receiving sustained normal forces as part of PIV.

Conclusion:

We have demonstrated that pressure-induced changes to the microcirculation perfusion can be visualized with previously documented autoregulation methods playing a part in the observed phenomena. Our iPPG implementation was useful in showing microcirculatory adaptations to pressure and is a necessary step towards designing an optical IAM.

1. 
   

   Flesher SN, Downey JE, Weiss JM, Hughes CL, Herrera AJ, Tyler-Kabara EC, Boninger ML, Collinger JL, Gaunt RA. A brain-computer interface that evokes tactile sensations improves robotic arm control. Science. 2021;372(6544):831-6. Epub 2021/05/22. doi: 10.1126/science.abd0380. PubMed PMID: 34016775.

   
   


2. 
   

   Hao H, Du L, Richardson AG, Lucas TH, Allen MG, Van der Spiegel J, Aflatouni F. A Wireless Artificial Mechanoreceptor in 180-nm CMOS. IEEE Transactions on Microwave Theory and Techniques. 2021;69(6):2907-20. doi: 10.1109/TMTT.2021.3072398.

   
   


3. 
   

   Du L, Hao H, Ding Y, Gabros A, Mier TCE, Van der Spiegel J, Lucas TH, Aflatouni F, Richardson AG, Allen MG. An implantable wireless tactile sensing system. Preprint available at Research Square. https://doi.org/10.21203/rs.3.rs-2515082/v1.

   
   


4. 
   

   Sun Y, Hollerbach JM, Mascaro SA. Predicting fingertip forces by imaging coloration changes in the fingernail and surrounding skin. IEEE Trans Biomed Eng. 2008;55(10):2363-71. Epub 2008/10/08. doi: 10.1109/TBME.2008.925691. PubMed PMID: 18838361.

   
   


5. 
   

   Wang W, Brinker AC den, Stuijk S, Haan G de, "Algorithmic Principles of Remote PPG," in IEEE Transactions on Biomedical Engineering, vol. 64, no. 7, pp. 1479-1491, July 2017, doi: 10.1109/TBME.2016.2609282.

   
   


6. 
   

   Cheng D, Wang J, Yokota T, Someya T. Spatiotemporal processing in photoplethysmography for skin microcirculatory perfusion imaging, Biomed. Opt. Express 13, 838-849 (2022). DOI: 10.1364/BOE.442764

   
   


7. Bergstrand S, Lindberg LG, Ek AC, Lindén M, Lindgren M. Blood flow measurements at different depths using photoplethysmography and laser Doppler techniques. Skin Res Technol. 2009 May;15(2):139-47. doi: 10.1111/j.1600-0846.2008.00337.x. PMID: 19622122.