Dynamic optical coherence tomography algorithm for label-free assessment of swiftness and occupancy of intratissue moving scatterers

TL;DR

This work tackles the interpretability gap in dynamic optical coherence tomography by introducing two open-source metrics, aLIV and Swiftness, derived from the LIV curve to quantify dynamic-scatterer occupancy and scatterer speed, respectively. Through numerical simulations based on a dispersed-scatterer model and a dedicated DOCT simulation framework, the authors demonstrate that aLIV correlates with the dynamic-scatterer ratio while Swiftness tracks motion speed, with OCDS acting as a speed fingerprint. They validate these metrics experimentally on in vitro tumor spheroids and an ex vivo mouse kidney, showing improved interpretability over conventional LIV and OCDS and robustness to acquisition-window size. The methods are released as open-source tools, enabling broader adoption and future validation across tissue types and motion models.

Abstract

Dynamic optical coherence tomography (DOCT) statistically analyzes fluctuations in time-sequential OCT signals, enabling label-free and three-dimensional visualization of intratissue and intracellular activities. Current DOCT methods, such as logarithmic intensity variance (LIV) and OCT correlation decay speed (OCDS) have several limitations.Namely, the DOCT values and intratissue motions are not directly related, and hence DOCT values are not interpretable in the context of the tissue motility. We introduce a new DOCT algorithm that provides more direct interpretation of DOCT in the contexts of dynamic scatterer ratio and scatterer speed in the tissue.The detailed properties of the new and conventional DOCT methods are investigated by numerical simulations, and the experimental validation with in vitro and ex vivo samples demonstrates the feasibility of the new method.

Figures (11)

• Figure 1: Schematic illustrating the acquisition-time-window (${A_{tw}\xspace}$)-size dependency of LIV. LIV inevitably decreases as ${A_{tw}\xspace}$ becomes smaller. Curves A, B, and C represent the cases of faster, moderate, and slower signal fluctuations, respectively.
• Figure 2: Diagram showing the principle of the new DOCT algorithm. (i) OCT frames are repeatedly acquired with a time interval of $\Delta t$ over ${A_{tw}\xspace}$. (ii) All combinations of data subsets are extracted for different time window (${T_w\xspace}$) sizes from the time-sequential OCT frames. (iii) LIV is computed from each data subset. (iv) All LIV values with identical ${T_w\xspace}$ are averaged. (v) A curve of the averaged LIV over ${T_w\xspace}$ is obtained, which is referred to as the "LIV curve". (vi) The LIV curve is fitted with a first-order saturation function of ${T_w\xspace}$.
• Figure 3: The scatterer-speed dependencies (a, c, e, g) and dynamic-scatterer-ratio (DSR) dependencies (b, d, f, h) of aLIV, LIV, Swiftness, and OCDS obtained by numerical simulations. The point colors of the DSR-dependency plots indicate the speed of the dynamic scatterers as shown in the legend in (b). In the magnified plot for the speed range [0, 1 $\muup$m/s] (g), the peak location of OCDS is indicated with the arrowhead.
• Figure 4: The en face aLIV, LIV, Swiftness, OCDS, and fluorescence images of the tumor spheroids treated by the anti-cancer drug (1-$\muup$M PTX) for 1, 3, and 6 days. The scale bars represent 100 $\muup$m. In the region with high aLIV (green), the dynamic scatterer ratio is expected to be high, while in the region with high Swiftness, the dynamic scatterers are expected to move rapidly. Examples of LIV curves are shown in (u), which were extracted at the points indicated by the arrows in (v, w). (v, w) correspond to the square regions in (a, c).
• Figure 5: The en face and B-scan aLIV, LIV, Swiftness, and OCDS images of the ex vivo mouse kidney. The scale bars represent 1 mm. The magnified images of LIV and Swiftness (i, j) correspond to the square regions in the B-scan images (f, g), respectively. Several pipe-like structures are visible, and they exhibit high aLIV values, which may indicate high occupancy of dynamic scatterers (high dynamic scatterer ratio), and low Swiftness values, suggesting the motion of the dynamic scatterers is slow.