Figure 5 . Total number of cracks and pits observed in the degrading bloodstains at various timepoints (in minutes).
4. Discussion
Overall, our findings corroborate and are supported by the literature on bloodstain drying, suggesting that optical profilometry is a viable technique to observe and characterize bloodstain drying and degradation. In the short-term experiments, scans showed significant changes within the first 35-40 minutes after deposition, after which the surface characteristics plateaued. The pre-gelation phase begins as soon as the droplet is deposited onto the plate, however, many of the surface characteristics could not be obtained immediately due to the liquid-state of the bloodstain. Topographical scans could detect the edge of bloodstains five minutes after deposition, signifying the beginning of drying. Within the first 15-20 minutes of drying, height profiles and maximum height observations showed a significant change, with differences as large as 60 µm within a 10-minute timespan. The change(s) in surface roughness, skewness, kurtosis and maximum height of the bloodstain over time is primarily driven by RBC movement [10,23]. As the RBCs gather along the edge of the bloodstain to form the rim, the maximum height will increase and the skewness and kurtosis values will change. The height distributions corroborate this and show that there is a decrease, then an increase in overall height as the rim of the bloodstain begins to form. The surface average roughness values could be correlated to RBC movement during drying. At earlier time points (before 20 minutes), the bloodstains contained no cracks, and the maximum height was in the centre of the bloodstain (Figure 2). As the RBCs moved towards the outside of the bloodstain and the surface temporarily flattened out, the surface average roughness value decreased. Then, as the rim began to form, the surface average roughness increased and cracks and pits formed. Cracks continued to form until ~30-35 minutes after deposition; all surface characteristics and height profile data then displayed a plateau, suggesting that no further changes occurred. The analysis of the area of cracks and pits over time was completed as a quantitative complement to the qualitative observations made using the topographical scans. Importantly, the profilometer will not interpolate a crack or pit. If a crack is present, but only one surface point is detected, it might interpolate the crack as much smaller than it is and, in some cases, it might only appear as one pixel after the scan is interpolated. This means that some cracks are larger than what the analysis determined; however, we decided rather than focus on the absolute value of a single crack/pit, the trend in the size of cracks and pits over time is more relevant. Similar to our work, Brutin et al. (2011) [23] found the drying time of approximately 10 µL of human blood to be 36 minutes (temperature = 22 ºC), while Ramsthaler et al. [8] found the drying time of ~25 µL of human blood to be 34 minutes (temperature = 24 ºC). Although outside the scope of this study, these findings suggest that different blood sources (bovine vs. human) display comparable drying mechanisms. The inclusion of an anticoagulant also can account for differences in drying time.
From an imaging and spectroscopy perspective, it is important to understand the surface profiles of bloodstains for TSD estimates. Where the light source is focused could have spectral implications – for example, unwanted scattering. Having an idea of surface roughness changes and cracks and pits formation over time may help explain observed spectral variance beyond inter- and intra-specific variation between bloodstain samples. While there is a significant effect of spectral signatures with time, the variation between TSD models is also high due to substrate-dependent interactions and environmental effects. However, given the large observed variance in surface topography of the dried bloodstain, we wonder how much this also influences models, particularly those that use solid-state imaging techniques.