Classification of Preeclamptic Placental Extracellular Vesicles Using Femtosecond Laser Fabricated Nanoplasmonic Sensors

Abstract

Placental extracellular vesicles (EVs) play an essential role in pregnancy by protecting and transporting diverse biomolecules that aid in fetomaternal communication. However, in preeclampsia, they have also been implicated in contributing to disease progression. Despite their potential clinical value, current technologies cannot provide a rapid and effective means of differentiating between healthy and diseased placental EVs. To address this, a fabrication process called laser-induced nanostructuring of SERS-active thin films (LINST) was developed to produce scalable nanoplasmonic substrates that provide exceptional Raman signal enhancement and allow the biochemical fingerprinting of EVs. After validating the performance of LINST substrates with chemical standards, placental EVs from tissue explant cultures were characterized, demonstrating that preeclamptic and normotensive placental EVs have classifiably distinct Raman spectra following the application of advanced machine learning algorithms. Given the abundance of placental EVs in maternal circulation, these findings encourage immediate exploration of surface-enhanced Raman spectroscopy (SERS) of EVs as a promising method for preeclampsia liquid biopsies, while this novel fabrication process will provide a versatile and scalable substrate for many other SERS applications.

Cascaded Deep Convolutional Neural Networks as Improved Methods of Preprocessing Raman Spectroscopy Data

“…The resulting accuracy is also comparable to the results from a state of art deep CNN using raw data and a bottleneck classifier (BC). 22,31 The proposed method achieved this while also reducing the training time by several orders of magnitude and providing interpretable spectral information for more specific biochemical comparisons. This is because the conventional classification methods are not trained using an augmented data, which is required for the other methods based on deep CNN classification of the raw data.…”

“…The second example is the Raman hyperspectral imaging of a chemical species droplet with a very low concentration (10 −6 M) released over the boundary of a previously pattered SERS substrate (known as LINST), 31 using the same confocal Raman microscope as the placental EV SERS experiment but with a signal acquisition time of 100 ms, 5× objective, and 532 nm laser. This is considered a difficult task for conventional prepossessing techniques as it deals with signals of various amplitudes and baselines and contains tens of thousands of spectra, such that manual tuning of hyper parameters for each spectrum is practically impossible.…”

“…To demonstrate the superiority of our proposed methods, three different examples were used which would have previously required extensive manual prepossessing. The first example includes recently published SERS spectra of placental EVs from tissue explant cultures, obtained using a Horiba LabRam Evolution confocal Raman microscope, 50× objective, 10 s acquisition time, and 785 nm laser . As previously mentioned, EV SERS spectra are highly diverse due to the heterogeneous cargo of each EV and their random placement relative to the SERS surface. − As a result, manual spectral preprocessing is often required to handle the wide range of noise and baselines.…”

“…Two examples of the spectra in the placental EV SERS data set showing (a) a spectrum with high SNR, (b) a spectrum with low SNR, and (c) the preprocessed results of the spectra depicted in (a) and (b) using the ResNet-based deep learning method, and (d) comparison of deep CNN classification of raw data and conventional machine learning classification accuracy. Data for conventional machine learning classification are shown when preprocessing is done by either AsLS or the ResNet-based deep learning preprocessing.…”

“…However, their proposed U-Net method could not perform denoising and baseline correction independently, preventing the direct quantitative comparison of their method to the conventional preprocessing counterparts. This also limits its application in the rapidly growing field of SERS characterization of extracellular vesicles (EVs), − in which the produced signals vary greatly in intensity, and recovering low signal-to-noise ratio (SNR) spectra can generate false peaks. This issue can potentially be avoided by having access to the signal that has only undergone baseline correction prior to denoising, but would require the development and validation of a new deep CNN preprocessing architecture.…”