![]() ![]() The schematic of our proposed setup ( Figure 1a) is as shown in Figure 1. Thus, the LSIT platform saves time and computation resources required to learn to classify the additional cell types along with the existing ones. Here, we have first introduced the transfer of learning scheme in a neural network, which can leverage the feasibility to introduce new cell types to the algorithm and thus learn their characteristics within a few iterations. ![]() ![]() Further, we have developed an auto characterization method based on a convolutional neural network (CNN) architecture to classify the various cell lines from the LSIT micrograph. For this, we employed the autoencoder-based denoising scheme. To address these limitations, in this work, we have developed an artificial intelligence (AI) powered signal enhancement scheme for the LSIT micrographs that can enhance the signal quality (signal to noise ratio (SNR)) for various cell lines in a heterogeneous cell sample. Further, the handcrafted approach of finding the features for every additional cell line is time-consuming and prone to subjective errors. Since the diffraction signature of a microparticle depends on the size as well as the signal-to-noise ratio of the particle, therefore any background noise can affect the overall performance of the auto characterization system. However, the performance of the system is dependent on the uniform illumination as well as the strong signatures of the microparticle samples. In our previous work, we have successfully developed the LSIT imaging system for the complete blood count using an analytical model based on handcrafted features that can automatically segment out the individual cells from a whole frame LSIT micrograph and subsequently analyze them based on the handcrafted parameters. This simple and cost-effective nature facilitates the feasibility of the LSIT for the applications in the fields of point-of-care systems or telemedicine systems. Since this arrangement consists of a few components, most of which are easily available at a low price, it therefore reduces the overall cost of the system. The absence of a lens or other optical arrangements allows it to fit into a very small space, thereby reducing the size of the overall system (as described in Figure 1a in the LSIT platform (Cellytics) built within a dimension of 100 × 120 × 80 mm 3). It comprises a lens-less detector, such as a complementary metal-oxide semiconductor (CMOS) image sensor, a semi-coherent light source, such as light-emitting diode (LED), and a disposable cell chip (C-Chip). This technique is widely popular for its simple imaging structure and cost-effectiveness. The lens-free shadow imaging technique (LSIT) is a well-established technique for the characterization of microparticles and biological cells. Furthermore, the model is adaptive to learn new type of samples within a few learning iterations and able to successfully classify the newly introduced sample along with the existing other sample types. The performance of our proposed method shows an increase in accuracy >98% along with the signal enhancement of >5 dB for most of the cell types, such as red blood cell (RBC) and white blood cell (WBC). In this work, we address these issues by leveraging the artificial intelligence-powered auto signal enhancing scheme such as denoising autoencoder and adaptive cell characterization technique based on the transfer of learning in deep neural networks. Further, its performance suffers from poor image (cell diffraction pattern) signatures due to their small signal or background noise. The developed auto characterization algorithm so far for this custom-developed LSIT cytometer was based on the handcrafted features of the cell diffraction patterns from the LSIT cytometer, that were determined from our empirical findings on thousands of samples of individual cell types, which limit the system in terms of induction of a new cell type for auto classification or characterization. Due to its simplicity and cost-effectiveness, various low-cost solutions have been developed, such as automatic analysis of complete blood count (CBC), cell viability, 2D cell morphology, 3D cell tomography, etc. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |