19,952 research outputs found
A 2.7-to-13.3μJ/boot/slot Flexible RNS-CKKS Processor in 28nm CMOS Technology for FHE-Based Privacy-Preserving Computing
No full text
Label-free sensor for automatic identification of erythrocytes using digital in-line holographic microscopy and machine learning
No full textCell types of erythrocytes should be identified because they are closely related to their functionality and viability. Conventional methods for classifying erythrocytes are time consuming and labor intensive. Therefore, an automatic and accurate erythrocyte classification system is indispensable in healthcare and biomedical fields. In this study, we proposed a new label-free sensor for automatic identification of erythrocyte cell types using a digital in-line holographic microscopy (DIHM) combined with machine learning algorithms. A total of 12 features, including information on intensity distributions, morphological descriptors, and optical focusing characteristics, is quantitatively obtained from numerically reconstructed holographic images. All individual features for discocytes, echinocytes, and spherocytes are statistically different. To improve the performance of cell type identification, we adopted several machine learning algorithms, such as decision tree model, support vector machine, linear discriminant classification, and k-nearest neighbor classification. With the aid of these machine learning algorithms, the extracted features are effectively utilized to distinguish erythrocytes. Among the four tested algorithms, the decision tree model exhibits the best identification performance for the training sets (n = 440, 98.18%) and test sets (n = 190, 97.37%). This proposed methodology, which smartly combined DIHM and machine learning, would be helpful for sensing abnormal erythrocytes and computer-aided diagnosis of hematological diseases in clinic.115sciescopu
Digital stereo-holographic microscopy for studying three-dimensional particle dynamics
No full textA digital stereo -holographic microscopy (DsHM) with two viewing angles is proposed to measure 3D information of microscale particles. This approach includes two volumetric recordings and numerical reconstruction, and it involves the combination of separately reconstructed holograms. The 3D positional information of a particle was determined by searching the center of the overlapped reconstructed volume. After confirming the proposed technique using static spherical particles, the 3D information of moving particles suspended in a Hagen-Poiseiulle flow was successfully obtained. Moreover, the 3D information of nonspherical particles, including ellipsoidal particles and red blood cells, were measured using the proposed technique. In addition to 3D positional information, the orientation and shape of the test samples were obtained from the plane images by slicing the overlapped volume perpendicular to the directions of the image recordings. This DsHM technique will be useful in analyzing the 3D dynamic behavior of various nonspherical particles, which cannot be measured by conventional digital holographic microscopy. (C) 2017 Elsevier Ltd. All rights reserved.11Nsciescopu
Machine learning-based in-line holographic sensing of unstained malaria-infected red blood cells
No full textAccurate and immediate diagnosis of malaria is important for medication of the infectious disease. Conventional methods for diagnosing malaria are time consuming and rely on the skill of experts. Therefore, an automatic and simple diagnostic modality is essential for healthcare in developing countries that lack the expertise of trained microscopists. In the present study, a new automatic sensing method using digital in-line holographic microscopy (DIHM) combined with machine learning algorithms was proposed to sensitively detect unstained malaria-infected red blood cells (iRBCs). To identify the RBC characteristics, 13 descriptors were extracted from segmented holograms of individual RBCs. Among the 13 descriptors, 10 features were highly statistically different between healthy RBCs (hRBCs) and iRBCs. Six machine learning algorithms were applied to effectively combine the dominant features and to greatly improve the diagnostic capacity of the present method. Among the classification models trained by the 6 tested algorithms, the model trained by the support vector machine (SVM) showed the best accuracy in separating hRBCs and iRBCs for training (n = 280, 96.78%) and testing sets (n = 120, 97.50%). This DIHM-based artificial intelligence methodology is simple and does not require blood staining. Thus, it will be beneficial and valuable in the diagnosis of malaria.11Nsciescopu
- …
