The electron detection performance of the “Icarus” hCMOS imaging sensor

Cells interact with their extracellular environments by means of a variety of cellular adhesion molecules (CAM) and surface ligands. In many instances, CAMs interact in a sequential temporal fashion which suggests that these adhesion molecules may occupy or be polarized to various membrane microdomains on the cell surface. Detection of CAMs can be accomplished by a variety of methods including immunofluorescent microscopy and flow cytometry, and by the use of immunocytochemical markers (i.e. colloidal gold) in electron microscopy. The development of high resolution field emission SEM in the mid 1980's and the Autrata modification of the YAG detector for backscatter electron detection at low voltage has greatly facilitated the recognition of colloidal gold probes for detection of surface CAMs. Low voltage FESEM with Bse imaging provides increased resolution of cell surface topography (~3nm at 3-4 keV) which can be observed in 3-dimensions, and simultaneously permits detection/high spatial resolution of immunogold label by atomic number contrast.

Download Full-text

Yag Single Crystals for Scintillators of Stem

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100179580 ◽

1990 ◽

Vol 48 (1) ◽

pp. 166-167

Author(s):

M. Hibino ◽

K. Irie ◽

R. Autrata ◽

P. schauer

Keyword(s):

Single Crystals ◽

Yttrium Aluminum Garnet ◽

Light Guide ◽

Detection System ◽

Polished Surface ◽

Detective Quantum Efficiency ◽

Backscattered Electrons ◽

Electron Detection ◽

Yttrium Aluminum ◽

Phosphor Screens

Although powdered phosphor screens are usually used for scintillators of STEM, it has been found that the phosphor screen of appropriate thickness should be used depending on the accelerating voltage, in order to keep high detective quantum efficiency. 1 It has been also found that the variation in sensitivity, due to granularity of phosphor screens, makes the measurement of fine electron probe difficult and that the sensitivity reduces with electron irradiation specially at high voltages.In order to find out a preferable scintillator for STEM, single crystals of YAG (yttrium aluminum garnet), which are used for detecting secondary and backscattered electrons in SEM were investigated and compared with powdered phosphor screens, at the accelerating voltages of 100kV and 1 MV. A conventional electron detection system, consisting of scintillator, light guide and PMT (Hamamatsu Photonics R268) was used for measurements. Scintillators used are YAG single crystals of 1.0 to 3.2mm thicknesses (with surfaces matted for good interface to the light guide) and of 0.8mm thickness (with polished surface), and powdered P-46 phosphor screens of 0.07mm and 1.0mm thicknesses for 100kV and 1MV, respectively. Surfaces on electron-incidence side of all scintillators are coated with reflecting layers.

Download Full-text

Understanding Target Detection Performance When Receiving Visual or Auditory Cues

PsycEXTRA Dataset ◽

10.1037/e618602012-013 ◽

2002 ◽

Author(s):

David C. Cibik ◽

Erich W. Meyerhoff

Keyword(s):

Target Detection ◽

Detection Performance ◽

Auditory Cues

Download Full-text

CNN performance dependence on linear image processing

Electronic Imaging ◽

10.2352/issn.2470-1173.2020.10.ipas-182 ◽

2020 ◽

Vol 2020 (10) ◽

pp. 310-1-310-7

Author(s):

Khalid Omer ◽

Luca Caucci ◽

Meredith Kupinski

Keyword(s):

Image Processing ◽

Texture Classification ◽

Full Rank ◽

Detection Performance ◽

Ideal Observer ◽

Training Data ◽

Image Texture ◽

Training Images ◽

Analytic Expressions ◽

Linear Compression

This work reports on convolutional neural network (CNN) performance on an image texture classification task as a function of linear image processing and number of training images. Detection performance of single and multi-layer CNNs (sCNN/mCNN) are compared to optimal observers. Performance is quantified by the area under the receiver operating characteristic (ROC) curve, also known as the AUC. For perfect detection AUC = 1.0 and AUC = 0.5 for guessing. The Ideal Observer (IO) maximizes AUC but is prohibitive in practice because it depends on high-dimensional image likelihoods. The IO performance is invariant to any fullrank, invertible linear image processing. This work demonstrates the existence of full-rank, invertible linear transforms that can degrade both sCNN and mCNN even in the limit of large quantities of training data. A subsequent invertible linear transform changes the images’ correlation structure again and can improve this AUC. Stationary textures sampled from zero mean and unequal covariance Gaussian distributions allow closed-form analytic expressions for the IO and optimal linear compression. Linear compression is a mitigation technique for high-dimension low sample size (HDLSS) applications. By definition, compression strictly decreases or maintains IO detection performance. For small quantities of training data, linear image compression prior to the sCNN architecture can increase AUC from 0.56 to 0.93. Results indicate an optimal compression ratio for CNN based on task difficulty, compression method, and number of training images.

Download Full-text