Recent studies have found that the position of mice
or rats can be decoded from calcium imaging of brain activity
offline. However, given the complex analysis pipeline, real-time
position decoding remains a challenging task, especially considering
strict requirements on hardware usage and energy cost
for closed-loop feedback applications. In this paper, we propose
two neural network based methods and corresponding hardware
designs for real-time position decoding from calcium images. Our
methods are based on: 1) convolutional neural network (CNN),
2) spiking neural network (SNN) converted from the CNN.
We implemented quantized CNN and SNN models on FPGA.
Evaluation results show that the CNN and the SNN methods
achieve 56.3%/83.1% and 56.0%/82.8% Hit-1/Hit-3 accuracy for
the position decoding across different rats, respectively. We also
observed an accuracy-latency tradeoff of the SNN method in
decoding positions under various time steps. Finally, we present
our SNN implementation on the neuromorphic chip Loihi.
Index Terms—calcium image, decoding, neural network.
more »
« less
Prediction of Inclusion Types From BSE Images: RF vs. CNN
The analysis of non-metallic inclusions is crucial for the assessment of steel properties. Scanning electron microscopy (SEM) coupled with energy dispersive spectroscopy (EDS) is one of the most prominent methods for inclusion analysis. This study utilizes the output generated from SEM/EDS analysis to predict inclusion types from BSE images. Prediction models were generated using two different algorithms, Random Forest (RF) and convolutional neural networks (CNN), for comparison. For each method, three separate models were developed. Starting with a simple binary model to differentiate between inclusions and non-inclusions, then developing to more complex four and five class models. For the 4-class model, inclusions were split into oxides, sulfides, and oxy-sulfides, in addition to the non-inclusion class. The 5-class model included specific types of inclusions only, namely alumina, calcium aluminates, calcium sulfides, complex calcium-manganese sulfides, and oxy-sulfide inclusions. CNN achieved better accuracy for the binary (92%) and 4-class (78%) models, compared to RF (binary 87%, 4-class 75%). For the 5-class model, the results were similar, 60% accuracy for RF and 59% for CNN.
more »
« less
- Award ID(s):
- 1826218
- PAR ID:
- 10333854
- Date Published:
- Journal Name:
- Frontiers in Materials
- Volume:
- 8
- ISSN:
- 2296-8016
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Cellular strategies and regulation of their crystallization mechanisms are essential to the formation of biominerals, and harnessing these strategies will be important for the future creation of novel non-native biominerals that recapitulate the impressive properties biominerals possess. Harnessing these biosynthetic strategies requires an understanding of the interplay between insoluble organic matrices, mineral precursors, and soluble organic and inorganic additives. Our long-range goal is to use a sea anemone model system (Nematostella vectensis) to examine the role of intrinsically disordered proteins (IDPs) found in native biomineral systems. Here, we study how ambient temperatures (25–37 °C) and seawater solution compositions (varying NaCl and Mg ratios) will affect the infiltration of organic matrices with calcium carbonate mineral precursors generated through a polymer-induced liquid-precursor (PILP) process. Fibrillar collagen matrices were used to assess whether solution conditions were suitable for intrafibrillar mineralization, and SEM with EDS was used to analyze mineral infiltration. Conditions of temperatures 30 °C and above and with low Mg:Ca ratios were determined to be suitable conditions for calcium carbonate infiltration. The information obtained from these observations may be useful for the manipulation and study of cellular secreted IDPs in our quest to create novel biosynthetic materials.more » « less
-
more » « less
Human mesenchymal stem cells (hMSCs) are multipotent progenitor cells with the potential to differentiate into various cell types, including osteoblasts, chondrocytes, and adipocytes. These cells have been extensively employed in the field of cell-based therapies and regenerative medicine due to their inherent attributes of self-renewal and multipotency. Traditional approaches for assessing hMSCs differentiation capacity have relied heavily on labor-intensive techniques, such as RT-PCR, immunostaining, and Western blot, to identify specific biomarkers. However, these methods are not only time-consuming and economically demanding, but also require the fixation of cells, resulting in the loss of temporal data. Consequently, there is an emerging need for a more efficient and precise approach to predict hMSCs differentiation in live cells, particularly for osteogenic and adipogenic differentiation. In response to this need, we developed innovative approaches that combine live-cell imaging with cutting-edge deep learning techniques, specifically employing a convolutional neural network (CNN) to meticulously classify osteogenic and adipogenic differentiation. Specifically, four notable pre-trained CNN models, VGG 19, Inception V3, ResNet 18, and ResNet 50, were developed and tested for identifying adipogenic and osteogenic differentiated cells based on cell morphology changes. We rigorously evaluated the performance of these four models concerning binary and multi-class classification of differentiated cells at various time intervals, focusing on pivotal metrics such as accuracy, the area under the receiver operating characteristic curve (AUC), sensitivity, precision, and F1-score. Among these four different models, ResNet 50 has proven to be the most effective choice with the highest accuracy (0.9572 for binary, 0.9474 for multi-class) and AUC (0.9958 for binary, 0.9836 for multi-class) in both multi-class and binary classification tasks. Although VGG 19 matched the accuracy of ResNet 50 in both tasks, ResNet 50 consistently outperformed it in terms of AUC, underscoring its superior effectiveness in identifying differentiated cells. Overall, our study demonstrated the capability to use a CNN approach to predict stem cell fate based on morphology changes, which will potentially provide insights for the application of cell-based therapy and advance our understanding of regenerative medicine.
-
The objective of this research was to create and appraise biodegradable polymer-based nanofibers containing distinct concentrations of calcium trimetaphosphate (Ca-TMP) for periodontal tissue engineering. Poly(ester urea) (PEU) (5% w/v) solutions containing Ca-TMP (15%, 30%, 45% w/w) were electrospun into fibrous scaffolds. The fibers were evaluated using SEM, EDS, TGA, FTIR, XRD, and mechanical tests. Degradation rate, swelling ratio, and calcium release were also evaluated. Cell/Ca-TMP and cell/scaffold interaction were assessed using stem cells from human exfoliated deciduous teeth (SHEDs) for cell viability, adhesion, and alkaline phosphatase (ALP) activity. Analysis of variance (ANOVA) and post-hoc tests were used (α = 0.05). The PEU and PEU/Ca-TMP-based membranes presented fiber diameters at 469 nm and 414–672 nm, respectively. Chemical characterization attested to the Ca-TMP incorporation into the fibers. Adding Ca-TMP led to higher degradation stability and lower dimensional variation than the pure PEU fibers; however, similar mechanical characteristics were observed. Minimal calcium was released after 21 days of incubation in a lipase-enriched solution. Ca-TMP extracts enhanced cell viability and ALP activity, although no differences were found between the scaffold groups. Overall, Ca-TMP was effectively incorporated into the PEU fibers without compromising the morphological properties but did not promote significant cell function.more » « less
-
This paper aims to clarify the influence of different types of fly ash on the mechanical properties and self-healing behavior of Engineered Cementitious Composite (ECC). Five types of fly ash with different chemical and physical properties were used in ECC mixtures. The fly ash to cement ratio was fixed at 3.0. The compressive and uniaxial tensile tests were conducted to evaluate the influence of fly ash type on mechanical properties. The permeability test was used to assess self-healing behavior of ECCs with different types of fly ash. The microtopography and chemical characteristics of the self-healing products in the crack were observed and examined by scanning electron microscope (SEM) and energy dispersive X-ray spectroscopy (EDS). The fly ash with relatively higher calcium content and smaller particle size was found conducive to a higher compressive strength. The lower combined Al2O3 and CaO content of this fly ash, however, was found to enhance the tensile strain capacity. Furthermore, high calcium fly ash accelerates the self-healing process of ECC for the same pre-damaged level. The self-healing product was a mixed CaCO3/C-S-H system with the CaCO3 as the main ingredient.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.3389/fmats.2021.754089
