The improved faster-RCNN for spinal fracture lesions detection

Purpose: Reading spinal CT (Computed Tomography) images is very important in the diagnosis of spondylosis, which is time-consuming and prones to make biases. In this paper, we propose a framework based on Faster-RCNN to improve detection performances of three spinal fracture lesions: cfracture (cervical fracture), tfracture (thoracic fracture) and lfracture (lumbar fracture). Methods: First, we use ResNet50 to replace VGG16 in backbone network in Faster-RCNN to increase depth of training network. Second, we utilize soft-NMS (Non-Maximum Suppression) instead of NMS to avoid missed detection of overlapped lesions. Third, we simplify RPN (Region Proposal Network) to accelerate training speed and reduce missed detection. Finally, we modify the classifier layer in Faster-RCNN and choose appropriate length-width ratio by changing anchor sizes in sliding window, then adopt multi-scale strategy in training to improve efficiency and accuracy. Results: The experimental results show that the proposed scheme has a good performance, mAP (mean average precision) is 90.6%, IOU (Intersection of Union) is 88.5 and detection time is 0.053 second per CT image, which means our proposed method can accurately detect spinal fracture lesions. Conclusion: Our proposed method can provide assistance and scientific references for both doctors and patients in clinically.

Download Full-text

The Shrank YoloV3-tiny for spinal fracture lesions detection

Journal of Intelligent & Fuzzy Systems ◽

10.3233/jifs-212255 ◽

2021 ◽

pp. 1-20

Author(s):

Gang Sha ◽

Junsheng Wu ◽

Bin Yu

Keyword(s):

Mobile Devices ◽

High Performance ◽

Spinal Fracture ◽

Computed Tomography Images ◽

Batch Normalization ◽

Lumbar Fracture ◽

Model Size ◽

Computational Resources ◽

Computer Resources ◽

Novel Model

Purpose: at present, more and more deep learning algorithms are used to detect and segment lesions from spinal CT (Computed Tomography) images. But these algorithms usually require computers with high performance and occupy large resources, so they are not suitable for the clinical embedded and mobile devices, which only have limited computational resources and also expect a relative good performance in detecting and segmenting lesions. Methods: in this paper, we present a model based on Yolov3-tiny to detect three spinal fracture lesions, cfracture (cervical fracture), tfracture (thoracic fracture), and lfracture (lumbar fracture) with a small size model. We construct this novel model by replacing the traditional convolutional layers in YoloV3-tiny with fire modules from SqueezeNet, so as to reduce the parameters and model size, meanwhile get accurate lesions detection. Then we remove the batch normalization layers in the fire modules after the comparative experiments, though the overall performance of fire module without batch normalization layers is slightly improved, we can reduce computation complexity and low occupations of computer resources for fast lesions detection. Results: the experiments show that the shrank model only has a size of 13 MB (almost a third of Yolov3-tiny), while the mAP (mean Average Precsion) is 91.3%, and IOU (intersection over union) is 90.7. The detection time is 0.015 second per CT image, and BFLOP/s (Billion Floating Point Operations per Second) value is less than Yolov3-tiny. Conclusion: the model we presented can be deployed in clinical embedded and mobile devices, meanwhile has a relative accurate and rapid real-time lesions detection.

Download Full-text

PCAN—Part-Based Context Attention Network for Thermal Power Plant Detection in Remote Sensing Imagery

Remote Sensing ◽

10.3390/rs13071243 ◽

2021 ◽

Vol 13 (7) ◽

pp. 1243

Author(s):

Wenxin Yin ◽

Wenhui Diao ◽

Peijin Wang ◽

Xin Gao ◽

Ya Li ◽

...

Keyword(s):

Remote Sensing ◽

Power Plants ◽

State Of The Art ◽

Thermal Power ◽

Image Interpretation ◽

Remote Sensing Image ◽

Thermal Power Plants ◽

Average Precision ◽

Deep Convolutional Neural Networks ◽

Multi Scale

The detection of Thermal Power Plants (TPPs) is a meaningful task for remote sensing image interpretation. It is a challenging task, because as facility objects TPPs are composed of various distinctive and irregular components. In this paper, we propose a novel end-to-end detection framework for TPPs based on deep convolutional neural networks. Specifically, based on the RetinaNet one-stage detector, a context attention multi-scale feature extraction network is proposed to fuse global spatial attention to strengthen the ability in representing irregular objects. In addition, we design a part-based attention module to adapt to TPPs containing distinctive components. Experiments show that the proposed method outperforms the state-of-the-art methods and can achieve 68.15% mean average precision.

Download Full-text

Diversity of Pod Shape in Pisum

Diversity ◽

10.3390/d13050203 ◽

2021 ◽

Vol 13 (5) ◽

pp. 203

Author(s):

Thomas Henry Noel Ellis ◽

Julie M. I. Hofer ◽

Eleni Vikeli ◽

Michael J. Ambrose ◽

Paola Higuera-Poveda ◽

...

Keyword(s):

Allelic Variation ◽

Size Variation ◽

Width Ratio ◽

Negative Regulators ◽

Wide Range ◽

Yield Determination ◽

Organ Specific ◽

Length Width ◽

End Use ◽

Pod Length

The seed-containing pod is the defining structure of plants in the legume family, yet pods exhibit a wide range of morphological variation. Within a species pod characters are likely to be correlated with reproductive strategy, and within cultivated forms will correspond to aspects of yield determination and/or end use. Here variation in pod size, described as pod length: pod width ratio, has been analyzed in pea germplasm represented by 597 accessions. This pod size variation is discussed with respect to population structure and to known classical pod morphology mutants. Variability of the pod length: width ratio can be explained by allelic variation at two genetic loci that may correspond to organ-specific negative regulators of growth.

Download Full-text

A Novel GIS-Based Approach for Automated Detection of Nearshore Sandbar Morphological Characteristics in Optical Satellite Imagery

Remote Sensing ◽

10.3390/rs13112233 ◽

2021 ◽

Vol 13 (11) ◽

pp. 2233

Author(s):

Rasa Janušaitė ◽

Laurynas Jukna ◽

Darius Jarmalavičius ◽

Donatas Pupienis ◽

Gintautas Žilinskas

Keyword(s):

Satellite Imagery ◽

Morphological Characteristics ◽

Unsupervised Classification ◽

Landsat 8 ◽

Multi Scale ◽

Medium Resolution ◽

Wide Range ◽

Position Index ◽

Local Relief ◽

Length Width

Satellite remote sensing is a valuable tool for coastal management, enabling the possibility to repeatedly observe nearshore sandbars. However, a lack of methodological approaches for sandbar detection prevents the wider use of satellite data in sandbar studies. In this paper, a novel fully automated approach to extract nearshore sandbars in high–medium-resolution satellite imagery using a GIS-based algorithm is proposed. The method is composed of a multi-step workflow providing a wide range of data with morphological nearshore characteristics, which include nearshore local relief, extracted sandbars, their crests and shoreline. The proposed processing chain involves a combination of spectral indices, ISODATA unsupervised classification, multi-scale Relative Bathymetric Position Index (RBPI), criteria-based selection operations, spatial statistics and filtering. The algorithm has been tested with 145 dates of PlanetScope and RapidEye imagery using a case study of the complex multiple sandbar system on the Curonian Spit coast, Baltic Sea. The comparison of results against 4 years of in situ bathymetric surveys shows a strong agreement between measured and derived sandbar crest positions (R2 = 0.999 and 0.997) with an average RMSE of 5.8 and 7 m for PlanetScope and RapidEye sensors, respectively. The accuracy of the proposed approach implies its feasibility to study inter-annual and seasonal sandbar behaviour and short-term changes related to high-impact events. Algorithm-provided outputs enable the possibility to evaluate a range of sandbar characteristics such as distance from shoreline, length, width, count or shape at a relevant spatiotemporal scale. The design of the method determines its compatibility with most sandbar morphologies and suitability to other sandy nearshores. Tests of the described technique with Sentinel-2 MSI and Landsat-8 OLI data show that it can be applied to publicly available medium resolution satellite imagery of other sensors.

Download Full-text

Meshing Method of High Precision FEM in Large Complex Structural Simulations

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.195-196.701 ◽

2012 ◽

Vol 195-196 ◽

pp. 701-704

Author(s):

Yan Hua Xue ◽

Zhi Guang Wang ◽

Xiao Hong Li ◽

Xin Jiang

Keyword(s):

Finite Element ◽

Complex Structure ◽

Width Ratio ◽

Two Dimensional ◽

Element Analysis ◽

Large Complex ◽

Mesh Density ◽

Length Width ◽

Complex Structural ◽

Dimensional Element

Shing is playing an important role in the large complex structural FEM simulations; it has a direct effect on calculating precision of structural simulations. For increasing the calculation accuracy and analysis accuracy of complex structure, the finite element meshing problems is proposed on the finite element analysis of large complicated structures. The effects caused by element type, mesh density and intergradations on calculating precision are studied and discussed. A research argues that with length-width ratio of 1~2 and length-thickness ration of 1.5~4.5 of two-dimensional rectangular element, the quality of meshing method of two-dimensional element is above normal. As the height of one-dimensional element is equal to the sum of reinforcing rib height of outer panel and half the thickness of panel, more accurate results can be obtained.

Download Full-text

Morphogenesis of the silkworm egg

Development ◽

10.1242/dev.36.1.13 ◽

1976 ◽

Vol 36 (1) ◽

pp. 13-18

Author(s):

J. M. Legay

Keyword(s):

Ovarian Follicle ◽

Temporal Organization ◽

Width Ratio ◽

Follicle Development ◽

Ovarian Follicle Development ◽

Length Width ◽

Form Stability

Ovarian follicle development, which accompanies morphogenesis of the silkworm egg has three distinct phases: spheric, ellipsoidal and flattened-ellipsoid. Transitions between phases are rapid and form-stability (characterized by length/width ratio) is preserved from the beginning of the ellipsoidal phase. The geometric stability of the follicle-oocyte-ovariole system, the polarity of the egg and the determinism in form changes reveal strikingly coordinated spatial and temporal organization.

Download Full-text

Comportamento tático coletivo em Pequenos Jogos no Futebol: influência de jogadores adicionais

Brazilian Journal of Kinanthropometry and Human Performance ◽

10.5007/1980-0037.2016v18n1p62 ◽

2016 ◽

Vol 18 (1) ◽

pp. 62 ◽

Cited By ~ 5

Author(s):

Gibson Moreira Praça ◽

Hugo Folgado ◽

André Gustavo Pereira de Andrade ◽

Pablo Juan Greco

Keyword(s):

The Other ◽

Soccer Players ◽

Width Ratio ◽

Collective Variables ◽

Other Hand ◽

Centroid Distance ◽

Friedman's Test ◽

Polar Coordinate ◽

Length Width ◽

Positional Data

DOI: http://dx.doi.org/10.5007/1980-0037.2016v18n1p62 The aim of this study was to compare the collective tactical behavior between numerically balanced and unbalanced small-sided soccer games. Eighteen male soccer players (mean age 16.4 years) participated in the study. Polar coordinate analysis was performed using positional data obtained with a 15-Hz GPS device. Collective variables including length, width, centroid distance (average point between teammates), and length per width ratio (LPWratio) were collected. Data were analyzed using Friedman’s test. The results showed greater length and width values in 4vs.3 games, while a higher LPWratiowas observed in 3vs.3+2 games compared to the other configurations. In games with an additional player (4vs.3), ball circulation and the increase in effective game space were alternatives to overcome the more concentrated defensive systems near the goal. On the other hand, 3vs.3+2 games allowed more actions in the length axis and a fast reach of the opponent’s goal.

Download Full-text

Estimating the Effect of Fractal Dimension on Representative Elementary Volume of Randomly Distributed Rock Fracture Networks

Geofluids ◽

10.1155/2018/7206074 ◽

2018 ◽

Vol 2018 ◽

pp. 1-13 ◽

Cited By ~ 1

Author(s):

Jing Zhang ◽

Liyuan Yu ◽

Hongwen Jing ◽

Richeng Liu

Keyword(s):

Fractal Dimension ◽

Functional Relationship ◽

Rock Fracture ◽

Representative Elementary Volume ◽

Width Ratio ◽

Elementary Volume ◽

Equivalent Permeability ◽

Length Width ◽

Rock Fracture Networks

The effect of fractal dimension (Df) on the determination of representative elementary volume (REV) was investigated through numerical experimentations, in which a new method was adopted to extract submodels that have different length-width ratios from original discrete facture networks (DFNs). Fluid flow in 1610 DFNs with different geometric characteristics of fractures and length-width ratios was simulated, and the equivalent permeability was calculated. The results show that the average equivalent permeability (KREV) at the REV size for DFNs increases with the increase in Df. The KREV shows a downward trend with increasing length-width ratio of the submodel. A strong exponent functional relationship is found between the REV size and Df. The REV size decreases with increasing Df. With the increment of the length-width ratio of submodels, the REV size shows a decreasing trend. The effects of length-width ratio and Df on the REV size can be negligible when Df≥1.5, but are significant when Df<1.5.

Download Full-text

Correlation of Ultrasound Findings and Fine-Needle Aspiration Cytology for the Diagnosis of Axillary Lymph Node Metastasis in Patients with Breast Carcinoma

Acta Cytologica ◽

10.1159/000493635 ◽

2018 ◽

Vol 63 (1) ◽

pp. 17-22

Author(s):

Kusum Kapila ◽

Preetha Alath ◽

Govind H. Hebbar ◽

Mohammed Jaragh ◽

Sara S. George ◽

...

Keyword(s):

Lymph Node ◽

Breast Carcinoma ◽

Fine Needle Aspiration ◽

Needle Aspiration ◽

Width Ratio ◽

Axillary Lymph ◽

Aspiration Cytology ◽

Fine Needle ◽

Length Width ◽

Needle Aspiration Cytology

Objective: Ultrasound-guided fine-needle aspiration cytology (USG-FNAC) is used for the detection of axillary lymph node (ALN) metastasis in patients with breast carcinoma (BC). US findings have a good diagnostic accuracy with high sensitivity and specificity. The aim of this study is to correlate the detection of ALN metastases on US with FNAC in BC patients. Study Design: In 75 BC patients, over a period of 9 months (January to September 2017), the size, cortical thickness (CT), presence or absence of hilar fat, and length/width ratio of ALN on US were reviewed and correlated with FNAC findings. Results: The age range was 29–78 (mean 52) years. There were 38 patients with a single ALN and 37 with multiple ALNs. ALNs with a maximum length of > 2.5 cm were malignant in 100% of cases while those ≥1.5 cm were malignant in 80.4%. ALNs with a CT of > 3 mm had metastasis in 78.1% cases. ALNs with absent hilar fat showed tumour in 87.5% cases. A length/width ratio of < 2 showed a metastatic tumour in 66.7% of aspirates. Conclusion: An association was seen between metastatic carcinoma on FNAC and axillary US features of a maximum length of ≥1.5 cm, the absence of hilar fat, and a CT of > 3 mm (p < 0.05).

Download Full-text

Do we have infraspecific taxa of Salvia multicaulis Vahl. (Lamiaceae) in Iran?

Ukrainian Journal of Ecology ◽

10.15421/2017_138 ◽

2017 ◽

Vol 7 (4) ◽

pp. 432-439 ◽

Cited By ~ 1

Author(s):

Seyed Mehdi Talebi ◽

Reza Rezakhanlou ◽

A V. Matsyura

Keyword(s):

Leaf Shape ◽

Morphological Characteristics ◽

Leaf Length ◽

Width Ratio ◽

Petiole Length ◽

Widespread Species ◽

Basal Leaf ◽

Leaf Petiole ◽

Upgma Cluster Analysis ◽

Length Width

Salvia multicaulis is a widespread species of Lamiaceae family in Iran. There are many discussions about its infraspecific variations. Although some varieties were definite for this species in various parts of the world, no infraspecific taxon was reported in Iran and all samples of this species were named as S. multicaulis. In this study, morphological characteristics of S. multicaulis populations, naturally growing in Iran, was examined. Twenty-two traits were examined in 94 individuals of this species to identify their phenotypic difference. Most of the investigated features were showing a high degree of variability, but it was highly pronounced for some characteristics such as basal leaf shape, basal leaf width, basal leaf length/ width ratio and basal petiole length. Significant positive/negative correlations were observed between some morphological variables. Furthermore, significant negative correlations were found between the eastern distribution of populations with basal leaf petiole length and basal leaf length/ width ratio. Based on the UPGMA cluster analysis, populations were divided into two main branches. The first branch contained four populations, while the second branch was bigger and clustered in two sub-branches. In one of them, three populations and in another one the rest populations arranged in two groups. CA joined plot confirmed that each of studied populations or group of populations had distinct morphological trait(s), which were useful in identification of them. Our findings supported population no. 13 had unique morphological traits such as the largest bracts and basal leaf petiole, highest flower number of each inflorescence cycle, widest and largest calyx. The conservation of the highly diverse populations of Iranian S. multicaulis is recommended.

Download Full-text