Python Source Code Research Articles

Bayesian Performance Analysis for Algorithm Ranking Comparison

In the field of optimization and machine learning, the statistical assessment of results has played a key role in conducting algorithmic performance comparisons. Classically, null hypothesis statistical tests have been used. However, recently, alternatives based on Bayesian statistics have shown great potential in complex scenarios, especially when quantifying the uncertainty in the comparison. In this work, we delve deep into the Bayesian statistical assessment of experimental results by proposing a framework for the analysis of several algorithms on several problems/instances. To this end, experimental results are transformed to their corresponding rankings of algorithms, assuming that these rankings have been generated by a probability distribution (defined on permutation spaces). From the set of rankings, we estimate the posterior distribution of the parameters of the studied probability models, and several inferences concerning the analysis of the results are examined. Particularly, we study questions related to the probability of having one algorithm in the first position of the ranking or the probability that two algorithms are in the same relative position in the ranking. Not limited to that, the assumptions, strengths, and weaknesses of the models in each case are studied. To help other researchers to make use of this kind of analysis, we provide a Python package and source code implementation at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://zenodo.org/record/6320599</uri> .

Image analysis and error detection in source software code

Subject of study. We study cognitive mechanisms for the analysis of images of source software code and detection of errors in it by a human. Aim. The aim of this study is to determine the influence of the professional skill of visual error detection in a Python source code on the control over oculomotor processes to be used for the subsequent modeling of the identified principles in artificial intelligence systems. Method. The eye movement properties of a human during error detection in the images of software code were investigated using eye movement tracking technology by means of the Russian software–hardware suite “Neurobureau,” which enabled a complete set of psychophysiological studies to be conducted. The test subjects completed two tasks: explaining the software code and finding an error in it. Each task comprised 10 stimuli with software code in Python language normalized to length and complexity and with syntax highlighting. The task completion time was not limited. Eight programmers with different professional experience (from 1 to 13 years) participated in the study. Main results. With increasing professional skill in image analysis, humans develop eye movement strategies enabling more effective performance of tasks with minimum effort. These strategies entail division of an integral software code into individual units relevant for the analysis. We discovered that experienced programmers exhibited fewer fixations, shorter scanpath length, and larger saccade amplitudes than novice programmers. Notably, the speed of saccades, in particular the broad searching eye movements during the task of software code explanation, increases with professional experience. We demonstrated that visual error detection is mainly determined by recognition of the details in a text that can be found exclusively based on the semantics and grammar of the programming language. We established that reading of software code texts is different from both viewing visual scenes and reading texts written in a natural language. Professional experience minimizes the efforts in this work, similar to any other type of professional activity. Eye movement control during code reading is guided by the knowledge of the programming language, understanding of context, and knowledge of the semantics and grammar of the language, as well as the low-spatial-frequency description of the images of strings and words, which is used to develop the skill of reading and recognition of the general configuration of strings and words in a natural language. Practical significance. The conclusions obtained in this study can be used to describe the existing techniques and create new neuromorphic algorithms (based on the strategies developed during human evolution) for the generation and correction of software code.

FindPFΔS: Non-Target Screening for PFAS─Comprehensive Data Mining for MS2 Fragment Mass Differences.

The limited availability of analytical reference standards makes non-target screening approaches based on high-resolution mass spectrometry increasingly important for the efficient identification of unknown PFAS (per- and polyfluoroalkyl substances) and their TPs. We developed and optimized a vendor-independent open-source Python-based algorithm (FindPFΔS = FindPolyFluoroDeltas) to search for distinct fragment mass differences in MS/MS raw data (.ms2-files). Optimization with PFAS standards, two pre-characterized paper and soil samples (iterative data-dependent acquisition), revealed Δ(CF2)n, ΔHF, ΔCnH3F2n-3, ΔCnH2F2n-4, ΔCnHF2n-5, ΔCnF2nSO3, ΔCF3, and ΔCF2O as relevant and selective fragment differences depending on applied collision energies. In a PFAS standard mix, 94% (36 of 38 compounds from 10 compound classes) could be found by FindPFΔS. The use of fragment differences was applicable to a wide range of PFAS classes and appears as a promising new approach for PFAS identification. The influence of mass tolerance and intensity threshold on the identification efficiency and on the detection of false positives was systematically evaluated with the use of selected HR-MS2-spectra (20,998) from MassBank. To this end, with the use of FindPFΔS, we could identify different unknown PFAS homologues in the paper extracts. FindPFΔS is freely available as both Python source code on GitHub (https://github.com/JonZwe/FindPFAS) and as an executable windows application (https://doi.org/10.5281/zenodo.6797353) with a graphical user interface on Zenodo.

Generation of Ince-Gaussian Beams Using Azocarbazole Polymer CGH.

Ince–Gaussian beams, defined as a solution to a wave equation in elliptical coordinates, have shown great advantages in applications such as optical communication, optical trapping and optical computation. However, to ingress these applications, a compact and scalable method for generating these beams is required. Here, we present a simple method that satisfies the above requirement, and is capable of generating arbitrary Ince–Gaussian beams and their superposed states through a computer-generated hologram of size 1 mm2, fabricated on an azocarbazole polymer film. Other structural beams that can be derived from the Ince–Gaussian beam were also successfully generated by changing the elliptical parameters of the Ince–Gaussian beam. The orthogonality relations between different Ince–Gaussian modes were investigated in order to verify applicability in an optical communication regime. The complete python source code for computing the Ince–Gaussian beams and their holograms are also provided.

Automatic Prostate Gleason Grading Using Pyramid Semantic Parsing Network in Digital Histopathology.

PurposeProstate biopsy histopathology and immunohistochemistry are important in the differential diagnosis of the disease and can be used to assess the degree of prostate cancer differentiation. Today, prostate biopsy is increasing the demand for experienced uropathologists, which puts a lot of pressure on pathologists. In addition, the grades of different observations had an indicating effect on the treatment of the patients with cancer, but the grades were highly changeable, and excessive treatment and insufficient treatment often occurred. To alleviate these problems, an artificial intelligence system with clinically acceptable prostate cancer detection and Gleason grade accuracy was developed.MethodsDeep learning algorithms have been proved to outperform other algorithms in the analysis of large data and show great potential with respect to the analysis of pathological sections. Inspired by the classical semantic segmentation network, we propose a pyramid semantic parsing network (PSPNet) for automatic prostate Gleason grading. To boost the segmentation performance, we get an auxiliary prediction output, which is mainly the optimization of auxiliary objective function in the process of network training. The network not only includes effective global prior representations but also achieves good results in tissue micro-array (TMA) image segmentation.ResultsOur method is validated using 321 biopsies from the Vancouver Prostate Centre and ranks the first on the MICCAI 2019 prostate segmentation and classification benchmark and the Vancouver Prostate Centre data. To prove the reliability of the proposed method, we also conduct an experiment to test the consistency with the diagnosis of pathologists. It demonstrates that the well-designed method in our study can achieve good results. The experiment also focused on the distinction between high-risk cancer (Gleason pattern 4, 5) and low-risk cancer (Gleason pattern 3). Our proposed method also achieves the best performance with respect to various evaluation metrics for distinguishing benign from malignant.AvailabilityThe Python source code of the proposed method is publicly available at https://github.com/hubutui/Gleason. All implementation details are presented in this paper.ConclusionThese works prove that the Gleason grading results obtained from our method are effective and accurate.

PyFocus – A Python package for vectorial calculations of focused optical fields under realistic conditions. Application to toroidal foci

Focused optical fields are key to a multitude of applications involving light-matter interactions, such as optical microscopy, single-molecule spectroscopy, optical tweezers, lithography, or quantum coherent control. A detailed vectorial characterization of the focused optical fields that includes a description beyond the paraxial approximation is key to optimize technological performance as well as for the design of meaningful experiments and interpret properly their results. Here, we present PyFocus, an open-source Python software package to perform fully vectorial calculations of focused electromagnetic fields after modulation by an arbitrary phase mask and in the presence of a multilayer system. We provide a graphical user interface and high-level functions to easily integrate PyFocus into custom scripts. Furthermore, to demonstrate the potential of PyFocus, we apply it to extensively characterize the generation of toroidal foci with a high numerical aperture objective, as it is commonly done in super-resolution fluorescence microscopy methods such as STED or MINFLUX. We provide examples of the effects of different experimental factors such as polarization, aberrations, and misalignments of key optical elements. Finally, we present calculations of toroidal foci through an interface of different mediums, and, to our knowledge, the first calculations of toroidal foci generated in total internal reflection conditions. Program summaryProgram Title: PyFocusCPC Library link to program files:https://doi.org/10.17632/sybnhnxbrt.1Developer's repository link:https://github.com/stefani-lab/, https://pypi.org/project/PyCustomFocus/, https://pyfocus.readthedocs.io/ (user manual)Licensing provisions: MITProgramming language: Python 3.6Nature of problem: Full vectorial calculation of focused of optical fields under realistic conditions, such as high numerical aperture, incident beams modulated in phase and/or intensity, and focusing through an interface or a multilayer system.Solution method: PyFocus incorporates all the equations necessary to calculate the focused optical field and provides suitable integration methods. Setting up of the parameters, running calculations, plotting and saving results can be done by means of a graphical user interface or high-level functions.Additional comments including restrictions and unusual features: In addition to the Python source code, an MS Windows executable is provided.

Audio recognition of Chinese traditional instruments based on machine learning

This paper is part of a special issue on Music Technology. We study the type recognition of traditional Chinese musical instrument audio in the common way. Using MEL spectrum characteristics as input, we train an 8-layer convolutional neural network, and finally achieve 99.3% accuracy. After that, this paper mainly studies the performance skill recognition of Chinese traditional musical instruments. Firstly, for a single instrument, the features were extracted by using the pre-trained ResNet model, and then the SVM algorithm was used to classify all the instruments with an accuracy of 99%. Then, in order to improve the generalization of the model, the paper proposes the performance skill recognition of the same kind of instruments. In this way, the regularity of the same playing technique of different instruments can be utilized. Finally, the recognition accuracy of the four kinds of instruments is as follows: 95.7% for blowing instruments, 82.2% for plucked-string instruments, 88.3% for strings instruments, and 97.5% for percussion instruments. We open source the audio database of traditional Chinese musical instruments and the Python source code of the whole experiment for further research.

PyComm: Malicious commands detection model for python scripts

Python is a concise language which can be used to build lightweight tools or dynamic object-orientated applications. The various attributes of Python have made it attractive to numerous malware authors. Attackers often embed malicious shell commands into Python scripts for illegal operations. However, traditional static analysis methods are not feasible to detect this kind of attack because they focus on common features and failure in finding those malicious commands. On the other hand, dynamic analysis is not optimal in this case for its time-consuming and inefficient. In this paper, we propose PyComm, a model for detecting malicious commands in Python scripts with multidimensional features based on machine learning, which considers both 12 statistical features and string sequences of Python source code. Meanwhile, three comparison experiments are designed to evaluate the validity of proposed method. Experimental results show that presented model has achieved an excellent performance based on those practical features and random forest (RF) algorithm, obtained an accuracy of 0.955 with a recall of 0.943.

Implementation of a practical Markov chain Monte Carlo sampling algorithm in PyBioNetFit.

Bayesian inference in biological modeling commonly relies on Markov chain Monte Carlo (MCMC) sampling of a multidimensional and non-Gaussian posterior distribution that is not analytically tractable. Here, we present the implementation of a practical MCMC method in the open-source software package PyBioNetFit (PyBNF), which is designed to support parameterization of mathematical models for biological systems. The new MCMC method, am, incorporates an adaptive move proposal distribution. For warm starts, sampling can be initiated at a specified location in parameter space and with a multivariate Gaussian proposal distribution defined initially by a specified covariance matrix. Multiple chains can be generated in parallel using a computer cluster. We demonstrate that am can be used to successfully solve real-world Bayesian inference problems, including forecasting of new Coronavirus Disease 2019 case detection with Bayesian quantification of forecast uncertainty. PyBNF version 1.1.9, the first stable release with am, is available at PyPI and can be installed using the pip package-management system on platforms that have a working installation of Python 3. PyBNF relies on libRoadRunner and BioNetGen for simulations (e.g. numerical integration of ordinary differential equations defined in SBML or BNGL files) and Dask.Distributed for task scheduling on Linux computer clusters. The Python source code can be freely downloaded/cloned from GitHub and used and modified under terms of the BSD-3 license (https://github.com/lanl/pybnf). Online documentation covering installation/usage is available (https://pybnf.readthedocs.io/en/latest/). A tutorial video is available on YouTube (https://www.youtube.com/watch?v=2aRqpqFOiS4&t=63s). Supplementary data are available at Bioinformatics online.

Technical considerations on the use of Granger causality in neuromonitoring

Neuromonitoring-derived indices play an important role in implementing personalised medicine for traumatic brain injury patients. A well-established example is the pressure reactivity index (PRx), calculated from spontaneous fluctuations of arterial blood pressure (ABP) and intracranial pressure (ICP). PRx assumes causal relationship between ABP and ICP but lacks the check for this assumption. Granger causality (GC) — a method of assessing causal interactions between time series data — is gaining popularity in neurosciences. In our work, we used ABP and ICP data recorded at the frequency of 100 Hz or higher from 235 traumatic brain injury patients. We focused on time domain GC. Analysis was first performed directly on high-resolution data, which included pulse waves. We showed that due to the measurement delay in high-resolution ABP data, GC analysis may erroneously indicate strong ICP→ABP causal relation. Subsequently, the data were downsampled to 0.1 Hz, effectively removing pulse and respiratory waves. We aimed to investigate how different ways of calculating GC influence results and which way should be recommended for ABP-ICP recordings. We considered aspects like selecting autoregressive model order and dealing with data non-stationarity. In addition, we generated simulated signals to investigate the influence of gaps and different procedures of missing data imputation on GC estimation. We showed that unlike methods which interpolate missing data, replacing missing data by white Gaussian noise did not increase the rate of false GC detection. Python source code used in this study is available at: https://github.com/m-m-placek/python-icmplus-granger-causality. Statement of significanceAssessing causality between time series data is of particular interest when neuromonitoring indices are derived from those time series and causal interaction between them is assumed. Causality assessment can improve reliability of such indices and open pathways for their safe clinical implementation. Granger Causality (GC) has recently been investigated in data collected from traumatic brain injury patients. However, there are two main issues related to applications suggested in these studies. Firstly, they considered GC for entire multi-day data recordings or for 24-h long episodes. There is interest in considering causal relationships in finer granularity, also in terms of their potential real-time applications at the bedside. Secondly, GC calculation requires selecting some parameters and there is no unique nor standardised way of doing that. Many papers often provide very brief description of data pre-processing and GC calculation. For this reason, it can be harder to reproduce and compare results derived from GC application. Different ways of obtaining GC may potentially lead to inconsistent results. Here, we attempted to explore possibility of time-varying GC of finer granularity and to provide general guidelines for application of GC to neurocritical care time series affected by periods of missing values.

Projecting social contact matrices to populations stratified by binary attributes with known homophily.

Contact networks are heterogeneous. People with similar characteristics are more likely to interact, a phenomenon called assortative mixing or homophily. Empirical age-stratified social contact matrices have been derived by extensive survey work. We lack however similar empirical studies that provide social contact matrices for a population stratified by attributes beyond age, such as gender, sexual orientation, or ethnicity. Accounting for heterogeneities with respect to these attributes can have a profound effect on model dynamics. Here, we introduce a new method, which uses linear algebra and non-linear optimization, to expand a given contact matrix to populations stratified by binary attributes with a known level of homophily. Using a standard epidemiological model, we highlight the effect homophily can have on model dynamics, and conclude by briefly describing more complicated extensions. The available Python source code enables any modeler to account for the presence of homophily with respect to binary attributes in contact patterns, ultimately yielding more accurate predictive models.

Multi-perspective social recommendation method with graph representation learning

Social recommender systems (SRS) aim to study how social relations influence users’ choices and how to use them for better learning users embeddings. However, the diversity of social relationships, which is instructive to the propagation of social influence, has been rarely explored. In this paper, we propose a graph convolutional network based representation learning method, namely multi-perspective social recommendation (MPSR), to construct hierarchical user preferences and assign friends’ influences with different levels of trust from varying perspectives. We further utilize the attributes of items to partition and excavate users’ explicit preferences and employ complementary perspective modeling to learn implicit preferences of users. To measure the trust degree of friends from different perspectives, the statistical information of users’ historical behavior is utilized to construct multi-perspective social networks. Experimental results on two public datasets of Yelp and Ciao demonstrate that the MPSR significantly outperforms the state-of-the-art methods. Further detailed analysis verifies the importance of mining explicit characteristics of users and the necessity for diverse social relationships, which show the rationality and effectiveness of the proposed model. The source Python code will be available upon request.

TRIPOD—A Treadmill Walking Dataset with IMU, Pressure-Distribution and Photoelectric Data for Gait Analysis

Inertial measurement units (IMUs) enable easy to operate and low-cost data recording for gait analysis. When combined with treadmill walking, a large number of steps can be collected in a controlled environment without the need of a dedicated gait analysis laboratory. In order to evaluate existing and novel IMU-based gait analysis algorithms for treadmill walking, a reference dataset that includes IMU data as well as reliable ground truth measurements for multiple participants and walking speeds is needed. This article provides a reference dataset consisting of 15 healthy young adults who walked on a treadmill at three different speeds. Data were acquired using seven IMUs placed on the lower body, two different reference systems (Zebris FDMT-HQ and OptoGait), and two RGB cameras. Additionally, in order to validate an existing IMU-based gait analysis algorithm using the dataset, an adaptable modular data analysis pipeline was built. Our results show agreement between the pressure-sensitive Zebris and the photoelectric OptoGait system (r = 0.99), demonstrating the quality of our reference data. As a use case, the performance of an algorithm originally designed for overground walking was tested on treadmill data using the data pipeline. The accuracy of stride length and stride time estimations was comparable to that reported in other studies with overground data, indicating that the algorithm is equally applicable to treadmill data. The Python source code of the data pipeline is publicly available, and the dataset will be provided by the authors upon request, enabling future evaluations of IMU gait analysis algorithms without the need of recording new data.

Deep learning ensemble method for classification of satellite hyperspectral images

Classification of hyperspectral image(HSI) is extensively utilized for the study of remotely sensed satellite images for various real-life applications. Convolutional Neural Networks (CNNs) are a commonly used deep learning technique for image data processing. The utilization of 2D CNNs and 3D CNNs have gained popularity in recent years for classification of hyperspectral image, and a lot of architectures with the combination of two have been proposed some of which include residual network based architectures. So, far individual models have been proposed and ensembling has not been explored much. In this paper, we propose an inception inspired architecture (IIA) and ensembled it with existing architectures HybridSN and inception residual network. The proposed IIA has 3D and 2D inception blocks which enable better spectral-spatial learning features. The Experiments are conducted on three well known publicly available HSI datasets and the results are compared to the state-of-the-art techniques. Experimental results yield that proposed deep learning ensemble method provides enhanced performance as compared to the state-of-the-art techniques. The python source code of the proposed method is available athttps://github.com/shyamfec/DL-Ensemble-Method.

A Python-WSGI and PHP-Apache Web Server Performance Analysis by Search Page Generator (SPG)

The web servers (WSGI-Python) and (PHP-Apache) are in middleware tier architecture. Middleware architecture is between frontend tier and backend tier, otherwise it’s a connection between frontend tier and backend tier for three tier architecture. The ELearning systems are designed by two different dynamic web technologies. First is by Python-WSGI and the second is by Personal Home Page (PHP-Apache). The two websites were designed with different open source and cross platform web technologies programming language namely; Python and PHP in the same structure and weight will evaluate perform over two different operating systems (OSs): 1) Windows-16 and 2) Linux-Ubuntu 20.4. Both systems run over the same computer architecture (64bit) as a server side with a common backend MySQL web database for both of them. Nevertheless, the middleware for PHP is a cross Apache MySQL PHP Perl (XAMPP), but the middleware for Python is Pycharm and the web server gateway interface (WSGI). WSGI and Apache are both web servers and this paper will show which of them has a better response time (RT). On the one hand, the experimental results demonstrate that the Python-WSGI is even weightier in Mbyte than PHP-Apache, on the other hand Python is still faster and more accurate than PHP. The designed SPG is by handwriting codes: one time designed the SPG by PHP source code and the other time designed by Python source code. Both Python-WSGI and PHP-Apache results are targeted to compare by the least time in milliseconds and take in to account enhanced performance.

PyBindingCurve, Simulation, and Curve Fitting to Complex Binding Systems at Equilibrium.

Understanding multicomponent binding interactions in protein–ligand, protein–protein, and competition systems is essential for fundamental biology and drug discovery. Hand-deriving equations quickly become unfeasible when the number of components is increased, and direct analytical solutions only exist to a certain complexity. To address this problem and allow easy access to simulation, plotting, and parameter fitting to complex systems at equilibrium, we present the Python package PyBindingCurve. We apply this software to explore homodimer and heterodimer formations culminating in the discovery that under certain conditions, homodimers are easier to break with an inhibitor than heterodimers and may also be more readily depleted. This is a potentially valuable and overlooked phenomenon of great importance to drug discovery. PyBindingCurve may be expanded to operate on any equilibrium binding system and allows definition of custom systems using a simple syntax. PyBindingCurve is available under the MIT license at https://github.com/stevenshave/pybindingcurve as the Python source code accompanied by examples and as an easily installable package within the Python Package Index.

The design of dynamic ensemble selection strategy for the error-correcting output codes family

Error-Correcting Output Codes (ECOC) is widely deployed to tackle the multiclass classification problem by reducing the original multi-class problem to several binary sub-problems. This study attempts to design a dynamic ensemble selection strategy to promote the performance of ECOC algorithms. Concretely, each column in a coding matrix is matched with a set of feature subsets generated by various feature selection methods. In the decoding process, a novel criterion based on the data complexity theory is proposed to pick up an optimal feature subset from the candidate subsets, so as to better distinguish unknown samples. As this strategy can be embedded in all types of ECOC algorithms, seven classical ECOC algorithms are deployed to verify the effectiveness of our strategy. Experiments are carried out on a set of UCI data sets, and the results confirm that despite different working principle, the proposed strategy can further improve the performance of various ECOC algorithms in most cases. Our python source code is available at: https://github.com/MLDMXM2017/ECOC_DES .

CARM: Confidence-aware recommender model via review representation learning and historical rating behavior in the online platforms

The recommendation systems in the online platforms often suffer from the rating data sparseness and information overload issues. Previous studies on this topic often leverage review information to construct an accurate user/item latent factor. To address this issue, we propose a novel confidence-aware recommender model via review representation learning and historical rating behavior in this article. It is motived that ratings are consistent with reviews in terms of user preferences, and reviews often contain misleading comments (e.g., fake good reviews, fake bad reviews). To this end, the interaction latent factor of user and item in the framework is constructed by exploiting review information interactivity. Then, the confidence matrix, which measures the relationship between the rating outliers and misleading reviews, is employed to further improve the model accuracy and reduce the impact of misleading reviews on the model. Furthermore, the loss function is constructed by maximum a posteriori estimation theory. Finally, the mini-batch gradient descent algorithm is introduced to optimize the loss function. Experiments conducted on four real-world datasets empirically demonstrate that our proposed method outperforms the state-of-the-art methods. The proposed method also further promotes the application in learning resource adaptation. The source Python code will be available upon request.

Quantifying Nanoparticle Assembly States in a Polymer Matrix through Deep Learning

There is great interest in controlling the spatial dispersion of inorganic nanoparticles (NPs) in an organic polymer matrix because this centrally underpins the property enhancements obtained from these hybrid materials. Currently, qualitative information on NP spatial distribution is obtained by visual inspection of transmission electron microscopy (TEM) images. Quantitative information is only indirectly obtained through the use of scattering probes such as small-angle X-ray/neutron scattering (SAXS/SANS). While the main challenge, that scattering probes operate in reciprocal space, can be remedied by Fourier inverting the data into real space, a much harder issue deconvolves the contribution of the particle form factor (which is affected by the details of the NP size and shape) from the structure factor that contains information on NP spatial distribution. These problems become acute when we deal with the popular topic of NPs grafted with polymer chains because the polymeric corona and hence the particle form factor becomes context-dependent and is hard to quantify. To make progress, we develop and apply a deep-learning-based image analysis method to quantify the distribution of spherical NPs in a polymer matrix directly from their real-space TEM images. A dataset of NP detection (DOPAD) is built by manually labeling particle positions on experimental TEM images of diverse polymer composite systems. A convolutional neural network object detection model is then trained on DOPAD. Together with sliding-window and merging algorithms, an automated pipeline is established, which takes a large TEM image as input and extracts NP locations and sizes. We validate the structural information resulting from this method against SAXS-derived structural information for NPs ordered by polymer crystallization and then use it to distinguish between different states of the assembly of polymer-grafted NPs in a polymer matrix achieved by using their surfactancy. We show that this data-rich protocol allows us to draw critical facets of experimental behavior which have previously not been accessible. The DOPAD dataset, Python source code, and trained model are shared on GitHub at https://dopad.github.io.

XAOM: A method for automatic alignment and orientation of radiographs for computer-aided medical diagnosis

Background and objectivesComputer-aided diagnosis relies on machine learning algorithms that require filtered and preprocessed data as the input. Aligning the image in the desired direction is an additional manual step in post-processing, commonly overlooked due to workload issues. Several state-of-the-art approaches for fracture detection and disease-struck region segmentation benefit from correctly oriented images, thus requiring such preprocessing of X-ray images. Furthermore, it is desirable to have archived studies in a standardized format. Radiograph hanging protocols also differ from case to case, which means that images are not always aligned and oriented correctly. As a solution, the paper proposes XAOM, an X-ray Alignment and Orientation Method for images from 21 different body regions. MethodsTypically, other methods are crafted for this purpose to suit a specific body region and form of usage. In contrast, the method proposed in this paper is comprehensive and easily tuned to align and orient X-ray images of any body region. XAOM consists of two stages. For the first stage of the method, aligning X-ray images, we experimented with the following approaches: Hough transform, Fast line detection algorithm, and Principal Component Analysis method. For the second stage, we have experimented with the adaptations of several well known convolutional neural network topologies for correctly predicting image orientation: LeNet5, AlexNet, VGG16, VGG19, and ResNet50. ResultsIn the first stage, the PCA-based approach performed best. The average difference between the angle detected by the algorithm and the angle marked by the experts on the test set containing 200 pediatric X-ray images was 1.65∘, while the median value was 0.11∘. In the second stage, the VGG16-based network topology achieved the best accuracy of 0.993 on a test set containing 4,221 images. ConclusionXAOM is highly accurate at aligning and orienting pediatric X-ray images of 21 common body regions according to a set standard. The proposed method is also robust and can be easily adjusted to the different alignment and rotation criteria. AvailabilityThe Python source code of the best performing implementation of XAOM is publicly available at https://github.com/fhrzic/XAOM.