Search

Medical image-to-image translation is often difficult and of limited effectiveness due to the differences in image acquisition mechanisms and the diverse structure of biological tissues. This work presents an unpaired image translation model between in-vivo optical coherence tomography (OCT) and ex-vivo Hematoxylin and eosin (H&E) stained images without the need for image stacking, registration, post-processing, and annotation. The model can generate high-quality and highly accurate virtual medical images, and is robust and bidirectional. Our framework introduces random noise to (1) blur redundant features, (2) defend against self-adversarial attacks, (3) stabilize inverse conversion, and (4) mitigate the impact of OCT speckles. We also demonstrate that our model can be pre-trained and then fine-tuned using images from different OCT systems in just a few epochs. Qualitative and quantitative comparisons with traditional image-to-image translation models show the robustness of our proposed signal-to-noise ratio (SNR) cycle-consistency method., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023. Published by Elsevier B.V.)

We present the results for CAPRI Round 54, the 5th joint CASP-CAPRI protein assembly prediction challenge. The Round offered 37 targets, including 14 homodimers, 3 homo-trimers, 13 heterodimers including 3 antibody-antigen complexes, and 7 large assemblies. On average ~70 CASP and CAPRI predictor groups, including more than 20 automatics servers, submitted models for each target. A total of 21 941 models submitted by these groups and by 15 CAPRI scorer groups were evaluated using the CAPRI model quality measures and the DockQ score consolidating these measures. The prediction performance was quantified by a weighted score based on the number of models of acceptable quality or higher submitted by each group among their five best models. Results show substantial progress achieved across a significant fraction of the 60+ participating groups. High-quality models were produced for about 40% of the targets compared to 8% two years earlier. This remarkable improvement is due to the wide use of the AlphaFold2 and AlphaFold2-Multimer software and the confidence metrics they provide. Notably, expanded sampling of candidate solutions by manipulating these deep learning inference engines, enriching multiple sequence alignments, or integration of advanced modeling tools, enabled top performing groups to exceed the performance of a standard AlphaFold2-Multimer version used as a yard stick. This notwithstanding, performance remained poor for complexes with antibodies and nanobodies, where evolutionary relationships between the binding partners are lacking, and for complexes featuring conformational flexibility, clearly indicating that the prediction of protein complexes remains a challenging problem., (© 2023 The Authors. Proteins: Structure, Function, and Bioinformatics published by Wiley Periodicals LLC.)

Covalent inhibitors have received extensive attentions in the past few decades because of their long residence time, high binding efficiency and strong selectivity. Therefore, it is valuable to develop computational tools like molecular docking for modeling of covalent protein-ligand interactions or screening of potential covalent drugs. Meeting the needs, we have proposed HCovDock, an efficient docking algorithm for covalent protein-ligand interactions by integrating a ligand sampling method of incremental construction and a scoring function with covalent bond-based energy. Tested on a benchmark containing 207 diverse protein-ligand complexes, HCovDock exhibits a significantly better performance than seven other state-of-the-art covalent docking programs (AutoDock, Cov&#95;DOX, CovDock, FITTED, GOLD, ICM-Pro and MOE). With the criterion of ligand root-mean-squared distance < 2.0 Å, HCovDock obtains a high success rate of 70.5% and 93.2% in reproducing experimentally observed structures for top 1 and top 10 predictions. In addition, HCovDock is also validated in virtual screening against 10 receptors of three proteins. HCovDock is computationally efficient and the average running time for docking a ligand is only 5 min with as fast as 1 sec for ligands with one rotatable bond and about 18 min for ligands with 23 rotational bonds. HCovDock can be freely assessed at http://huanglab.phys.hust.edu.cn/hcovdock/., (© The Author(s) 2022. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com.)

Protein-protein interactions are crucial in many biological processes. Therefore, determining the structure of a protein-protein complex is valuable for understanding its molecular mechanisms and developing drugs. Molecular docking is a powerful computational tool in the prediction of protein-protein complex structures, in which a scoring function with good performance is very important. In this study, we have proposed a hybrid scoring function of atomic contact-based desolvation energies and distance-dependent interatomic potentials for protein-protein interactions, named HITScorePP, where the atomic contact desolvation energies were derived using an iterative method and the distance-dependent potentials were directly taken from our ITScorePP scoring function. Integrating the hybrid scoring function into our fast Fourier transform (FFT) based HDOCK docking scheme, the updated docking program, named HDOCK2.0, significantly improved the docking performance on the 55 newly added complexes in the protein docking benchmark 5.0 and a data set of 19 antibacterial protein complexes. HDOCK2.0 was also compared with four other state-of-the-art docking programs including Rosetta, ZDOCK3.0.2, FRODOCK3.0, ATTRACT, and PatchDock and obtained the overall best performance in binding mode predictions. These results demonstrated the accuracy of our hybrid scoring function and the necessity of included desolvation effects in protein-protein docking.

Purpose: Treatment planning systems (TPSs) from different vendors can involve different implementations of Monte Carlo dose calculation (MCDC) algorithms for pencil beam scanning (PBS) proton therapy. There are currently no guidelines for validating non-water materials in TPSs. Furthermore, PBS-specific parameters can vary by 1-2 orders of magnitude among different treatment delivery systems (TDSs). This paper proposes a standardized framework on the use of commissioning data and steps to validate TDS-specific parameters and TPS-specific heterogeneity modeling to potentially reduce these uncertainties., Methods: A standardized commissioning framework was developed to commission the MCDC algorithms of RayStation 8A and Eclipse AcurosPT v13.7.20 using water and non-water materials. Measurements included Bragg peak depth-dose and lateral spot profiles and scanning field outputs for Varian ProBeam. The phase-space parameters were obtained from in-air measurements and the number of protons per MU from output measurements of 10 × 10 cm 2 square fields at a 2 cm depth. Spot profiles and various PBS field measurements at additional depths were used to validate TPS. Human tissues in TPS, Gammex phantom materials, and artificial materials were used for the TPS benchmark and validation., Results: The maximum differences of phase parameters, spot sigma, and divergence between MCDC algorithms are below 4.5 µm and 0.26 mrad in air, respectively. Comparing TPS to measurements at depths, both MC algorithms predict the spot sigma within 0.5 mm uncertainty intervals, the resolution of the measurement device. Beam Configuration in AcurosPT is found to underestimate number of protons per MU by ~2.5% and requires user adjustment to match measured data, while RayStation is within 1% of measurements using Auto model. A solid water phantom was used to validate the range accuracy of non-water materials within 1% in AcurosPT., Conclusions: The proposed standardized commissioning framework can detect potential issues during PBS TPS MCDC commissioning processes, and potentially can shorten commissioning time and improve dosimetric accuracies. Secondary MCDC can be used to identify the root sources of disagreement between primary MCDC and measurement., (© 2020 American Association of Physicists in Medicine.)

Purpose: Monte Carlo (MC) dose calculation is generally superior to analytical dose calculation (ADC) used in commercial TPS to model the dose distribution especially for heterogeneous sites, such as lung and head/neck patients. The purpose of this study was to provide a validated, fast, and open-source MC code, MCsquare, to assess the impact of approximations in ADC on clinical pencil beam scanning (PBS) plans covering various sites., Methods: First, MCsquare was validated using tissue-mimicking IROC lung phantom measurements as well as benchmarked with the general purpose Monte Carlo TOPAS for patient dose calculation. Then a comparative analysis between MCsquare and ADC was performed for a total of 50 patients with 10 patients per site (including liver, pelvis, brain, head-and-neck, and lung). Differences among TOPAS, MCsquare, and ADC were evaluated using four dosimetric indices based on the dose-volume histogram (target Dmean, D95, homogeneity index, V95), a 3D gamma index analysis (using 3%/3 mm criteria), and estimations of tumor control probability (TCP)., Results: Comparison between MCsquare and TOPAS showed less than 1.8% difference for all of the dosimetric indices/TCP values and resulted in a 3D gamma index passing rate for voxels within the target in excess of 99%. When comparing ADC and MCsquare, the variances of all the indices were found to increase as the degree of tissue heterogeneity increased. In the case of lung, the D95s for ADC were found to differ by as much as 6.5% from the corresponding MCsquare statistic. The median gamma index passing rate for voxels within the target volume decreased from 99.3% for liver to 75.8% for lung. Resulting TCP differences can be large for lung (≤10.5%) and head-and-neck (≤6.2%), while smaller for brain, pelvis and liver (≤1.5%)., Conclusions: Given the differences found in the analysis, accurate dose calculation algorithms such as Monte Carlo simulations are needed for proton therapy, especially for disease sites with high heterogeneity, such as head-and-neck and lung. The establishment of MCsquare can facilitate patient plan reviews at any institution and can potentially provide unbiased comparison in clinical trials given its accuracy, speed and open-source availability., (© 2018 American Association of Physicists in Medicine.)

Protein-ligand docking has been playing an important role in modern drug discovery. To model drug-target binding in real systems, a number of flexible-ligand docking algorithms with different sampling strategies and scoring methods have been subsequently developed over the past three decades, while rigid-ligand docking is still being used because of its compelling computational efficiency. Here, a comprehensive assessment has been conducted to investigate the effectiveness of flexible-ligand docking versus rigid-ligand docking for three representative docking algorithms (global optimization, incremental construction and multi-conformer docking) in virtual screening and pose prediction on the Directory of Useful Decoys. It was found that overall flexible-ligand docking did not achieve a statistically significant improvement in enrichments over rigid-ligand docking in virtual screening, but all docking programs significantly improved the success rates when considering ligand flexibility in pose prediction. The worse effectiveness of flexible-ligand docking in virtual screening than in pose prediction suggests that the challenges of current docking algorithms exist in ranking more than docking, although the use of flexible-ligand docking in virtual screening was justified by its better effectiveness for more flexible ligand in virtual screening. Challenges for scoring, including internal energy, charge polarization, entropy and flexibility, were investigated and discussed. An empirical way was also proposed to consider loss of ligand conformational entropy for virtual screening.

A major subclass of protein-protein interactions is formed by homo-oligomers with certain symmetry. Therefore, computational modeling of the symmetric protein complexes is important for understanding the molecular mechanism of related biological processes. Although several symmetric docking algorithms have been developed for Cn symmetry, few docking servers have been proposed for Dn symmetry. Here, we present HSYMDOCK, a web server of our hierarchical symmetric docking algorithm that supports both Cn and Dn symmetry. The HSYMDOCK server was extensively evaluated on three benchmarks of symmetric protein complexes, including the 20 CASP11-CAPRI30 homo-oligomer targets, the symmetric docking benchmark of 213 Cn targets and 35 Dn targets, and a nonredundant test set of 55 transmembrane proteins. It was shown that HSYMDOCK obtained a significantly better performance than other similar docking algorithms. The server supports both sequence and structure inputs for the monomer/subunit. Users have an option to provide the symmetry type of the complex, or the server can predict the symmetry type automatically. The docking process is fast and on average consumes 10∼20 min for a docking job. The HSYMDOCK web server is available at http://huanglab.phys.hust.edu.cn/hsymdock/.

Protein-peptide interactions are crucial in many cellular functions. Therefore, determining the structure of protein-peptide complexes is important for understanding the molecular mechanism of related biological processes and developing peptide drugs. HPEPDOCK is a novel web server for blind protein-peptide docking through a hierarchical algorithm. Instead of running lengthy simulations to refine peptide conformations, HPEPDOCK considers the peptide flexibility through an ensemble of peptide conformations generated by our MODPEP program. For blind global peptide docking, HPEPDOCK obtained a success rate of 33.3% in binding mode prediction on a benchmark of 57 unbound cases when the top 10 models were considered, compared to 21.1% for pepATTRACT server. HPEPDOCK also performed well in docking against homology models and obtained a success rate of 29.8% within top 10 predictions. For local peptide docking, HPEPDOCK achieved a high success rate of 72.6% on a benchmark of 62 unbound cases within top 10 predictions, compared to 45.2% for HADDOCK peptide protocol. Our HPEPDOCK server is computationally efficient and consumed an average of 29.8 mins for a global peptide docking job and 14.2 mins for a local peptide docking job. The HPEPDOCK web server is available at http://huanglab.phys.hust.edu.cn/hpepdock/.

AcurosPT is a Monte Carlo algorithm in the Eclipse 13.7 treatment planning system, which is designed to provide rapid and accurate dose calculations for proton therapy. Computational run-time in minimized by simplifying or eliminating less significant physics processes. In this article, the accuracy of AcurosPT was benchmarked against both measurement and an independent MC calculation, TOPAS. Such a method can be applied to any new MC calculation for the detection of potential inaccuracies. To validate multiple Coulomb scattering (MCS) which affects primary beam broadening, single spot profiles in a Solidwater ® phantom were compared for beams of five selected proton energies between AcurosPT, measurement and TOPAS. The spot Gaussian sigma in AcurosPT was found to increase faster with depth than both measurement and TOPAS, suggesting that the MCS algorithm in AcurosPT overestimates the scattering effect. To validate AcurosPT modeling of the halo component beyond primary beam broadening, field size factors (FSF) were compared for multi-spot profiles measured in a water phantom. The FSF for small field sizes were found to disagree with measurement, with the disagreement increasing with depth. Conversely, TOPAS simulations of the same FSF consistently agreed with measurement to within 1.5%. The disagreement in absolute dose between AcurosPT and measurement was smaller than 2% at the mid-range depth of multi-energy beams. While AcurosPT calculates acceptable dose distributions for typical clinical beams, users are cautioned of potentially larger errors at distal depths due to overestimated MCS and halo implementation., (© 2017 The Authors. Journal of Applied Clinical Medical Physics published by Wiley Periodicals, Inc. on behalf of American Association of Physicists in Medicine.)

Protein-protein docking is an important computational tool for predicting protein-protein interactions. With the rapid development of proteomics projects, more and more experimental binding information ranging from mutagenesis data to three-dimensional structures of protein complexes are becoming available. Therefore, how to appropriately incorporate the biological information into traditional ab initio docking has been an important issue and challenge in the field of protein-protein docking. To address these challenges, we have developed a Hybrid DOCKing protocol of template-based and template-free approaches, referred to as HDOCK. The basic procedure of HDOCK is to model the structures of individual components based on the template complex by a template-based method if a template is available; otherwise, the component structures will be modeled based on monomer proteins by regular homology modeling. Then, the complex structure of the component models is predicted by traditional protein-protein docking. With the HDOCK protocol, we have participated in the CPARI experiment for rounds 28-35. Out of the 25 CASP-CAPRI targets for oligomer modeling, our HDOCK protocol predicted correct models for 16 targets, ranking one of the top algorithms in this challenge. Our docking method also made correct predictions on other CAPRI challenges such as protein-peptide binding for 6 out of 8 targets and water predictions for 2 out of 2 targets. The advantage of our hybrid docking approach over pure template-based docking was further confirmed by a comparative evaluation on 20 CASP-CAPRI targets. Proteins 2017; 85:497-512. © 2016 Wiley Periodicals, Inc., (© 2016 Wiley Periodicals, Inc.)

Inspiratory flow limitation (IFL) is a critical symptom of sleep breathing disorders. A characteristic flattened flow-time curve indicates the presence of highest resistance flow limitation. This study involved investigating a real-time algorithm for detecting IFL during sleep. Three categories of inspiratory flow shape were collected from previous studies for use as a development set. Of these, 16 cases were labeled as non-IFL and 78 as IFL which were further categorized into minor level (20 cases) and severe level (58 cases) of obstruction. In this study, algorithms using polynomial functions were proposed for extracting the features of IFL. Methods using first- to third-order polynomial approximations were applied to calculate the fitting curve to obtain the mean absolute error. The proposed algorithm is described by the weighted third-order (w.3rd-order) polynomial function. For validation, a total of 1,093 inspiratory breaths were acquired as a test set. The accuracy levels of the classifications produced by the presented feature detection methods were analyzed, and the performance levels were compared using a misclassification cobweb. According to the results, the algorithm using the w.3rd-order polynomial approximation achieved an accuracy of 94.14% for IFL classification. We concluded that this algorithm achieved effective automatic IFL detection during sleep.

Ultrasound tissue characterization is crucial for the detection of tissue abnormalities. Since the statistics of the backscattered ultrasound signals strongly depend on density and spatial arrangement of local scatterers, appropriate modeling of the backscattered signals may be capable of providing unique physiological information on local tissue properties. Among various techniques, the Nakagami imaging, realized in a window-based estimation scheme, has a good performance in assessing different scatterer statistics in tissues. However, inconsistent m values have been reported in literature and obtained only from a local tissue region, abating the reliability of Nakagami imaging in tissue characterization. The discrepancies in m values in relevant literature may stem from the nonuniformity of the ultrasound image resolution, which is often neglected. We therefore hypothesized that window-based Nakagami m estimation was highly associated with the regional spatial resolution of ultrasound imaging. To test this hypothesis, our study investigated the effect of beamforming methods, including synthetic aperture (SA), coherent plane wave compounding (CPWC), multi-focusing (MF), and single-focusing (SF), on window-based m parameter estimation from the perspective of the resolution cell. The statistics of m parameter distribution as a function of imaging depth were characterized by their mean, variance, and skewness. The phantom with a low scatterer density (16 scatterers mm(-3)) had significantly lower m values compared to the ones with high scatterer densities (32 and 64 scatterers mm(-3)). Results from the homogeneous phantom with 64 scatterers mm(-3) showed that SA, MF, and CPWC had relatively uniform lateral resolutions compared to SF and thus relatively constant m estimates at different imaging depths. Our findings suggest that an ultrasound imaging regime exhibiting invariant spatial resolution throughout the entire imaging field of view would be the most appropriate for Nakagami imaging for tissue characterization.

Protein-protein docking is an important computational tool for studying protein-protein interactions. A variety of docking programs with different sampling algorithms and scoring functions as well as computational efficiencies have been subsequently developed over the last decades. Here, we have reviewed the trend and performance of current global docking programs through a comprehensive assessment of the 18 docking/scoring protocols of 14 global docking programs on the latest protein docking benchmark 4.0. The effects of docking algorithms, interaction types, and conformational changes on the docking performance were investigated and discussed. The findings are expected to provide a general guideline for the choice of an appropriate docking protocol and offer insights into the optimization and development of docking and scoring algorithms., (Copyright © 2015 Elsevier Ltd. All rights reserved.)

Objective: To evaluate the feasibility of clinical application for genetic based dosing algorithm in the predication of warfarin maintenance dose in Chinese population., Methods: The clinical data were collected and blood samples harvested from a total of 126 patients undergoing heart valve replacement. The genotypes of VKORC1 and CYP2C9 were determined by melting curve analysis after PCR. They were divided randomly into the study and control groups. In the study group, the first three doses of warfarin were prescribed according to the predicted warfarin maintenance dose while warfarin was initiated at 2.5 mg/d in the control group. The warfarin doses were adjusted according to the measured international normalized ratio (INR) values. And all subjects were followed for 50 days after an initiation of warfarin therapy., Results: At the end of a 50-day follow-up period, the proportions of the patients on a stable dose were 82.4% (42/51) and 62.5% (30/48) for the study and control groups respectively. The mean durations of reaching a stable dose of warfarin were (27.5 ± 1.8) and (34.7 ± 1.8) days and the median durations were (24.0 ± 1.7) and (33.0 ± 4.5) days in the study and control groups respectively. Significant differences existed in the durations of reaching a stable dose between the two groups (P = 0.012). Compared with the control group, the hazard ratio (HR) for the duration of reaching a stable dose was 1.786 in the study group (95%CI 1.088 - 2.875, P = 0.026)., Conclusion: The predicted dosing algorithm incorporating genetic and non-genetic factors may shorten the duration of achieving efficiently a stable dose of warfarin. And the present study validates the feasibility of its clinical application.

A hierarchical approach has been developed for protein-protein docking. In the first step, a Fast Fourier Transform (FFT)-based docking algorithm is used to globally sample all putative binding modes, in which the protein is represented by a reduced model, that is, each side chain on the protein surface is represented by its center of mass. Compared to conventional FFT docking with all-atom models, the FFT docking method with a reduced model is expected to generate more hits because it allows larger side-chain flexibility. Next, the filtered binding modes (normally several thousands) are refined by an iteratively derived knowledge-based scoring function ITScorePP and by considering backbone/loop flexibility using an ensemble docking algorithm. The distance-dependent potentials of ITScorePP were extracted by a physics-based iterative method, which circumvents the long-standing reference state problem in the knowledge-based approaches. With this hierarchical protocol, we have participated in the CAPRI experiments for Rounds 15-19 of 11 targets (T32-T42). In the predictor experiments, we achieved correct binding modes for six targets: three are with high accuracy (T40 for both distinct binding modes, T41, and T42), two are with medium accuracy (T34 and T37), and one is acceptable (T32). In the scorer experiments, of the seven target complexes that contain at least one acceptable mode submitted by the CAPRI predictor groups, we obtained correct binding modes for four targets: three are with high accuracy (T37, T40, and T41) and one is with medium accuracy (T34), suggesting good accuracy and robustness of ITScorePP., (© 2010 Wiley-Liss, Inc.)

Correlation-based speckle tracking methods are commonly used in elasticity imaging to estimate displacements. In the presence of local strain, a larger window size results in larger displacement error. To reduce tracking error, we proposed a short correlation window followed by a correlation coefficient filter. Although simulation and experimental results demonstrated the efficacy of the method, it was not clear why correlation coefficient filtering reduces tracking error since tracking error increases if normalization before filtering is not applied. In this paper, we analyzed tracking errors by estimating phase variances of the cross-correlation function and the correlation coefficient at the true time lag based on statistical properties of these functions' real and imaginary parts. The role of normalization is clarified by identifying the effect of the cross-correlation function's amplitude fluctuation on the function's imaginary part. Furthermore, we present analytic forms for predicting axial displacement error as a function of strain, system parameters (signal-to-noise ratio, center frequency, and signal and noise bandwidths), and tracking parameters (window and filter sizes) for cases with and without normalization before filtering. Simulation results correspond to theory well for both noise-free cases and general cases with an empirical correction term included for strains up to 4%.

Surgery to correct severe heart arrhythmias usually requires detailed maps of the cardiac activation wave prior to ablation. The pinpoint electrical mapping procedure is laborious and limited by its spatial resolution (5-10 mm). We propose ultrasound current source density imaging (UCSDI), a direct 3-D imaging technique that potentially facilitates existing mapping procedures with superior spatial resolution. The technique is based on a pressure-induced change in resistivity known as the acoustoelectric (AE) effect, which is spatially confined to the ultrasound focus. AE-modulated voltage recordings are used to map and reconstruct current densities. In this preliminary study, we tested UCSDI under controlled conditions and compared it with conventional electrical mapping techniques. A 2-D dipole field was produced by a pair of electrodes in a bath of 0.9% NaCl solution. Boundary electrodes detected the AE signal while a 7.5-MHz focused ultrasound transducer was scanned across the bath. UCSDI located the current source and sink to within 1 mm of their actual positions. A future UCSDI system potentially provides real-time 3-D images of the cardiac activation wave coregistered with anatomical ultrasound and would greatly facilitate corrective procedures for heart abnormalities.

Tumor vascularity is an important factor that has been shown to correlate with tumor malignancy and was demonstrated as a prognostic indicator for a wide range of cancers. Three-dimensional (3-D) power Doppler ultrasound (PDUS) offers a convenient tool for investigators to inspect the signals of blood flow and vascular structures in breast cancer. In this paper, a new computer-aided diagnosis (CAD) system for quantifying Doppler ultrasound images based on 3-D thinning algorithm and neural network is proposed. We extracted the skeleton of blood vessels from 3-D PDUS data to facilitate the capturing of morphological changes. Nine features including vessel-to-volume ratio, number of vascular trees, length of vessels, number of branching, mean of radius, number of cycles, and three tortuosity measures, were extracted from the thinning result. Benign and malignant tumors can therefore be differentiated by a score computed by a multilayered perceptron (MLP) neural network using these features as parameters. The proposed system was tested on 221 breast tumors, including 110 benign and 111 malignant lesions. The accuracy, sensitivity, specificity, and positive and negative predictive values were 88.69% (196/221), 91.89% (102/111), 85.45% (94/110), 86.44% (102/118), and 91.26% (94/103), respectively. The Az value of the ROC curve was 0.94. The results demonstrate a correlation between the morphology of blood vessels and tumor malignancy, indicating that the newly proposed method can retrieves a high accuracy in the classification of benign and malignant breast tumors.

In this study, photoacoustic flow measurement methods based on wash-in analysis are presented. These methods use the rod-to-sphere shape transformations of gold nanorods induced by pulsed-laser irradiation. Due to the shape dependence of the optical absorption of the gold nanorods, these shape transitions are associated with a change in the peak optical absorption wavelength. Pulsed-laser irradiation at the wavelength corresponding to the peak optical absorption of the original gold nanorods allows the particles that undergo shape changes to be viewed as "being destructed" by the laser irradiation at that wavelength, hence, flow information can be derived from the change in ultrasound intensity that is directly related to the wash-in rate of the gold nanorods and the laser intensity. Two flow estimation methods based on the wash-in analysis are described. The first method first applies high-energy laser pulses that induce shape changes in all the nanorods. A series of low-energy pulses then are applied to monitor the acoustic signal change as new nanorods flow into the region of interest. The second method uses single-energy laser pulses such that the "destruction" and "detection" are performed simultaneously. The simulation results show that it is valid to fit the time-intensity curves by exponential models. To demonstrate the validity of the proposed methods, an Nd:YAG pulsed laser operating at 1064 nm was used for optical irradiation, and a 1-MHz ultrasonic transducer was used for acoustic detection. Gold nanorods with a peak optical absorption at 1018 nm and a concentration of 0.26 nM were used to estimate flow velocities ranging from 0.35 to 2.83 mm/s. The linear regression results show that the correlation coefficients between the measured velocities and the true values are close to unity (> or = 0.94), thus demonstrating the feasibility of the proposed photoacoustic techniques for relative flow estimation.

Polymer microring resonators are demonstrated as high-frequency, ultrasound detectors. An optical microring resonator consists of a ring waveguide closely coupled to a straight bus waveguide, serving as light input and output. Acoustic waves irradiating the ring induce strain, deforming the waveguide dimensions and changing the refractive index of the waveguide via the elasto-optic effect. These effects modify the effective refractive index of the guided mode inside the waveguide. The sharp wavelength dependence of the microring resonance can enhance the optical response to acoustic strain. Such polymer microring resonators are experimentally demonstrated in detecting broadband ultrasound pulses from a 50 MHz transducer. Measured frequency response shows that these devices have potential in high-frequency, ultrasound detection. Design guidelines for polymer microring resonators forming an ultrasound detector array are discussed.

This paper proposes an approach to designing binary codes suitable for high-frequency applications of coded excitation in medical ultrasound. For a high-frequency ultrasound system, transmitting well-designed binary codes with a low sampling ratio (i.e., the bit rate divided by the transducer center frequency) is a practical way to improve the signal-to-noise ratio (SNR) because the challenge of implementing arbitrary-waveform generators for transmitting nonbinary codes increases with the frequency and the switching speed of square-wave pulsers are limited. One conventional approach designs codes using a base sequence that modulates wideband sequences up to the transducer passband. Because a major portion of codes is excluded as a candidate, codes designed using this approach typically need long compression filters for restoring the axial resolution, and they do not improve the SNR efficiently. In contrast, the approach proposed here searches all the codes that match the transducer passband; hence, the resultant codes exhibit better performance. The technique was tested using a bit rate of 50 MHz and a sampling ratio of 2. For a transducer with an ideal Gaussian frequency response with a center frequency of 25 MHz and a -6 dB bandwidth of 15 MHz, the SNR for the same side-lobe extent was 1 to 6 dB higher for the codes designed using the proposed approach compared with those designed using the conventional approach. When a real transducer response with a center frequency of 26.4 MHz and a one-way -6 dB bandwidth of 20.7 MHz was considered, the codes designed using the proposed approach were superior by 0.5 to 5 dB. Therefore, our approach is better than the conventional approach for designing binary codes for high-frequency ultrasound, with the results indicating that the moderate bit rate of 50 MHz will suffice when the ultrasonic center frequency is 25 MHz.

One approach to incorporate protein flexibility in molecular docking is the use of an ensemble consisting of multiple protein structures. Sequentially docking each ligand into a large number of protein structures is computationally too expensive to allow large-scale database screening. It is challenging to achieve a good balance between docking accuracy and computational efficiency. In this work, we have developed a fast, novel docking algorithm utilizing multiple protein structures, referred to as ensemble docking, to account for protein structural variations. The algorithm can simultaneously dock a ligand into an ensemble of protein structures and automatically select an optimal protein structure that best fits the ligand by optimizing both ligand coordinates and the conformational variable m, where m represents the m-th structure in the protein ensemble. The docking algorithm was validated on 10 protein ensembles containing 105 crystal structures and 87 ligands in terms of binding mode and energy score predictions. A success rate of 93% was obtained with the criterion of root-mean-square deviation <2.5 A if the top five orientations for each ligand were considered, comparable to that of sequential docking in which scores for individual docking are merged into one list by re-ranking, and significantly better than that of single rigid-receptor docking (75% on average). Similar trends were also observed in binding score predictions and enrichment tests of virtual database screening. The ensemble docking algorithm is computationally efficient, with a computational time comparable to that for docking a ligand into a single protein structure. In contrast, the computational time for the sequential docking method increases linearly with the number of protein structures in the ensemble. The algorithm was further evaluated using a more realistic ensemble in which the corresponding bound protein structures of inhibitors were excluded. The results show that ensemble docking successfully predicts the binding modes of the inhibitors, and discriminates the inhibitors from a set of noninhibitors with similar chemical properties. Although multiple experimental structures were used in the present work, our algorithm can be easily applied to multiple protein conformations generated by computational methods, and helps improve the efficiency of other existing multiple protein structure(MPS)-based methods to accommodate protein flexibility., (Copyright 2006 Wiley-Liss, Inc.)

Angiogenesis provides blood supply for tumor expansion and also increases the opportunity for tumor cells to enter the blood or lymph circulation. Several proangiogenic factors as well as the contribution of the microenvironment to tumor-induced angiogenesis have been identified. Among these, vascular endothelial growth factor (VEGF) and the angiopoietin (Ang) family play a predominant role involved in the growth for endothelial cells. Tumor vessels are structurally and functionally abnormal because of an imbalance of these angiogenic regulators. In contrast to normal vessels, tumor vasculature is highly disorganized, tortuous and dilated, with uneven diameter and excessive branching. In other words, the morphologic features are likely to carry additional clues that, when used in conjunction with more established parameters, can improve the present diagnostic approaches. In our study, we present a new method that helps to capture the morphologic features from three-dimensional (3-D) power Doppler ultrasound (PDUS) images. After narrowing down the vessels into their skeletons using a 3-D thinning algorithm, we extracted seven features including vessel-to-volume ratio, number of vascular trees, number of bifurcation, mean of radius and three tortuosity measures, from the skeleton and applied a neural network to classify the tumors by using these features. In investigations into 221 solid breast tumors, including 110 benign and 111 malignant cases, the p values using the Student's t-test for all features were less than 0.05, indicating that the proposed features were deemed statistically significant. The A(Z) values for these seven features were 0.84, 0.87, 0.84, 0.75, 0.77, 0.79 and 0.69, respectively. The accuracy, sensitivity, specificity, and positive and negative predictive values were 80.09% (177 of 221), 80.18% (89 of 111), 80% (88 of 110), 80.18% (89 of 111) and 80% (88 of 110), respectively, with an A(Z) value of 0.89. The preliminary results show that the proposed method is feasible and has a good agreement with the diagnosis of the pathologists.

This paper presents a new coded excitation scheme that efficiently synthesizes codes for arbitrary waveforms using a bipolar square wave pulser. In a coded excitation system, pulse compression is performed to restore the axial resolution. In order to maintain low range sidelobes, the system needs to transmit signals that have smooth spectra. However, such a transmitter requires the generation of arbitrary waveforms and, therefore, is more expensive. In other words, a trade-off is necessary between the compression performance and the transmitter cost. Here we propose a method that preserves the low-cost advantage of a bipolar pulser while achieving approximately the same compression performance as an arbitrary waveform generator. The key idea of the proposed method is the conversion of a nonbinary code (i.e., requiring an arbitrary waveform generator) with good compression performance into a binary code (i.e., requiring only a bipolar pulser) by code translation and code tuning. The code translation is implemented by sending the nonbinary code into a virtual one-bit, sigma-delta modulator, and the code tuning involves minimizing the root-mean-square error between the resultant binary code and the original nonbinary code by sequential and iterative tuning while taking the transducer response into account. Tukey-windowed chirps are known to have good compression performance. Such chirps of different durations (16, 20, and 24 micros), all with a taper ratio of 0.15, a center frequency of 2.5 MHz, and an equivalent bandwidth of 1.5 MHz, were converted into binary Tukey-windowed chirps that were compared with pseudochirps (i.e., direct binary approximations of the original chirp) over the same spectral band. The bit rate was 40 MHz. Simulation results show that the use of binary Tukey-windowed chirps can reduce the code duration by 20.6% or the peak sidelobe level by 6 dB compared to the commonly used pseudochirps. Experimental results obtained under the same settings were in agreement with the simulations. Our results demonstrate that arbitrary waveform coded excitation can be realized using bipolar square wave pulsers for applications in medical ultrasound.

Sidelobe contribution from off-axis targets degrades image quality in a coherent array imaging system. In ultrasound imaging, focusing errors resulting from sound-velocity inhomogeneities in human tissue--also known as phase aberrations reduce the coherence of the received signals and elevate the sidelobe level. This paper proposes an adaptive receive-aperture technique based on thresholding of the coherence factor (CF). The CF describes the coherence of the received array signals and can be used as an index of focusing quality. This paper demonstrates that thresholding of the CF allows the mainlobe-dominated signals to be distinguished from the sidelobe-dominated signals, after which the receive-aperture size at each imaging position can be optimally determined so as to enhance the mainlobe-dominated signals and suppress the sidelobe-dominated signals. Thus, image quality degradation resulting from sound-velocity inhomogeneities can be reduced. Simulations and measured ultrasound data are used to evaluate the efficacy of the proposed technique. The characteristics of the proposed technique including the effects of the signal-to-noise ratio (SNR) and the transmit focal depth, and speckle reduction are discussed. The proposed technique is also compared with the parallel adaptive receive compensation algorithm and shown to produce a better improvement in image quality.

An approach based on limited-angle transmission tomography for reconstruction of the sound velocity distribution in the breast is proposed. The imaging setup is similar to that of x-ray mammography. With this setup, the time-of-flight data are acquired by a linear array positioned at the top of the compressed breast that both transmits and receives, and a metal plate is placed at the bottom as a reflector. The setup allows acoustic data acquisition for simultaneous B-mode image formation and the tomographic sound velocity reconstruction. In order to improve the sound velocity estimation accuracy, a new reconstruction algorithm based on a convex programming formulation has been developed. Extensive simulations for both imaging and time-of-flight data based on a 5-MHz linear array were performed on tissues with different geometries and acoustic parameters. Results show that the sound velocity error was generally 1-3 m/s, with a maximum of 5.8 m/s. The radii of the objects under investigation varied from 2 to 6 mm, and all of them were detected successfully. Thus, the proposed approach has been shown to be both feasible and accurate. The approach can be used to complement conventional B-mode imaging to further enhance the detection of breast cancer.

Spiculation is a stellate distortion caused by the intrusion of breast cancer into surrounding tissue. Its existence is an important clue to characterizing malignant tumors. Many successful mammographic methods have been proposed to detect tumors with spiculation. Traditional two-dimensional (2-D) ultrasound cannot easily find spiculations because spiculations normally appear parallel to the surface of the skin. Recently, three-dimensional (3-D) ultrasound has been gradually used in clinical applications and it has been proven to be useful in determining the architectural distortion or spiculation that surrounds a breast tumor. This paper aims to identify spiculation from 3-D ultrasonic volume data of a tumor found by a physician. In the proposed method, each coronal slice of volume data is successively extracted and then analyzed as a 2-D ultrasound image by the proposed spiculation detection method. First, in each horizontal slice, the modified rotating structuring element (ROSE) operation is used to find the central region in which spiculation lines converge. Second, the stick algorithm is used to estimate the direction of the edge of each pixel around the central region. A pixel whose edge points toward the central region is marked as a potential spiculation. Finally, the marked pixels are collected around the central region and their distribution is analyzed to determine whether spiculation is present. The 3-D test datasets were obtained using the Voluson 530 or 730, Kretztechnik, Austria. First, the proposed method was tested on 104 2-D typical coronal images (selected by an experienced physician) extracted from 52 3-D ultrasonic datasets. Finally, 225 3-D pathologically proven datasets were tested to evaluate the performance. Spiculations are more easily observed in the coronal view than in the other two views. That is, the 3-D ultrasound is a powerful tool for identifying spiculations. Furthermore, 16% (19/120) of benign cases and 90% (94/105) of malignant cases are detected as spiculations.

Monte Carlo (MC)-based dose calculations are generally superior to analytical dose calculations (ADC) in modeling the dose distribution for proton pencil beam scanning (PBS) treatments. The purpose of this paper is to present a methodology for commissioning and validating an accurate MC code for PBS utilizing a parameterized source model, including an implementation of a range shifter, that can independently check the ADC in commercial treatment planning system (TPS) and fast Monte Carlo dose calculation in opensource platform (MCsquare). The source model parameters (including beam size, angular divergence and energy spread) and protons per MU were extracted and tuned at the nozzle exit by comparing Tool for Particle Simulation (TOPAS) simulations with a series of commissioning measurements using scintillation screen/CCD camera detector and ionization chambers. The range shifter was simulated as an independent object with geometric and material information. The MC calculation platform was validated through comprehensive measurements of single spots, field size factors (FSF) and three-dimensional dose distributions of spread-out Bragg peaks (SOBPs), both without and with the range shifter. Differences in field size factors and absolute output at various depths of SOBPs between measurement and simulation were within 2.2%, with and without a range shifter, indicating an accurate source model. TOPAS was also validated against anthropomorphic lung phantom measurements. Comparison of dose distributions and DVHs for representative liver and lung cases between independent MC and analytical dose calculations from a commercial TPS further highlights the limitations of the ADC in situations of highly heterogeneous geometries. The fast MC platform has been implemented within our clinical practice to provide additional independent dose validation/QA of the commercial ADC for patient plans. Using the independent MC, we can more efficiently commission ADC by reducing the amount of measured data required for low dose "halo" modeling, especially when a range shifter is employed.

Simple Summary: We proposed a deep learning algorithm to detect lymph nodes and classify them in the head and neck region on computed tomography. We further analyzed the inference result from the model and found that the size of the lymph nodes may be a characteristic for the model to classify them. This finding is consistent with current clinical aspects. We will deploy the model in clinical practice and hope to assist clinicians in finding out the lesions more correctly and efficiently. Background: Head and neck cancer is highly prevalent in Taiwan. Its treatment mainly relies on clinical staging, usually diagnosed from images. A major part of the diagnosis is whether lymph nodes are involved in the tumor. We present an algorithm for analyzing clinical images that integrates a deep learning model with image processing and attempt to analyze the features it uses to classify lymph nodes. Methods: We retrospectively collected pretreatment computed tomography images and surgery pathological reports for 271 patients diagnosed with, and subsequently treated for, naïve oral cavity, oropharynx, hypopharynx, and larynx cancer between 2008 and 2018. We chose a 3D UNet model trained for semantic segmentation, which was evaluated for inference in a test dataset of 29 patients. Results: We annotated 2527 lymph nodes. The detection rate of all lymph nodes was 80%, and Dice score was 0.71. The model has a better detection rate at larger lymph nodes. For those identified lymph nodes, we found a trend where the shorter the short axis, the more negative the lymph nodes. This is consistent with clinical observations. Conclusions: The model showed a convincible lymph node detection on clinical images. We will evaluate and further improve the model in collaboration with clinical physicians. [ABSTRACT FROM AUTHOR]

The technique of human action recognition can be applied in a number of fields, such as medical rehabilitation, posture in the domain of sports, and emotion perception in counselling. Deep learning action recognition models that consider the continuous changes in the human skeleton can efficiently identify action states, such as rehabilitation processes, posture, and changes in emotion. However, the keypoints of the human skeleton are susceptible to occlusion due to body movement, resulting in a lower confidence level in their positions, which in turn affects the training and execution of human action recognition models. This paper proposes a keypoint correction algorithm based on continuous-time images of temporal variation that can predict and correct the skeleton keypoints with low confidence, based on the previous and subsequent multiple images. The proposed temporal-variation skeleton keypoint correction algorithm can provide accurate feature association training for the locations of the skeleton keypoints. The proposed method is shown to improve the accuracy of human action recognition by at least 30% over three alternative models in the literature: STV-GCN, GCN-NAS, and 2S-AGCN. The proposed algorithm is also improved by at least 30% compared with the above-mentioned methods in the literature, based on an action recognition model with a Gaussian filter. [ABSTRACT FROM AUTHOR]

Background: This study aims to develop a machine learning-based application in a real-world medical domain to assist anesthesiologists in assessing the risk of complications in patients after a hip surgery. Methods: Data from adult patients who underwent hip repair surgery at Chi-Mei Medical Center and its 2 branch hospitals from January 1, 2013, to March 31, 2020, were analyzed. Patients with incomplete data were excluded. A total of 22 features were included in the algorithms, including demographics, comorbidities, and major preoperative laboratory data from the database. The primary outcome was a composite of adverse events (in-hospital mortality, acute myocardial infarction, stroke, respiratory, hepatic and renal failure, and sepsis). Secondary outcomes were intensive care unit (ICU) admission and prolonged length of stay (PLOS). The data obtained were imported into 7 machine learning algorithms to predict the risk of adverse outcomes. Seventy percent of the data were randomly selected for training, leaving 30% for testing. The performances of the models were evaluated by the area under the receiver operating characteristic curve (AUROC). The optimal algorithm with the highest AUROC was used to build a web-based application, then integrated into the hospital information system (HIS) for clinical use. Results: Data from 4,448 patients were analyzed; 102 (2.3%), 160 (3.6%), and 401 (9.0%) patients had primary composite adverse outcomes, ICU admission, and PLOS, respectively. Our optimal model had a superior performance (AUROC by DeLong test) than that of ASA-PS in predicting the primary composite outcomes (0.810 vs. 0.629, p < 0.01), ICU admission (0.835 vs. 0.692, p < 0.01), and PLOS (0.832 vs. 0.618, p < 0.01). Conclusions: The hospital-specific machine learning model outperformed the ASA-PS in risk assessment. This web-based application gained high satisfaction from anesthesiologists after online use. [ABSTRACT FROM AUTHOR]

Featured Application: This study proposes a new path planning method called the navigation strategy with path priority for multiple robots moving. It also completely designs the robot's action strategy. The proposed study is suitable for handling or moving more than one robot in a trackless factory environment. This paper proposes a new path planning strategy called the navigation strategy with path priority (NSPP) for multiple robots moving in a large flat space. In the space, there may be some static or/and dynamic obstacles. Suppose we have the path-priority order for each robot, then this article aims to find an efficient path for each robot from its starting point to its target point without any collision. Here, a generalized Voronoi diagram (GVD) is used to perform the map division based on each robot's path-priority order, and the proposed NSPP is used to do the path planning for the robots in the space. This NSPP can be applied to any number of robots. At last, there are several simulations with a different number of robots in a circular or rectangular space to be shown that the proposed method can complete the task effectively and has better performance in average trajectory length than those by using the benchmark methods of the shortest distance algorithm (SDA) and reciprocal orientation algorithm (ROA). [ABSTRACT FROM AUTHOR]

In real-world recognition/classification tasks, limited by various objective factors, it is usually difficult to collect training samples to exhaust all classes when training a recognizer or classifier. A more realistic scenario is open set recognition (OSR), where incomplete knowledge of the world exists at training time, and unknown classes can be submitted to an algorithm during testing, requiring the classifiers to not only accurately classify the seen classes, but also effectively deal with unseen ones. This paper provides a comprehensive survey of existing open set recognition techniques covering various aspects ranging from related definitions, representations of models, datasets, evaluation criteria, and algorithm comparisons. Furthermore, we briefly analyze the relationships between OSR and its related tasks including zero-shot, one-shot (few-shot) recognition/learning techniques, classification with reject option, and so forth. Additionally, we also review the open world recognition which can be seen as a natural extension of OSR. Importantly, we highlight the limitations of existing approaches and point out some promising subsequent research directions in this field. [ABSTRACT FROM AUTHOR]

Motivation Protein structure alignment is one of the fundamental problems in computational structure biology. A variety of algorithms have been developed to address this important issue in the past decade. However, due to their heuristic nature, current structure alignment methods may suffer from suboptimal alignment and/or over-fragmentation and thus lead to a biologically wrong alignment in some cases. To overcome these limitations, we have developed an accurate topology-independent and global structure alignment method through an FFT-based exhaustive search algorithm, which is referred to as FTAlign. Results Our FTAlign algorithm was extensively tested on six commonly used datasets and compared with seven state-of-the-art structure alignment approaches, TMalign, DeepAlign, Kpax, 3DCOMB, MICAN, SPalignNS and CLICK. It was shown that FTAlign outperformed the other methods in reproducing manually curated alignments and obtained a high success rate of 96.7 and 90.0% on two gold-standard benchmarks, MALIDUP and MALISAM, respectively. Moreover, FTAlign also achieved the overall best performance in terms of biologically meaningful structure overlap (SO) and TMscore on both the sequential alignment test sets including MALIDUP, MALISAM and 64 difficult cases from HOMSTRAD, and the non-sequential sets including MALIDUP-NS, MALISAM-NS, 199 topology-different cases, where FTAlign especially showed more advantage for non-sequential alignment. Despite its global search feature, FTAlign is also computationally efficient and can normally complete a pairwise alignment within one second. Availability and implementation http://huanglab.phys.hust.edu.cn/ftalign/. [ABSTRACT FROM AUTHOR]

Abstract Due to absence of quantitative and evolutionary view on flame-front's wrinkling morphonology, descriptions upon instability onset of premixed expanding spherical flames in previous literatures are implicit and inconsistent. This paper utilized shadowgraphy and high-magnification lens in the optical configuration, which clearly manifested cracks in wrinkled flame images and facilitated precise segmentation of them. A simple algorithm was developed to reconstruct real lengths of cracks from these planar projected images. Furthermore, a meaningful parameter, the averaged cell radius, was deducted from the total crack length. Application of this novel methodology on the near-equidiffusive C 2 H 2 -air flame reveals similar growth speeds of flame cracks under different pressures before the flame front transits into cellular appearance. Before this transition, the averaged cell radius kept fluctuating around a roughly stable level which was intimately dependent on initial disturbances. Once the transition started, the effect of inherent instability dominates over initial disturbances and averaged cell radii of repeated tests gradually converged to a similar value. The above methodology builds up an automatic way to quantitatively characterizing cracks and cells' developing history, which extracts more useful and precise information from image series than merely manual estimation of cell sizes at individual shoots in previous studies. Highlights • Establish a new algorithm to detect wrinkling cracks in spherical flame images. • Two new parameters deduced to indicate length of cracks and size of flame cells. • Quantitative evolving features of crack development and cellular transition. [ABSTRACT FROM AUTHOR]

Motivation: With the discovery of more and more noncoding RNAs and their versatile functions, RNA-RNA interactions have received increased attention. Therefore, determination of their complex structures is valuable to understand the molecular mechanism of the interactions. Given the high cost of experimental methods, computational approaches like molecular docking have played an important role in the determination of complex structures, in which a benchmark is critical for the development of docking algorithms. Results: Meeting the need, we have developed the first comprehensive and nonredundant RNA- RNA docking benchmark (RRDB). The diverse dataset of 123 targets consists of 78 unboundunbound and 45 bound-unbound (or unbound-bound) test cases. The dataset was classified into three groups according to the interface conformational changes between bound and unbound structures: 47 'easy', 38 'medium' and 38 'difficult' targets. A docking test with the benchmark using ZDOCK 2.1 demonstrated the challenging nature of the RNA-RNA docking problem and the important value of the present benchmark. The bound and unbound cases of the benchmark will be beneficial for the development and optimization of docking and scoring algorithms for RNA-RNA interactions. [ABSTRACT FROM AUTHOR]

Motivated by the fact that the signals tend to have a representation biased towards their own classes, we propose a novel Sparse Representation-based Classifier (SRC) named Class Specific Sparse Representation-based Classifier (CSSRC), which incorporates the class information in the representation learning. Unlike the conventional SRC algorithms, CSSRC defines each class as a group and then impels these groups to compete for representing the test sample. To achieve such property, CSSRC imposes a L 1 -norm constraint to the classes for compulsively selecting the most relevant classes and introduces a L 2 -norm constraint to the samples belonging to the same class for making sure that all homogeneous samples can be sufficiently exploited for representation. Since CSSRC is a typical structure sparse representation issue, it can be efficiently solved by the convex optimization. Seven popular visual and audio signal databases are employed for evaluation. The results demonstrate its effectiveness in comparison with the state-of-the-art classifiers. [ABSTRACT FROM AUTHOR]

Linear discriminant analysis (LDA) is a powerful dimensionality reduction technique, which has been widely used in many applications. Although, LDA is well-known for its discriminant capability, it clearly does not capture the geometric structure of the data. However, from the geometric perspective, the high-dimensional data resides on some low-dimensional manifolds in the sample space and these manifold structures are essential for data clustering and classification. In this paper, we propose a novel LDA algorithm named graph regularized linear discriminant analysis (GRLDA) to further improve the conventional LDA by incorporating such geometric information of data. GRLDA is achieved by penalizing the LDA with a Graph regularization, which is an affinity matrix encoding the geometric relationship of the data points. To take high-order geometric relationship among samples into consideration, we generalize GRLDA via using the hypergraph regularization instead of the graph regularization. We name this new version as hyper graph regularized linear discriminant analysis. Moreover, we exploit the null space of LDA via using an identity matrix to regularize the between-class scatter matrix. This strategy can further improve the discriminating power of LDA algorithms. Four popular face databases are used to evaluate our proposed LDA algorithms and the results of experiments demonstrate that they outperform the state-of-the-art dimensionality reduction algorithms. [ABSTRACT FROM AUTHOR]

Abstract: Data mining is to extract large amounts of data in the potential, unknown useful information. Based on decision tree classification algorithm in data mining is widely used, Compared with other classification algorithms, Decision tree algorithm has a computation is relatively small, easy to extract explicit rules, you can display important decision attributes and advantages of the classification accuracy. This article is for specific customer relationship management system, decision tree algorithm for data mining research and analysis, data mining and decision tree techniques are described in detail. [Copyright &y& Elsevier]

Abstract: In this paper, we propose the MIML (Multi-Instance Multi-Label learning) framework where an example is described by multiple instances and associated with multiple class labels. Compared to traditional learning frameworks, the MIML framework is more convenient and natural for representing complicated objects which have multiple semantic meanings. To learn from MIML examples, we propose the MimlBoost and MimlSvm algorithms based on a simple degeneration strategy, and experiments show that solving problems involving complicated objects with multiple semantic meanings in the MIML framework can lead to good performance. Considering that the degeneration process may lose information, we propose the D-MimlSvm algorithm which tackles MIML problems directly in a regularization framework. Moreover, we show that even when we do not have access to the real objects and thus cannot capture more information from real objects by using the MIML representation, MIML is still useful. We propose the InsDif and SubCod algorithms. InsDif works by transforming single-instances into the MIML representation for learning, while SubCod works by transforming single-label examples into the MIML representation for learning. Experiments show that in some tasks they are able to achieve better performance than learning the single-instances or single-label examples directly. [Copyright &y& Elsevier]

• We propose a DSNnet optimization model with non-convex equality constraints for unsupervised representation learning. • We design a neural network based algorithm to solve the proposed DSNnet model. • We introduce a novel similarity measurement between input data and learned features as a regularization term of DSNnet. Deep Semi-NMF (DSN), which learns hierarchical representations by stacking multiple layers Semi-NMF, shows competitive performance in unsupervised data analysis. However, the features learned from DSN always lack of representativity and discriminativity. In this paper, we build a novel Deep Semi-NMF network (DSNnet) to address the issues of DSN. Specifically, DSNnet contains multiple fully-connected layers, in which the activation function of each layer adopts Smoothly Clipped Absolute Deviation (SCAD). The non-negative hidden features are computed forwardly, while the network parameters are updated by the stochastic gradient descent method. Moreover, to enhance the discriminativity of features, we suggest simultaneously minimizing the reconstruction error of input and output, and maximizing the similarity between input and learned features. The proposed similarity measurement, which consists of global geometric similarity and local pointwise similarity, encourages the compactness between similar points and separateness between dissimilar points in the feature space, and is beneficial to preserve intrinsic information of original data. Extensive experiments conducted on several datasets illustrate the superiority of the proposed approach in comparison with state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Abstract: Focused ultrasound thermal therapy relies on temperature monitoring for treatment guidance and assurance of targeting and dose control. One potential approach is to monitor temperature change through ultrasonic-backscattered signal processing. The current approach involves the detection of echo time-shifts based on cross-correlation processing from segmented radiofrequency (RF) data. In this study, we propose a novel ultrasonic temperature-measurement approach that detects changes in instantaneous frequency along the imaging beam direction. Focused ultrasound was used as the heating source, and the 1-D beamformed RF signals provided from an ultrasound imager were used to verify the proposed algorithm for temperature change estimation. For comparison, a conventional cross-correlation technique was also evaluated. Heating experiments testing tissue-mimicking phantoms and ex vivo porcine muscles were conducted. The results showed that temperature can be well estimated by the proposed algorithm in the temperature range, where the relationship of sound speed versus temperature is linear. Compared with the cross-correlation–based algorithm, the proposed new algorithm yields a six-fold increase in computational efficiency, along with comparable contrast-detection ability and precision. This new algorithm may serve as an alternative method for implementing temperature estimation into a clinical ultrasound imager for thermal therapy guidance. (E-mail: kcju@isu.edu.tw) [Copyright &y& Elsevier]