Characterization of sucrose nonfermenting-1-related protein kinase 2 (SnRK2) gene family in Haynaldia villosa demonstrated SnRK2.9-V enhances drought and salt stress tolerance of common wheat.

BACKGROUND: The sucrose nonfermenting-1-related protein kinase 2 (SnRK2) plays a crucial role in responses to diverse biotic/abiotic stresses. Currently, there are reports on these genes in Haynaldia villosa, a diploid wild relative of wheat. RESULTS: To understand the evolution of SnRK2-V family genes and their roles in various stress conditions, we performed genome-wide identification of the SnRK2-V gene family in H. villosa. Ten SnRK2-V genes were identified and characterized for their structures, functions and spatial expressions. Analysis of gene exon/intron structure further revealed the presence of evolutionary paths and replication events of SnRK2-V gene family in the H. villosa. In addition, the features of gene structure, the chromosomal location, subcellular localization of the gene family were investigated and the phylogenetic relationship were determined using computational approaches. Analysis of cis-regulatory elements of SnRK2-V gene members revealed their close correlation with different phytohormone signals. The expression profiling revealed that ten SnRK2-V genes expressed at least one tissue (leave, stem, root, or grain), or in response to at least one of the biotic (stripe rust or powdery mildew) or abiotic (drought or salt) stresses. Moreover, SnRK2.9-V was up-regulated in H. villosa under the drought and salt stress and overexpressing of SnRK2.9-V in wheat enhanced drought and salt tolerances via enhancing the genes expression of antioxidant enzymes, revealing a potential value of SnRK2.9-V in wheat improvement for salt tolerance. CONCLUSION: Our present study provides a basic genome-wide overview of SnRK2-V genes in H. villosa and demonstrates the potential use of SnRK2.9-V in enhancing the drought and salt tolerances in common wheat.

Intratumor injection of BCG Ag85A high-affinity peptides enhanced anti-tumor efficacy in PPD-positive melanoma.

As one of the scheduled immunization vaccines worldwide, virtually all individuals have been vaccinated with BCG vaccine. In order to verify the hypothesis that delivering BCG high-affinity peptides to tumor areas could activate the existing BCG memory T cells to attack tumor, we firstly predicted the HLA-A*0201 high-affinity peptides of BCG Ag85A protein (KLIANNTRV, GLPVEYLQV), and then, A375 melanoma cells and HLA-A*0201 PBMCs (from PPD-positive adults) were added to co-incubated with the predicted peptides in vitro. We found that the predicted BCG high-affinity peptides could be directly loaded onto the surface of tumor cells, enhancing the tumor-killing efficacy of PBMCs from PPD-positive volunteer. Then, we constructed PPD-positive mice model bearing B16F10 subcutaneous tumors and found that intratumor injection of BCG Ag85A high-affinity peptides (SGGANSPAL, YHPQQFVYAGAMSGLLD) enhanced the anti-tumor efficacy in PPD-positive melanoma mice. Along with the better anti-tumor efficacy, the expression of PDL1 on tumor cell surface was also increased, and stronger antitumor effects occurred when further combined with anti-PD1 antibody. For microenvironment analysis, the proportion of effector memory T cells was increased and the better treatment efficacy may be attributed to the elevated effector memory CD4 + T cells within the tumor. In conclusion, using the existing immune response of BCG vaccine by delivering high-affinity peptides of BCG to tumor area is a safe and promising therapy for cancer.

TNFSF14/LIGHT promotes cardiac fibrosis and atrial fibrillation vulnerability via PI3Kγ/SGK1 pathway-dependent M2 macrophage polarisation.

BACKGROUND: Tumour necrosis factor superfamily protein 14 (TNFSF14), also called LIGHT, is an important regulator of immunological and fibrosis diseases. However, its specific involvement in cardiac fibrosis and atrial fibrillation (AF) has not been fully elucidated. The objective of this study is to examine the influence of LIGHT on the development of myocardial fibrosis and AF. METHODS: PCR arrays of peripheral blood mononuclear cells (PBMCs) from patients with AF and sinus rhythm was used to identify the dominant differentially expressed genes, followed by ELISA to evaluate its serum protein levels. Morphological, functional, and electrophysiological changes in the heart were detected in vivo after the tail intravenous injection of recombinant LIGHT (rLIGHT) in mice for 4 weeks. rLIGHT was used to stimulate bone marrow-derived macrophages (BMDMs) to prepare a macrophage-conditioned medium (MCM) in vitro. Then, the MCM was used to culture mouse cardiac fibroblasts (CFs). The expression of relevant proteins and genes was determined using qRT-PCR, western blotting, and immunostaining. RESULTS: The mRNA levels of LIGHT and TNFRSF14 were higher in the PBMCs of patients with AF than in those of the healthy controls. Additionally, the serum protein levels of LIGHT were higher in patients with AF than those in the healthy controls and were correlated with left atrial reverse remodelling. Furthermore, we demonstrated that rLIGHT injection promoted macrophage infiltration and M2 polarisation in the heart, in addition to promoting atrial fibrosis and AF inducibility in vivo, as detected with MASSON staining and atrial burst pacing respectively. RNA sequencing of heart samples revealed that the PI3Kγ/SGK1 pathway may participate in these pathological processes. Therefore, we confirmed the hypothesis that rLIGHT promotes BMDM M2 polarisation and TGB-ß1 secretion, and that this process can be inhibited by PI3Kγ and SGK1 inhibitors in vitro. Meanwhile, increased collagen synthesis and myofibroblast transition were observed in LIGHT-stimulated MCM-cultured CFs and were ameliorated in the groups treated with PI3Kγ and SGK1 inhibitors. CONCLUSION: LIGHT protein levels in peripheral blood can be used as a prognostic marker for AF and to evaluate its severity. LIGHT promotes cardiac fibrosis and AF inducibility by promoting macrophage M2 polarisation, wherein PI3Kγ and SGK1 activation is indispensable.

Enhanced systemic tumor suppression by in situ vaccine combining radiation and OX40 agonist with CpG therapy.

BACKGROUND: In situ tumor vaccine has been gradually becoming a hot research field for its advantage of achieving personalized tumor therapy without prior antigen identification. Various in situ tumor vaccine regimens have been reported to exert considerable antitumor efficacy in preclinical and clinical studies. However, the design of in situ tumor vaccines still needs further optimization and the underlying immune mechanism also waits for deeper investigation. METHODS: A novel triple in situ vaccine strategy that combining local radiation with intratumoral injection of TLR9 agonist CpG and OX40 agonist was established in this sturdy. Local and abscopal antitumor efficacy as well as survival benefit were evaluated in the bilateral tumors and pulmonary metastasis model of B16F10 melanoma. In situ vaccine-induced immune responses and immune-associated variation in tumor environment were further investigated using multiparameter flow cytometry and RNA sequencing. Base on the analysis, the RT + CpG + αOX40 triple in situ vaccine was combined with checkpoint blockade therapy to explore the potential synergistic antitumor efficacy. RESULTS: Enhanced tumor suppression was observed with minimal toxicity in both treated and untreated abscopal tumors after receiving RT + CpG + αOX40 triple vaccine. The introduction of local radiation and OX40 agonist benefit more to the inhibition of local and abscopal lesions respectively, which might be partially attributed to the increase of effector memory T cells in the tumor microenvironment. Further analysis implied that the triple in situ vaccine did not only activate the microenvironment of treated tumors, with the upregulation of multiple immune-associated pathways, but also enhanced systemic antitumor responses, thus achieved superior systemic tumor control and survival benefit. Moreover, the triple in situ vaccine synergized with checkpoint blockade therapy, and significantly improved the therapeutic effect of anti-programmed cell death protein (PD)-1 antibody. CONCLUSION: This triple combining in situ vaccine induced intensive antitumor responses, mediated effective systemic tumor control and survival benefit, and displayed impressive synergistic antitumor effect with checkpoint blockade therapy. These data preliminary confirmed the efficacy, feasibility and safety of the triple combining in situ vaccine, suggesting its great application potential as both monotherapy and a part of combined immunotherapeutic regimens in clinical scenario.

Microwave photonic radar system with improved SNR performance utilizing optical resonant amplification and random Fourier coefficient waveforms.

A microwave photonic (MWP) radar system with improved signal-to-noise ratio (SNR) performance is proposed and experimentally demonstrated. By improving the SNR of echoes through properly designed radar waveforms and resonant amplification in the optical domain, the proposed radar system can detect and image weak targets that were previously hidden in noise. Echoes with a common low-level SNR obtain high optical gain and the in-band noise is suppressed during resonant amplification. The designed radar waveforms, based on random Fourier coefficients, reduce the effect of optical nonlinearity while providing reconfigurable waveform performance parameters for different scenarios. A series of experiments are developed to verify the feasibility of the SNR improvement of the proposed system. Experimental results show a maximum SNR improvement of 3.6 dB with an optical gain of 28.6 dB for the proposed waveforms over a wide input SNR range. From a comparison with linear frequency modulated signals in microwave imaging of rotating targets, significant quality enhancement is observed. The results confirm the ability of the proposed system to improve SNR performance of MWP radars and its great application potential in SNR-sensitive scenarios.

Construction of a knowledge graph for breast cancer diagnosis based on Chinese electronic medical records: development and usability study.

BACKGROUND: Electronic medical records (EMRs) contain a wealth of information related to breast cancer diagnosis and treatment. Extracting relevant features from these medical records and constructing a knowledge graph can significantly contribute to an efficient data analysis and decision support system for breast cancer diagnosis. METHODS: An approach was proposed to develop a workflow for effectively extracting breast cancer-related features from Chinese breast cancer mammography reports and constructing a knowledge graph for breast cancer diagnosis. Firstly, the concept layer of the knowledge graph for breast cancer diagnosis was constructed based on breast cancer diagnosis and treatment guidelines, along with insights from clinical experts. .Next, a BiLSTM-Highway-CRF model was designed to extract the mammography features, which formed the data layer of the knowledge graph. Finally, the knowledge graph was constructed by combining the concept layer and the data layer in a Neo4j graph data platform, and then applied in visualization analysis, semantic query and computer assisted diagnosis. RESULTS: Mammographic features were extracted from a total of 1171 mammography examination reports. The overall extraction performance of the model achieved an accuracy rate of 97.16%, a recall rate of 98.06%, and a F1 score of 97.61%. Additionally, 47,660 relationships between entities were identified based on the four different types of relationships defined in the concept layer. The knowledge graph for breast cancer diagnosis was constructed after inputting mammographic features and relationships into the Neo4j graph data platform. The model was assessed from the concept layer, data layer, and application layer perspectives, and showed promising results. CONCLUSIONS: The proposed workflow is applicable for constructing knowledge graphs for breast cancer diagnosis based on Chinese EMRs. This study serves as a reference for the rapid design, construction, and application of knowledge graphs for diagnosis and treatment of other diseases. Furthermore, it offers a potential solution to address the issues of limited data sharing and format inconsistencies present in Chinese EMR data.

Investigation of signal-to-noise ratio performance of microwave photonic links enhanced by optical injection locking and channelized spectrum stitching.

A signal-to-noise ratio (SNR) improvement method for microwave photonic (MWP) links enhanced by optical injection locking (OIL) and channelized spectrum stitching (CSS) is investigated and experimentally demonstrated. By exploiting the resonant amplification characteristics of OIL, both optical gain and in-band noise suppression of the input radio frequency signal can be achieved. The injection bandwidth is channelized to further suppress noise during OIL, and the input signal can be well reconstructed by spectrum stitching in the digital domain. Experimental results show that the optimal improvement in SNR of 3.6 dB is achieved for linear frequency modulated signals and at least an additional improvement of 7.2 dB can be obtained by adopting CSS. Other broadband signals for radar and communication are used to further verify the ability to improve SNR. The potential for application scenarios with large operating bandwidth and high optical gain is also demonstrated.

Single Target SAR 3D Reconstruction Based on Deep Learning.

Synthetic aperture radar tomography (TomoSAR) is an important 3D mapping method. Traditional TomoSAR requires a large number of observation orbits however, it is hard to meet the requirement of massive orbits. While on the one hand, this is due to funding constraints, on the other hand, because the target scene is changing over time and each observation orbit consumes lots of time, the number of orbits can be fewer as required within a narrow time window. When the number of observation orbits is insufficient, the signal-to-noise ratio (SNR), peak-to-sidelobe ratio (PSR), and resolution of 3D reconstruction results will decline severely, which seriously limits the practical application of TomoSAR. In order to solve this problem, we propose to use a deep learning network to improve the resolution and SNR of 3D reconstruction results under the condition of very few observation orbits by learning the prior distribution of targets. We use all available orbits to reconstruct a high resolution target, while only very few (around 3) orbits to reconstruct a low resolution input. The low-res and high-res 3D voxel-grid pairs are used to train a 3D super-resolution (SR) CNN (convolutional neural network) model, just like ordinary 2D image SR tasks. Experiments on the Civilian Vehicle Radar dataset show that the proposed deep learning algorithm can effectively improve the reconstruction both in quality and in quantity. In addition, the model also shows good generalization performance for targets not shown in the training set.

Estimation and correction of vibration-induced range cell migration for FMCW synthetic aperture ladar.

Synthetic aperture ladar (SAL) is a newly developed imaging device for remote sensing application. Owing to its short wavelength (3-5 orders of magnitude shorter than radar), SAL is very sensitive to platform vibration. For frequency-modulated continuous-wave SAL (FMCW-SAL), the platform vibration induces an additional range cell migration (RCM) to the SAL image. The vibration-induced RCM (VI-RCM) deteriorates the image quality. The VI-RCM is a unique problem for the FMCW-SAL imaging. To address this problem, a raw-data-driven method is proposed to correct the VI-RCM in this paper. First, the signal model was developed to show the VI-RCM in FMCW-SAL echo. Then, based on the model, the differential phase function (DPF) is constructed for the adjacent range profiles. The DPF is a single-frequency signal with its frequency being proportional to the relative range shift between the adjacent range profiles. Based on the DPF, the relative range shift is estimated. After the estimation of all the relative range shifts, the VI-RCM is calculated and corrected. Experiments are performed. The simulated experiment demonstrated the feasibility, accuracy, and efficiency of the proposed method, and the real data processing result verified the effectiveness of the proposed method for FMCW-SAL in practical applications.

A Novel MIMO-SAR System Based on Simultaneous Digital Beam Forming of Both Transceiver and Receiver.

Orthogonal frequency division multiplexing (OFDM) chirp waveform, which is composed of two or more successive identical linear frequency modulated sub pulses, is a newly proposed orthogonal waveform scheme for multi-input multi-output (MIMO) synthetic aperture radar (SAR) systems. However, according to the waveform model, there will be range ambiguity if the mapping width exceeds the maximum unambiguous width determined by the transmitted signal. This greatly limits its application in high-resolution wide-swath (HRWS) remote sensing. The traditional system divides the echoes by digital beam forming (DBF) to solve this problem, but the energy utilization rate is low. A MIMO-SAR system using simultaneous digital beam forming of both transceiver and receiver to avoid range ambiguity is designed in this paper. Compared with traditional system, the novel system designed in this paper obtain higher energy utilization and waveform orthogonality.

Single-frequency LADAR super-resolution Doppler tomography for extended targets.

Doppler tomography is an important means to obtain two-dimensional (2-D) images of remote targets. It is especially suitable for imaging spinning targets such as space debris, warheads, and aircraft blades. However, related research is mostly focused on the microwave band rather than the laser. Higher resolution can be achieved by implementing Doppler tomography in the laser band compared to the existing Doppler tomography in the microwave. Moreover, existing imaging methods are mostly directed at point targets. When these methods deal with extended target echoes, the image quality is unsatisfactory. These problems severely limit the application of Doppler tomography. Here, a novel laser Doppler tomography method has been proposed. The method is based on a single-frequency laser radar (LADAR) that does not require any form of wideband modulation of the transmitted signal. The imaging process is based on the precise relationship between the scattering coefficient of the target and the statistical characteristics of the Doppler spectrum and finds the maximum a posteriori (MAP) estimate of the scattering coefficient distribution. The imaging resolution depends on the Doppler frequency resolution, which exceeds the diffraction limit and is independent of the imaging distance. A laser Doppler tomography experimental system was established. With this system, high-quality laser Doppler tomograms of extended targets were obtained for the first time. In the experiment, the targets have different rotational speeds from 100 to 1000 r/min. The images of these targets with a resolution of 0.4 mm are obtained at a distance of 5 m indoors. In these images, the target details such as textures on the surfaces can be rendered. The quality of these images is greatly improved compared to existing processing methods. The experimental results confirm the effectiveness of the proposed laser Doppler tomography method.

Feature extraction algorithm of a precession target based on a Doppler frequency profile of dual-aspect observation.

In ballistic missile defense, using precession parameters is an effective way to determine warheads from decoys. Due to the geometric theory of diffraction, the backscattered field from a smooth object for a microwave radar consists of contributions from only isolated scattering centers. Further, thanks to the short wavelength of the laser, the ladar could obtain complete Doppler information of illuminated parts of precessing targets. On the basis of modeling the observation geometry of a rotating target with precession, a novel method for extracting precession parameters using coherent ladar based on a Doppler frequency profile (DFP) of dual-view observation is first proposed, to the best of our knowledge. By analyzing the periodic changes of bandwidth of a cone-shaped precessing object, we indicate that the sequence of DFP bandwidth can be approximate to a sinusoid function, based on which the precession period can be obtained through sinusoidal fitting. The precession angle can be obtained by combining dual-aspect receiving observation. A laboratory experiment verified the effectiveness of the proposed algorithm, and the errors of extracted parameters are about 10%.

Analysis and compensation of telescopes' gaps effect on aperture synthesis in a multi-channel synthetic aperture ladar system.

The azimuth multi-channel synthetic aperture ladar (SAL), which arranges multiple telescopes along the flight direction of the platform, transmits signals through a single telescope and receives echoes by multiple telescopes simultaneously to obtain data. The aperture synthesis technology, which has the ability to achieve high resolution through multiple small telescopes, is applied to the multi-channel SAL system to realize the reconstruction of the complete azimuth Doppler spectrum in a short observation time. However, there are gaps inevitably between telescopes, which degrade the results of aperture synthesis. In this work, the effect of gaps on the instantaneous Doppler spectrum of each channel and the influence on the result of the azimuth impulse compression after aperture synthesis are analyzed. In addition, an estimation method of gaps based on the phase errors between channels is proposed to reduce the influence. The estimation accuracy of the proposed method is analyzed, and the effectiveness of the method is verified with simulations. The estimated gaps are used to compensate for the phase discontinuity of the azimuth signal after aperture synthesis caused by gaps. The method improves the result of aperture synthesis and reduces the side-lobe of the azimuth impulse compression after aperture synthesis.

Accurate Wide Angle SAR Imaging Based on LS-CS-Residual.

Wide angle synthetic aperture radar (WASAR) receives data from a large angle, which causes the problem of aspect dependent scattering. L 1 regularization is a common compressed sensing (CS) model. The L 1 regularization based WASAR imaging method divides the whole aperture into subapertures and reconstructs the subaperture images individually. However, the aspect dependent scattering recovery of it is not accurate. The subaperture images of WASAR can be regarded as the SAR video. The support set among the different frames of SAR video are highly overlapped. Least squares on compressed sensing residuals (LS-CS-Residuals) can reconstruct the time sequences of sparse signals which change slowly with time. This is to replace CS on the observation by CS on the least squares (LS) residual computed using the prior estimate of the support. In this paper, we introduce LS-CS-Residual into WASAR imaging. In the iteration of LS-CS-Residual, the azimuth-range decoupled operators are used to avoid the huge memory cost. Real data processing results show that LS-CS-Residual can estimate the aspect dependent scatterings of the targets more accurately than CS based methods.

Efficient Parameter Estimation for Sparse SAR Imaging Based on Complex Image and Azimuth-Range Decouple.

Sparse signal processing theory has been applied to synthetic aperture radar (SAR) imaging. In compressive sensing (CS), the sparsity is usually considered as a known parameter. However, it is unknown practically. For many functions of CS, we need to know this parameter. Therefore, the estimation of sparsity is crucial for sparse SAR imaging. The sparsity is determined by the size of regularization parameter. Several methods have been presented for automatically estimating the regularization parameter, and have been applied to sparse SAR imaging. However, these methods are deduced based on an observation matrix, which will entail huge computational and memory costs. In this paper, to enhance the computational efficiency, an efficient adaptive parameter estimation method for sparse SAR imaging is proposed. The complex image-based sparse SAR imaging method only considers the threshold operation of the complex image, which can reduce the computational costs significantly. By utilizing this feature, the parameter is pre-estimated based on a complex image. In order to estimate the sparsity accurately, adaptive parameter estimation is then processed in the raw data domain, combining with the pre-estimated parameter and azimuth-range decouple operators. The proposed method can reduce the computational complexity from a quadratic square order to a linear logarithm order, which can be used in the large-scale scene. Simulated and Gaofen-3 SAR data processing results demonstrate the validity of the proposed method.

Three-dimensional inverse synthetic aperture lidar imaging for long-range spinning targets.

We present a three-dimensional (3D) imaging method for long-range spinning targets. This method acquires multi-angle two-dimensional (2D) images of spinning targets by the inverse synthetic aperture lidar (ISAL) imaging technique. The 3D distribution of the scattering coefficients of a target has a mapping relationship with the series of 2D images. This mapping is analyzed, and a 3D Hough transform is used to implement inverse mapping. The parameter space of the Hough transform is the estimation of the 3D distribution of the scattering coefficients. The 3D point spread function obtained by the method has narrow main lobe widths and sufficiently low side lobes to achieve high image quality, which is verified by computer simulations. In the simulations, the main lobe widths in the three dimensions are 0.29 cm, 0.29 cm, and 3.48 cm, respectively. In outdoor experiments, 3D images of targets at 1 km away from the lidar were obtained. The images clearly show the 3D shape of targets.

Inverse synthetic aperture LADAR demonstration: system structure, imaging processing, and experiment result.

A long-distance inverse synthetic aperture LADAR (ISAL) imaging experiment outdoors over 1 km for cooperative targets is demonstrated, which gets a two-dimensional high-resolution image with resolution exceeding 2.5 cm. The system utilizes an electro-optic in-phase and quadrature modulator to output a linear frequency-modulated continuous waveform (LFMCW) with a bandwidth of 6 GHz and pulse repetition frequency (PRF) of 16.7 KHz. For the problem of the coherence of the laser, the effects of the coherent processing interval (CPI) and time delay of the local oscillator (LO) on the coherence are discussed. The fiber delay line is set and the CPI is reduced to lower the requirement of the frequency stability of the laser source. The images are formed by two-dimensional Fourier transform and joint time-frequency transform methods, respectively. In this paper, we present the system structure, imaging processing, and the experiment result in detail. The experiment result validates the performance of our system for ISAL imaging.

Linear frequency-modulated continuous-wave ladar system for synthetic aperture imaging.

A novel and high-efficiency linear frequency-modulated continuous-wave (FMCW) ladar system for synthetic aperture imaging is proposed and experimentally demonstrated. This novel system generates wide-bandwidth linear FMCW ladar signals by employing an electro-optic LiNbO3- in-phase and quadrature modulator with an effective bias controller. The effectiveness of the proposed system is experimentally validated. Optical synthetic aperture images are obtained by using two 0.41 cm aperture diameter telescopes at the distance of 1 km. The resolution of these images can reach to 4 cm. A resolution improvement by about 10 times is achieved when compared with the conventional real aperture imaging system.

A Robust Range Grating Lobe Suppression Method Based on Image Contrast for Stepped-Frequency SAR.

The magnitude error and phase error (MEPE) in the transfer function of a stepped-frequency synthetic aperture radar (SAR) system results in a periodic MEPE in the synthesized wideband waveform (SWW), which induces the grating lobes in the high-resolution range profile (HRRP). In this paper, a robust data-driven grating lobe suppression (GLS) method is proposed. Based on a contrast-based error estimation method and the grating lobes of the brightest scatterers in the SAR image, the periodic MEPE can be robustly estimated using the proposed method. By compensating the estimated periodic MEPE, the range grating lobes can be suppressed to the background level of the SAR image. Simulation results and real data processing have demonstrated the superiority of the proposed method.

Development of an online, publicly accessible naive Bayesian decision support tool for mammographic mass lesions based on the American College of Radiology (ACR) BI-RADS lexicon.

PURPOSE: To develop and validate a decision support tool for mammographic mass lesions based on a standardized descriptor terminology (BI-RADS lexicon) to reduce variability of practice. MATERIALS AND METHODS: We used separate training data (1,276 lesions, 138 malignant) and validation data (1,177 lesions, 175 malignant). We created naïve Bayes (NB) classifiers from the training data with tenfold cross-validation. Our "inclusive model" comprised BI-RADS categories, BI-RADS descriptors, and age as predictive variables; our "descriptor model" comprised BI-RADS descriptors and age. The resulting NB classifiers were applied to the validation data. We evaluated and compared classifier performance with ROC-analysis. RESULTS: In the training data, the inclusive model yields an AUC of 0.959; the descriptor model yields an AUC of 0.910 (P < 0.001). The inclusive model is superior to the clinical performance (BI-RADS categories alone, P < 0.001); the descriptor model performs similarly. When applied to the validation data, the inclusive model yields an AUC of 0.935; the descriptor model yields an AUC of 0.876 (P < 0.001). Again, the inclusive model is superior to the clinical performance (P < 0.001); the descriptor model performs similarly. CONCLUSION: We consider our classifier a step towards a more uniform interpretation of combinations of BI-RADS descriptors. We provide our classifier at www.ebm-radiology.com/nbmm/index.html . KEY POINTS: • We provide a decision support tool for mammographic masses at www.ebm-radiology.com/nbmm/index.html . • Our tool may reduce variability of practice in BI-RADS category assignment. • A formal analysis of BI-RADS descriptors may enhance radiologists' diagnostic performance.