A Two-Stage Generative Model with CycleGAN and Joint Diffusion for MRI-based Brain Tumor Detection.

Accurate detection and segmentation of brain tumors is critical for medical diagnosis. However, current supervised learning methods require extensively annotated images and the state-of-the-art generative models used in unsupervised methods often have limitations in covering the whole data distribution. In this paper, we propose a novel framework Two-Stage Generative Model (TSGM) that combines Cycle Generative Adversarial Network (CycleGAN) and Variance Exploding stochastic differential equation using joint probability (VE-JP) to improve brain tumor detection and segmentation. The CycleGAN is trained on unpaired data to generate abnormal images from healthy images as data prior. Then VE-JP is implemented to reconstruct healthy images using synthetic paired abnormal images as a guide, which alters only pathological regions but not regions of healthy. Notably, our method directly learned the joint probability distribution for conditional generation. The residual between input and reconstructed images suggests the abnormalities and a thresholding method is subsequently applied to obtain segmentation results. Furthermore, the multimodal results are weighted with different weights to improve the segmentation accuracy further. We validated our method on three datasets, and compared with other unsupervised methods for anomaly detection and segmentation. The DSC score of 0.8590 in BraTs2020 dataset, 0.6226 in ITCS dataset and 0.7403 in In-house dataset show that our method achieves better segmentation performance and has better generalization.

The CT and MRI features of benign calvarium and skull base osteoblastoma.

OBJECTIVE: We retrospectively reviewed the CT and MRI features of patients with benign osteoblastoma in the calvarium and skull base (CSBOB). METHODS: Nine cases of pathologically confirmed benign CSBOB were analysed retrospectively. The patients had undergone CT and/or MRI. Tumour location, size, and imaging features were reviewed and recorded. RESULTS: The patients included four males and five females with a mean age of 27.0 years (age 14-40 years). The tumours were located in the frontal bone in 3 patients, the occipital bone in 3 patients, and in the parietal bone, sphenoid bone, and skull base in 1 patient each. On CT, the tumours measured 5.1 ± 3.3 (1.8-8.4) cm. Seven tumours were shown to have caused expansile bony destruction with an eggshell appearance and varying degrees of calcification or matrix mineralization. Multiple septa were observed in 5 tumours. Intracranial growth was observed in 5 tumours. On MRI, 7 tumours showed heterogeneous hypo- to isointensity on T1WI. Heterogeneous high signal patterns with low signal rims and septa were observed in 6 tumours on T2WI, and 4 showed a fluid-fluid level. On contrast-enhanced imaging, 6 tumours showed peripheral and septal enhancement, and 2 showed the dural tail sign. CONCLUSIONS: Benign CSBOB is a rare tumour characterized by expansile bony destruction, septa, a sclerotic rim and calcification or matrix mineralization on CT and MRI. ADVANCES IN KNOWLEDGE: The findings from this study contribute to a better understanding of benign CSBOB and provide valuable imaging features that can aid in its diagnosis and differentiation from other tumours in the calvarium and skull base.

MnO2/Ce6 microbubble-mediated hypoxia modulation for enhancing sono-photodynamic therapy against triple negative breast cancer.

Sono-photodynamic therapy (SPDT) has emerged as a promising treatment modality for triple negative breast cancer (TNBC). However, the hypoxic tumor microenvironment hinders the application of SPDT. Herein, in this study, a multifunctional platform (MnO2/Ce6@MBs) was designed to address this issue. A sono-photosensitizer (Ce6) and a hypoxia modulator (MnO2) were loaded into microbubbles and precisely released within tumor tissues under ultrasound irradiation. MnO2in situ reacted with the excess H2O2 and H+ and produced O2 within the TNBC tumor, which alleviated hypoxia and augmented SPDT by increasing ROS generation. Meanwhile, the reaction product Mn2+ was able to achieve T1-weighted MRI for enhanced tumor imaging. Additionally, Ce6 and microbubbles served as a fluorescence imaging contrast agent and a contrast-enhanced ultrasound imaging agent, respectively. In in vivo anti-tumor studies, under the FL/US/MR imaging guidance, MnO2/Ce6@MBs combined with SPDT significantly reversed tumor hypoxia and inhibited tumor growth in 4T1-tumor bearing mice. This work presents a theragnostic system for reversing tumor hypoxia and enhancing TNBC treatment.

Correction: MnO2/Ce6 microbubble-mediated hypoxia modulation for enhancing sono-photodynamic therapy against triple negative breast cancer.

Correction for 'MnO2/Ce6 microbubble-mediated hypoxia modulation for enhancing sono-photodynamic therapy against triple negative breast cancer' by Ping Li et al., Biomater. Sci., 2024, https://doi.org/10.1039/d3bm00931a.

A novel automatic segmentation method directly based on magnetic resonance imaging K-space data for auxiliary diagnosis of glioma.

Background: The use of segmentation architectures in medical imaging, particularly for glioma diagnosis, marks a significant advancement in the field. Traditional methods often rely on post-processed images; however, key details can be lost during the fast Fourier transformation (FFT) process. Given the limitations of these techniques, there is a growing interest in exploring more direct approaches. The adaption of segmentation architectures originally designed for road extraction for medical imaging represents an innovative step in this direction. By employing K-space data as the modal input, this method completely eliminates the information loss inherent in FFT, thereby potentially enhancing the precision and effectiveness of glioma diagnosis. Methods: In the study, a novel architecture based on a deep-residual U-net was developed to accomplish the challenging task of automatically segmenting brain tumors from K-space data. Brain tumors from K-space data with different under-sampling rates were also segmented to verify the clinical application of our method. Results: Compared to the benchmarks set in the 2018 Brain Tumor Segmentation (BraTS) Challenge, our proposed architecture had superior performance, achieving Dice scores of 0.8573, 0.8789, and 0.7765 for the whole tumor (WT), tumor core (TC), and enhanced tumor (ET) regions, respectively. The corresponding Hausdorff distances were 2.5649, 1.6146, and 2.7187 for the WT, TC, and ET regions, respectively. Notably, compared to traditional image-based approaches, the architecture also exhibited an improvement of approximately 10% in segmentation accuracy on the K-space data at different under-sampling rates. Conclusions: These results show the superiority of our method compared to previous methods. The direct performance of lesion segmentation based on K-space data eliminates the time-consuming and tedious image reconstruction process, thus enabling the segmentation task to be accomplished more efficiently.

Demonstration of magnetic resonance Z-spectral imaging for fatty acid characterization of bone marrow at 3 T.

Magnetic resonance Z-spectral imaging (ZSI) has emerged as a new approach to measure fat fraction (FF). However, its feasibility for fat spectral imaging remains to be elucidated. In this study, a single-slice ZSI sequence dedicated to fat spectral imaging was designed, and its capability for fatty acid characterization was investigated on peanut oil samples, a multiple-vial fat-water phantom with varied oil volumes, and vertebral body marrow in healthy volunteers and osteoporosis patients at 3 T. The peanut oil spectrum was also recorded with a 400-MHz NMR spectrometer. A Gaussian-Lorentzian sum model was used to resolve water and six fat signals of the pure oil sample or four fat signals of the fat-water phantom or vertebral bone marrow from Z spectra. Fat peak amplitudes were normalized to the total peak amplitude of water and all fat signals. Normalized fat peak amplitudes and FF were quantified and compared among vials of the fat-water phantom or between healthy volunteers and osteoporosis patients. An unpaired student's t-test and Pearson's correlation were conducted, with p less than 0.05 considered statistically significant. The results showed that the peanut oil spectra measured with the ZSI technique were in line with respective NMR spectra, with amplitudes of the six fat signal peaks significantly correlated between the two methods (y = x + 0.001, r = 0.996, p < 0.001 under a repetition time of 1.6 s; and y = 1.026x - 0.003, r = 0.996, p < 0.001 under a repetition time of 3.1 s). Moreover, ZSI-measured FF exhibited a significant correlation with prepared oil volumes (y = 0.876x + 1.290, r = 0.996, p < 0.001). The osteoporosis patients showed significantly higher normalized fat peak amplitudes and FF in the L4 vertebral body marrow than the healthy volunteers (all p < 0.01). In summary, the designed ZSI sequence is feasible for fatty acid characterization, and has the potential to facilitate the diagnosis and evaluation of diseases associated with fat alterations at 3 T.

Synthetic high-energy computed tomography image via a Wasserstein generative adversarial network with the convolutional block attention module.

Background: Computed tomography (CT) is now universally applied into clinical practice with its non-invasive quality and reliability for lesion detection, which highly improves the diagnostic accuracy of patients with systemic diseases. Although low-dose CT reduces X-ray radiation dose and harm to the human body, it inevitably produces noise and artifacts that are detrimental to information acquisition and medical diagnosis for CT images. Methods: This paper proposes a Wasserstein generative adversarial network (WGAN) with a convolutional block attention module (CBAM) to realize a method of directly synthesizing high-energy CT (HECT) images through low-energy scanning, which greatly reduces X-ray radiation from high-energy scanning. Specifically, our proposed generator structure in WGAN consists of Visual Geometry Group Network (Vgg16), 9 residual blocks, upsampling and CBAM, a subsequent attention block. The convolutional block attention module is integrated into the generator for improving the denoising ability of the network as verified by our ablation comparison experiments. Results: Experimental results of the generator attention module ablation comparison indicate an optimization boost to the overall generator model, obtaining the synthesized high-energy CT with the best metric and denoising effect. In different methods comparison experiments, it can be clearly observed that our proposed method is superior in the peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM) and most of the statistics (average CT value and its standard deviation) compared to other methods. Because P<0.05, the samples are significantly different. The data distribution at the pixel level between the images synthesized by the method in this paper and the high-energy CT images is also most similar. Conclusions: Experimental results indicate that CBAM is able to suppress the noise and artifacts effectively and suggest that the image synthesized by the proposed method is closest to the high-energy CT image in terms of visual perception and objective evaluation metrics.

HLA-DQA1 expression is associated with prognosis and predictable with radiomics in breast cancer.

BACKGROUND: High HLA-DQA1 expression is associated with a better prognosis in many cancers. However, the association between HLA-DQA1 expression and prognosis of breast cancer and the noninvasive assessment of HLA-DQA1 expression are still unclear. This study aimed to reveal the association and investigate the potential of radiomics to predict HLA-DQA1 expression in breast cancer. METHODS: In this retrospective study, transcriptome sequencing data, medical imaging data, clinical and follow-up data were downloaded from the TCIA ( https://www.cancerimagingarchive.net/ ) and TCGA ( https://portal.gdc.cancer.gov/ ) databases. The clinical characteristic differences between the high HLA-DQA1 expression group (HHD group) and the low HLA-DQA1 expression group were explored. Gene set enrichment analysis, Kaplan‒Meier survival analysis and Cox regression were performed. Then, 107 dynamic contrast-enhanced magnetic resonance imaging features were extracted, including size, shape and texture. Using recursive feature elimination and gradient boosting machine, a radiomics model was established to predict HLA-DQA1 expression. Receiver operating characteristic (ROC) curves, precision-recall curves, calibration curves, and decision curves were used for model evaluation. RESULTS: The HHD group had better survival outcomes. The differentially expressed genes in the HHD group were significantly enriched in oxidative phosphorylation (OXPHOS) and estrogen response early and late signalling pathways. The radiomic score (RS) output from the model was associated with HLA-DQA1 expression. The area under the ROC curves (95% CI), accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the radiomic model were 0.866 (0.775-0.956), 0.825, 0.939, 0.7, 0.775, and 0.913 in the training set and 0.780 (0.629-0.931), 0.659, 0.81, 0.5, 0.63, and 0.714 in the validation set, respectively, showing a good prediction effect. CONCLUSIONS: High HLA-DQA1 expression is associated with a better prognosis in breast cancer. Quantitative radiomics as a noninvasive imaging biomarker has potential value for predicting HLA-DQA1 expression.

Classification of asymmetry in mammography via the DenseNet convolutional neural network.

Purpose: To investigate the effectiveness of a deep learning system based on the DenseNet convolutional neural network in diagnosing benign and malignant asymmetric lesions in mammography. Methods: Clinical and image data from 460 women aged 23-82 years (47.57 ± 8.73 years) with asymmetric lesions who underwent mammography at Shenzhen People's Hospital, Shenzhen Luohu District People's Hospital, and Shenzhen Hospital of Peking University from December 2019 to December 2020 were retrospectively analyzed. Two senior radiologists, two junior radiologists, and the DL system read the mammographic images of 460 patients, respectively, and finally recorded the BI-RADS classification of asymmetric lesions. We then used the area under the curve (AUC) of the receiver operating characteristic (ROC) to evaluate the diagnostic efficacy and the difference between AUCs by the Delong method. Results: Specificity (0.909 vs. 0.835, 0.790, χ2=8.21 and 17.22, p＜0.05) and precision (0.872 vs. 0.763, 0.726, χ2=9.23 and 5.22, p＜0.05) of the DL system in the diagnosis of benign and malignant asymmetric lesions were higher than those of junior radiologist A and B, and there was a statistically significant difference between AUCs (0.778 vs. 0.579, 0.564, Z = 4.033 and 4.460, p＜0.05). Furthermore, the AUC (0.778 vs. 0.904, 0.862, Z = 3.191, and 2.167, p＜0.05) of benign and malignant asymmetric lesions diagnosed by the DL system was lower than that of senior radiologist A and senior radiologist B. Conclusions: The DL system based on the DenseNet convolution neural network has high diagnostic efficiency, which can help junior radiologists evaluate benign and malignant asymmetric lesions more accurately. It can also improve diagnostic accuracy and reduce missed diagnoses caused by inexperienced junior radiologists.

Deep learning for fully automated segmentation and volumetry of Couinaud liver segments and future liver remnants shown with CT before major hepatectomy: a validation study of a predictive model.

Background: Recent reports have shown the potential for deep learning (DL) models to automatically segment of Couinaud liver segments and future liver remnant (FLR) for liver resections. However, these studies have mainly focused on the development of the models. Existing reports lack adequate validation of these models in diverse liver conditions and thorough evaluation using clinical cases. This study thus aimed to develop and perform a spatial external validation of a DL model for the automated segmentation of Couinaud liver segments and FLR using computed tomography (CT) in various liver conditions and to apply the model prior to major hepatectomy. Methods: This retrospective study developed a 3-dimensional (3D) U-Net model for the automated segmentation of Couinaud liver segments and FLR on contrast-enhanced portovenous phase (PVP) CT scans. Images were obtained from 170 patients from January 2018 to March 2019. First, radiologists annotated the Couinaud segmentations. Then, a 3D U-Net model was trained in Peking University First Hospital (n=170) and tested in Peking University Shenzhen Hospital (n=178) in cases with various liver conditions (n=146) and in candidates for major hepatectomy (n=32). The segmentation accuracy was evaluated using the dice similarity coefficient (DSC). Quantitative volumetry to evaluate the resectability was compared between manual and automated segmentation. Results: The DSC in the test data sets 1 and 2 for segments I to VIII was 0.93±0.01, 0.94±0.01, 0.93±0.01, 0.93±0.01, 0.94±0.00, 0.95±0.00, 0.95±0.00, and 0.95±0.00, respectively. The mean automated FLR and FLR% assessments were 493.51±284.77 mL and 38.53%±19.38%, respectively. The mean manual FLR and FLR% assessments were 500.92±284.38 mL and 38.35%±19.14%, respectively, in test data sets 1 and 2. For test data set 1, when automated segmentation of the FLR% was used, 106, 23, 146, and 57 cases were categorized as candidates for a virtual major hepatectomy of types 1, 2, 3, and 4, respectively; however, when manual segmentation of the FLR% was used, 107, 23, 146, and 57 cases were categorized as candidates for a virtual major hepatectomy of types 1, 2, 3, and 4, respectively. For test data set 2, all cases were categorized as candidates for major hepatectomy when automated and manual segmentation of the FLR% was used. No significant differences in FLR assessment (P=0.50; U=185,545), FLR% assessment (P=0.82; U=188,337), or the indications for major hepatectomy were noted between automated and manual segmentation (McNemar test statistic 0.00; P>0.99). Conclusions: The DL model could be used to fully automate the segmentation of Couinaud liver segments and FLR with CT prior to major hepatectomy in an accurate and clinically practicable manner.

Improved breast lesion detection in mammogram images using a deep neural network.

PURPOSE: This study aimed to investigate the effect of using a deep neural network (DNN) in breast cancer (BC) detection. METHODS: In this retrospective study, a DNN-based model was constructed from a total of 880 mammograms that 220 patients underwent between April and June 2020. The mammograms were reviewed by two senior and two junior radiologists with and without the aid of the DNN model. The performance of the network was assessed by comparing the area under the curve (AUC) and receiver operating characteristic curves for the detection of four features of malignancy (masses, calcifications, asymmetries, and architectural distortions), with and without the aid of the DNN model and by the senior and junior radiologists. Additionally, the effect of utilizing the DNN on diagnosis time for both the senior and junior radiologists was evaluated. RESULTS: The AUCs of the model for the detection of mass and calcification were 0.877 and 0.937, respectively. In the senior radiologist group, the AUC values for evaluation of mass, calcification, and asymmetric compaction were significantly higher with the DNN model than those obtained without the model. Similar effects were observed in the junior radiologist group, but the increase in the AUC values was even more dramatic. The median mammogram assessment time of the junior and senior radiologists was 572 (357-951) s, and 273.5 (129-469) s, respectively, with the DNN model, and the corresponding assessment time without the model, was 739 (445-1003) s and 321 (195-491) s, respectively. CONCLUSION: The DNN model exhibited high accuracy in detecting the four named features of BC and effectively shortened the review time by both senior and junior radiologists.

Intermediate-weighted MRI with fat suppression (IW-FS): diagnostic performance for bone marrow edema and erosion detection in axial spondyloarthritis.

BACKGROUND: Bone marrow edema (BME) and erosion of the sacroiliac joint are both key lesions for diagnosing axial spondyloarthritis (axSpA) on magnetic resonance imaging (MRI). PURPOSE: To qualitatively and quantitatively compare intermediate-weighted MRI with fat suppression (IW-FS) with T2-weighted short tau inversion recovery (T2-STIR) in assessment of sacroiliac BME and erosion in axSpA. MATERIAL AND METHODS: Patients aged 18-60 years with axSpA were prospectively enrolled. All patients underwent a 3.0-T MRI examination of the sacroiliac joints. Para-coronal IW-FS, T2-STIR, and T1-weighted (T1W) images were acquired. BME and erosion were scored by two readers in consensus on IW-FS and STIR using a modified Spondyloarthritis Research Consortium of Canada (SPARCC) scoring system. Consensus scores on T1WI were used as the reference for erosion. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were measured for BME. RESULTS: In total, 49 patients (mean age=33.4 ± 7.6 years) were included. More patients were scored as having BME on T2-STIR (36 vs. 29, P = 0.016). SPARCC-BME score on IW-FS was lower than that acquired on T2-STIR (mean, 11.5 vs. 14.7, P = 0.002). SNR and CNR of BME were both lower on IW-FS than on T2-STIR (mean SNR, 118 vs. 218, P < 0.001; mean CNR, 44 vs. 137, P < 0.001). The sensitivity of erosion detection was higher on IW-FS (83%) than on T2-STIR (54%, P = 0.006). CONCLUSION: IW-FS is not sufficient for BME detection using T2-STIR as the reference standard in patients with axSpA. IW-FS has a much higher sensitivity than T2-STIR for erosion detection in the sacroiliac joint.

Evaluation of atrial anatomical remodeling in atrial fibrillation with machine-learned morphological features.

PURPOSE: To elucidate the role of atrial anatomical remodeling in atrial fibrillation (AF), we proposed an automatic method to extract and analyze morphological characteristics in left atrium (LA), left atrial appendage (LAA) and pulmonary veins (PVs) and constructed classifiers to evaluate the importance of identified features. METHODS: The LA, LAA and PVs were segmented from contrast computed tomography images using either a commercial software or a self-adaptive algorithm proposed by us. From these segments, geometric and fractal features were calculated automatically. To reduce the model complexity, a feature selection procedure is adopted, with the important features identified via univariable analysis and ensemble feature selection. The effectiveness of this approach is well illustrated by the high accuracy of our models. RESULTS: Morphological features, such as LAA ostium dimensions and LA volume and surface area, statistically distinguished ([Formula: see text]) AF patients or AF with LAA filling defects (AF(def+)) patients among all patients. On the test set, the best model to predict AF among all patients had an area under the receiver operating characteristic curve (AUC) of 0.91 (95% CI, 0.8-1) and the best model to predict AF(def+) among all patients had an AUC of 0.92 (95% CI, 0.81-1). CONCLUSION: This study automatically extracted and analyzed atrial morphology in AF and identified atrial anatomical remodeling that statistically distinguished AF or AF(def+). The importance of identified atrial morphological features in characterizing AF or AF(def+) was validated by corresponding classifiers. This work provides a good foundation for a complete computer-assisted diagnostic workflow of predicting the occurrence of AF or AF(def+).

Proton density fat fraction (PDFF) maps aid fat metaplasia evaluation in the sacroiliac joints in additional to T1WI: Improved diagnostic accuracy in axial spondyloarthritis.

PURPOSE: To evaluate the added value of qualitative and quantitative fat metaplasia analysis using proton-density fat fraction (PDFF) map in additional to T1-weighted imaging (T1WI) of the sacroiliac joints (SIJ) for diagnosis of axial spondyloarthritis (axSpA). METHOD: Patients aged 18-45 years with axSpA were enrolled. Non-SpA patients and healthy volunteers were included as controls. All participants underwent 3.0T MRI of the SIJs including semi-coronal T1WI and semi-coronal chemical-shift encoded MRI sequence for generating PDFF map. Each joint was divided into four quadrants for analysis. Two independent readers scored fat metaplasia on T1WI alone or with additional PDFF map and measured PDFF values in different reading sessions. Using clinical diagnosis as the reference, diagnostic accuracy of visual scores and PDFF measurements was evaluated by area under the receiver operating characteristic curve (AUC). Inter-reader agreement was evaluated by the intra-class correlation coefficient (ICC). RESULTS: Forty-nine patients with axSpA and thirty-six controls were included. Qualitative fat metaplasia scores using additional PDFF map performed better than using T1WI alone (AUC: Reader 1, 0.847 vs 0.795, p = 0.082; Reader 2, 0.785 vs 0.719, p = 0.048). AUCs of quantitative analysis using number of quadrants with PDFF value ≥75 % were higher than qualitative analysis using T1WI alone (Reader 1, 0.863 vs 0.795, p = 0.046; Reader 2, 0.823 vs 0.785, p = 0.011). ICCs were 0.854 to 0.922 for qualitative analysis and 0.935 for quantitative analysis. CONCLUSIONS: Additional PDFF map can increase the diagnostic accuracy for axSpA by qualitative and quantitative fat metaplasia analysis, in comparison to using T1WI alone.

Combining Magnetic Resonance Diffusion-Weighted Imaging with Prostate-Specific Antigen to Differentiate Between Malignant and Benign Prostate Lesions.

BACKGROUND We aimed to develop a combined model of quantitative parameters derived from 3 different magnetic resonance imaging (MRI) diffusion models and laboratory data related to prostate-specific antigen (PSA) for differentiating between prostate cancer (PCa) and benign lesions. MATERIAL AND METHODS Eighty-four patients pathologically confirmed as having PCa or benign disease were enrolled. All patients underwent multiparametric MRI before biopsy, added intravoxel incoherent motion (IVIM) imaging, and diffusion kurtosis imaging (DKI). The following data were collected: quantitative parameters of diffusion-weighted imaging (DWI), IVIM, and DKI, preoperative total PSA, free/total PSA ratio, and PSA density (PSAD) values. A combined logistic regression model was established by above MRI quantitative parameters and PSA data to diagnose PCa. The Prostate Imaging Reporting and Data System version 2 (PI-RADS v2) was used to assess the lesions for comparison. RESULTS Thirty-two patients had PCa and 52 patients had benign lesions. In multivariate logistic regression analysis, only apparent diffusion coefficient (ADC) and PSAD were significant variables (P<0.05) and were thus retained in the model. The area under curve value of the combined model (0.911) was higher than that of ADC, PSAD, and PI-RADS v2 (0.887, 0.861, and 0.859, respectively) in univariate analysis, but without any statistically significant differences. The combined model generated greater clinical benefit than the independent application of ADC, PSAD, and PI-RADS v2. CONCLUSIONS ADC and PSAD were the 2 most important metrics for distinguishing PCa from benign lesions. The combined model of ADC and PSAD demonstrated satisfactory discrimination and improved clinical net benefit.

Deep learning-based pulmonary tuberculosis automated detection on chest radiography: large-scale independent testing.

Background: It is critical to have a deep learning-based system validated on an external dataset before it is used to assist clinical prognoses. The aim of this study was to assess the performance of an artificial intelligence (AI) system to detect tuberculosis (TB) in a large-scale external dataset. Methods: An artificial, deep convolutional neural network (DCNN) was developed to differentiate TB from other common abnormalities of the lung on large-scale chest X-ray radiographs. An internal dataset with 7,025 images was used to develop the AI system, including images were from five sources in the U.S. and China, after which a 6-year dynamic cohort accumulation dataset with 358,169 images was used to conduct an independent external validation of the trained AI system. Results: The developed AI system provided a delineation of the boundaries of the lung region with a Dice coefficient of 0.958. It achieved an AUC of 0.99 and an accuracy of 0.948 on the internal data set, and an AUC of 0.95 and an accuracy of 0.931 on the external data set when it was used to detect TB from normal images. The AI system achieved an AUC of more than 0.9 on the internal data set, and an AUC of over 0.8 on the external data set when it was applied to detect TB, non-TB abnormal and normal images. Conclusions: We conducted a real-world independent validation, which showed that the trained system can be used as a TB screening tool to flag possible cases for rapid radiologic review and guide further examinations for radiologists.

Intratumoral and peritumoral radiomics for preoperative prediction of neoadjuvant chemotherapy effect in breast cancer based on contrast-enhanced spectral mammography.

OBJECTIVE: To investigative the performance of intratumoral and peritumoral radiomics based on contrast-enhanced spectral mammography (CESM) to preoperatively predict the effect of the neoadjuvant chemotherapy (NAC) of breast cancers. MATERIALS AND METHODS: A total of 118 patients with breast cancer who underwent preoperative CESM and NAC from July 2017 to June 2020 were retrospectively analyzed, and the patients were grouped into training (n = 81) and test sets (n = 37) according to the CESM examination time. NAC effect for each patient was assessed by pathology. Intratumoral and peritumoral radiomics features were extracted from CESM images, and feature selection was performed through the Mann-Whitney U test and least absolute shrinkage and selection operator regression (LASSO). Five radiomics signatures based on intratumoral regions, 5-mm peritumoral regions, 10-mm peritumoral regions, intratumoral regions + 5-mm peritumoral regions, and intratumoral regions + 10-mm peritumoral regions were calculated through a linear combination of selected features weighted by their respective coefficients. The prediction performance of radiomics signatures was assessed by the area under the receiver operator characteristic (ROC) curve, the precision-recall (P-R) curve, the calibration curve, and decision curve analysis (DCA). RESULTS: Ten radiomics features were selected to establish the radiomics signature of intratumoral regions + 5-mm peritumoral regions, which yielded a maximum AUC of 0.85 (95% CI, 0.72-0.98) in the test set. The calibration curves, P-R curves, and DCA showed favorable predictive performance of the five radiomics signatures. CONCLUSION: The intratumoral and peritumoral radiomics based on CESM exhibited potential for predicting the NAC effect in breast cancer, which could guide treatment decisions. KEY POINTS: • The intratumoral and peritumoral CESM-based radiomics signatures show good performance in predicting the NAC effect in breast cancer.

An iterative reconstruction method for sparse-projection data for low-dose CT.

Reducing X-ray radiation is beneficial for reducing the risk of cancer in patients. There are two main approaches for achieving this goal namely, one is to reduce the X-ray current, and another is to apply sparse-view protocols to do image scanning and projections. However, these techniques usually lead to degradation of the reconstructed image quality, resulting in excessive noise and severe edge artifacts, which seriously affect the diagnosis result. In order to overcome such limitation, this study proposes and tests an algorithm based on guided kernel filtering. The algorithm combines the characteristics of anisotropic edges between adjacent image voxels, expresses the relevant weights with an exponential function, and adjusts the weights adaptively through local gray gradients to better preserve the image structure while suppressing noise information. Experiments show that the proposed method can effectively suppress noise and preserve the image structure. Comparing with similar algorithms, the proposed algorithm greatly improves the peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and root mean square error (RMSE) of the reconstructed image. The proposed algorithm has the best effect in quantitative analysis, which verifies the effectiveness of the proposed method and good image reconstruction performance. Overall, this study demonstrates that the proposed method can reduce the number of projections required for repeated CT scans and has potential for medical applications in reducing radiation doses.

PWLS-PR: low-dose computed tomography image reconstruction using a patch-based regularization method based on the penalized weighted least squares total variation approach.

BACKGROUND: Radiation exposure computed tomography (CT) scans and the associated risk of cancer in patients have been major clinical concerns. Existing research can achieve low-dose CT imaging by reducing the X-ray current and the number of projections per rotation of the human body. However, this method may produce excessive noise and fringe artifacts in the traditional filtered back projection (FBP)-reconstructed image. METHODS: To solve this problem, iterative image reconstruction is a promising option to obtain high-quality images from low-dose scans. This paper proposes a patch-based regularization method based on penalized weighted least squares total variation (PWLS-PR) for iterative image reconstruction. This method uses neighborhood patches instead of single pixels to calculate the nonquadratic penalty. The proposed regularization method is more robust than the conventional regularization method in identifying random fluctuations caused by sharp edges and noise. Each iteration of the proposed algorithm can be described in the following three steps: image updating via the total variation based on penalized weighted least squares (PWLS-TV), image smoothing, and pixel-by-pixel image fusion. RESULTS: Simulation and real-world projection experiments show that the proposed PWLS-PR algorithm achieves a higher image reconstruction performance than similar algorithms. Through the qualitative and quantitative evaluation of simulation experiments, the effectiveness of the method is also verified. CONCLUSIONS: Furthermore, this study shows that the PWLS-PR method reduces the amount of projection data required for repeated CT scans and has the useful potential to reduce the radiation dose in clinical medical applications.

Isolated interrupted aortic arch in an adult: A case report.

BACKGROUND: Interrupted aortic arch (IAA) is a rare congenital heart disease defined by an interruption of the lumen and anatomical continuity between the ascending and descending major arteries. It is usually found within a few hours or days of birth. Without surgery, the chances of survival are low. If IAA patients have an effective collateral circulation established, they can survive into adulthood. However, IAA in adults is extremely rare, with few reported cases. CASE SUMMARY: A 27-year-old woman presented with a 6-year history of progressively worsening shortness of breath and chest tightness on exertion. She had cyanotic lips and clubbing of the fingers. A transthoracic echocardiogram revealed an enlarged heart and dilation of the main pulmonary artery. There was an abnormal 9 mm passage between the descending aorta and pulmonary artery. The ventricular septal outflow tract had a 14 mm defect. Doppler ultrasound suggested a patent ductus arteriosus and computed tomographic angiography showed the absence of the aortic arch. The diagnoses were ventricular septal defect, patent ductus arteriosus, and definite interruption of the aortic arch. Although surgical correction was recommended, the patient declined due to the surgical risks and was treated with medications to reduce pulmonary artery pressure and treat heart failure. Her condition has been stable for 12 mo of follow-up. CONCLUSION: Although rare, IAA should be considered in adults with refractory hypertension or unexplained congestive heart failure.