Combination of leucine-rich alpha-2 glycoprotein and fecal markers detect Crohn's disease activity confirmed by balloon-assisted enteroscopy.

BACKGROUND/AIMS: Endoscopic activity confirmed by enteroscopy is associated with poor clinical outcome in Crohn's disease (CD). We investigated which of the existing biomarkers best reflects endoscopic activity in CD patients including the small bowel, and whether their combined use can improve accuracy. METHODS: One hundred and four consecutive patients with ileal and ileocolonic type CD who underwent balloon-assisted enteroscopy (BAE) from October 2021 to August 2022 were enrolled, with clinical and laboratory data prospectively collected and analyzed. RESULTS: Hemoglobin, platelet count, C-reactive protein, leucine-rich alpha-2 glycoprotein (LRG), fecal calprotectin, and fecal hemoglobin all showed significant difference in those with ulcers found on BAE. LRG and fecal calprotectin showed the highest areas under the curve (0.841 and 0.853) for detecting ulcers. LRG showed a sensitivity of 78% and specificity of 80% at a cutoff value of 13 µg/mL, whereas fecal calprotectin showed a sensitivity of 91% and specificity of 67% at a cutoff value of 151 µg/g. Dual positivity for LRG and fecal calprotectin, as well as LRG and fecal hemoglobin, both predicted ulcers with an improved specificity of 92% and 100%. A positive LRG or fecal calprotectin/hemoglobin showed an improved sensitivity of 96% and 91%. Positivity for LRG and either of the fecal biomarkers was associated with increased risk of hospitalization, surgery, and relapse. CONCLUSIONS: The biomarkers LRG, fecal calprotectin, and fecal hemoglobin can serve as noninvasive and accurate tools for assessing activity in CD patients confirmed by BAE, especially when used in combination.

Association of AI-determined Kellgren-Lawrence grade with medial meniscus extrusion and cartilage thickness by AI-based 3D MRI analysis in early knee osteoarthritis.

The associations among Kellgren-Lawrence (KL) grade, medial meniscus extrusion (MME), and cartilage thickness in knee osteoarthritis (OA) remain insufficiently understood. Our aim was to determine these associations in early to moderate medial tibiofemoral knee OA. We included 469 subjects with no lateral OA from the Kanagawa Knee Study. KL grade was assessed using artificial intelligence (AI) software. The MME was measured by MRI, and the cartilage thickness was evaluated in 18 subregions of the medial femorotibial joint by another AI system. The median MME width was 1.4 mm in KL0, 1.5 mm in KL1, 2.4 mm in KL2, and 6.0 mm in KL3. Cartilage thinning in the medial femur occurred in the anterior central subregion in KL1, expanded inwardly in KL2, and further expanded in KL3. Cartilage thinning in the medial tibia occurred in the anterior and middle external subregions in KL1, expanded into the anterior and middle central subregions in KL2, and further expanded in KL3. The absolute correlation coefficient between MME width and cartilage thickness increased as the KL grade increased in some subregions. This study provides novel insights into the early stages of knee OA and potentially has implications for the development of early intervention strategies.

Novel utility of Vesical Imaging-Reporting and Data System in multimodal treatment for muscle-invasive bladder cancer.

OBJECTIVES: To examine the clinical significance of the Vesical Imaging-Reporting and Data System (VI-RADS) in predicting outcome of multimodal treatment (MMT) in muscle-invasive bladder cancer (MIBC) patients. METHODS: We reviewed 78 pathologically proven MIBC patients who underwent MMT including transurethral resection and chemoradiotherapy, followed by partial or radical cystectomy. Treatment response was assessed through histologic evaluation of cystectomy specimens. Two radiologists categorized the index lesions of pretherapeutic MRI according to the 5-point VI-RADS score. The associations of VI-RADS score with the therapeutic effect of MMT were analyzed. The diagnostic performance of VI-RADS scores with a cut-off VI-RADS scores ≤ 2 or ≤ 3 for predicting pathologic complete response to MMT (MMT-CR) was evaluated. RESULTS: MMT-CR was achieved in 2 (100%) of VI-RADS score 1 (n = 2), 16 (84%) of score 2 (n = 19), 12 (86%) of score 3 (n = 14), 7 (64%) of score 4 (n = 11), and 14 (44%) of score 5 (n = 32). VI-RADS score was inversely associated with the incidence of MMT-CR (p = 0.00049). The cut-off VI-RADS score ≤ 2 and ≤ 3 could predict the favorable therapeutic outcome of MMT with high specificity (0.89 with 95% confidence interval [CI]: 0.71-0.98 and 0.82 with 95% CI: 0.62-0.94, respectively) and high positive predictive value (0.86 with 95% CI: 0.64-0.97 and 0.86 with 95% CI: 0.70-0.95, respectively). CONCLUSION: VI-RADS score may serve as an imaging marker in MIBC patients for predicting the therapeutic outcome of MMT. CLINICAL RELEVANCE STATEMENT: Muscle-invasive bladder cancer patients with a lower Vesical Imaging-Reporting and Data System score can be a good candidate for bladder-sparing treatment incorporating multimodal treatment. KEY POINTS: • Vesical Imaging-Reporting and Data System (VI-RADS) score was potentially valuable for classifying pathologic tumor response in patients with muscle-invasive bladder cancer. • The likelihood of achieving complete response of multimodal treatment (MMT) decreased with increasing VI-RADS score. • VI-RADS score could serve as an imaging marker that optimizes patient selection for MMT.

Transmural Remission Characterized by High Biologic Concentrations Demonstrates Better Prognosis in Crohn's Disease.

BACKGROUND: The importance and pathophysiology of transmural healing in patients with Crohn's disease [CD] remains to be verified. We aimed to examine the association between serum concentrations of biologics and transmural remission evaluated via magnetic resonance enterography [MRE]. METHODS: We enrolled patients with CD who received maintenance biologics 1 year after induction and prospectively followed up for at least 1 year after baseline laboratory, endoscopic and MRE examination. We evaluated the relationship between baseline factors including the presence of transmural remission and patient prognosis, as well as between serum concentrations and transmural remission. RESULTS: We included 134 patients, of whom 65, 31, 27 and 11 received infliximab, adalimumab, ustekinumab and vedolizumab, respectively. Those who achieved transmural remission showed a lower risk of hospitalization and surgery than those who did not achieve remission [p < 0.01]. Adjusted hazard ratios of transmural remission for predicting hospitalization and surgery were 0.11 and 0.02, respectively, which were lower than those of clinical remission, biochemical remission and endoscopic remission. Regarding serum concentrations, the median concentration was higher in patients with transmural remission than in patients with transmural activity for all agents [p < 0.01 for infliximab, p = 0.04 for adalimumab, p < 0.01 for ustekinumab, p = 0.08 for vedolizumab]. CONCLUSIONS: Transmural remission was the best predictor for prognosis in CD patients who received maintenance biologic therapy. High drug concentration levels were associated with transmural remission confirmed via MRE.

Serum Leucine-Rich α2 Glycoprotein: A Novel Biomarker for Transmural Inflammation in Crohn's Disease.

INTRODUCTION: Leucine-rich alpha-2 glycoprotein (LRG) is a newly studied biomarker for inflammatory diseases. This study aimed to investigate whether LRG can be used for evaluating transmural activity in patients with Crohn's disease (CD). METHODS: We performed magnetic resonance enterography (MRE) in 227 consecutive patients with CD from June 2020 to August 2021. We prospectively compared MRE findings with clinical and laboratory data including LRG. MRE was evaluated using 2 validated scoring systems, and transmural inflammation was defined as having a maximum simplified magnetic resonance index of activity (sMaRIA) score of ≥4 and a 5-point classification score of ≥9, respectively. RESULTS: The correlation between LRG and the total MRE score showed a positive correlation ( r = 0.576 for the sMaRIA score, P < 0.01, and r = 0.633 for the 5-point score, P < 0.01). Serum concentrations of LRG significantly increased as MRE scores increased ( P < 0.01). The area under the curve of LRG for a sMaRIA score of ≥4 and a 5-point score of ≥9 was 0.845 and 0.869, respectively, which was significantly higher than that of CDAI ( P < 0.01) or C-reactive protein ( P < 0.01). LRG levels of ≥14 µg/mL had a 67% sensitivity and 90% specificity for a sMaRIA score of ≥4 and a 73% sensitivity and 89% specificity for a 5-point score of ≥9. Patients with high LRG levels were also strongly associated with CD-related hospitalization, surgery, and clinical relapse compared with those with low LRG levels ( P < 0.01 for all). DISCUSSION: LRG is a highly accurate serum biomarker for detecting transmural activity in patients with CD. Results need to be validated in further multicenter studies.

Motility Mapping Quantification Using the Classical Optical Flow Algorithm for Small Bowel Crohn's Disease: Comparison with Balloon-assisted Enteroscopy Findings.

PURPOSE: To quantify bowel motility shown on cine MRI using the classical optical flow algorithm and compare it with balloon-assisted enteroscopy (BAE) findings in patients with Crohn's disease (CD). METHODS: This retrospective study included 29 consecutive patients with CD who had undergone MR enterocolonography (MREC) and BAE between March and May 2017. We developed computer software to present motion vector magnitudes between consecutive cine MR images as bowel motility maps via a classical optical flow algorithm using the Horn-Schunck method. Cine MR images were acquired with a balanced steady-state free precession sequence in the coronal direction to capture small bowel motility. The small bowels were divided into three segments. In total, 63 bowel segments were assessed via BAE and MREC. Motility scores on the maps, simplified MR index of activity (sMaRIA), and MREC score derived from a 5-point MR classification were assessed independently by two radiologists and compared with the CD endoscopic index of severity (CDEIS). Correlations were assessed using Spearman's rank coefficient. The areas under the receiver-operating characteristic curve (AUCs) of motility score for differentiating CDEIS was calculated; a P value < 0.05 was considered statistically significant. RESULTS: Motility score was negatively correlated with CDEIS (r = -0.59 [P < 0.001] and -0.54 [P < 0.001]), and the AUCs of motility scores for detecting CDEIS ≥ 3 were 88.2% and 78.6% for observers 1 and 2, respectively. There were no significant differences in the AUC for detecting CDEIS ≥ 3 and CDEIS ≥ 12 between motility and sMaRIA or MREC score. CONCLUSION: The motility map was feasible for locally quantifying the bowel motility. In addition, the motility score on the map reflected the endoscopic inflammatory activity of each small bowel segment in patients with CD; hence, it could be used as a tool in objectively interpreting cine MREC to predict inflammatory activity in CD.

Can Deep Learning-Based Volumetric Analysis Predict Oxygen Demand Increase in Patients with COVID-19 Pneumonia?

Background and Objectives: This study aimed to investigate whether predictive indicators for the deterioration of respiratory status can be derived from the deep learning data analysis of initial chest computed tomography (CT) scans of patients with coronavirus disease 2019 (COVID-19). Materials and Methods: Out of 117 CT scans of 75 patients with COVID-19 admitted to our hospital between April and June 2020, we retrospectively analyzed 79 CT scans that had a definite time of onset and were performed prior to any medication intervention. Patients were grouped according to the presence or absence of increased oxygen demand after CT scan. Quantitative volume data of lung opacity were measured automatically using a deep learning-based image analysis system. The sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) of the opacity volume data were calculated to evaluate the accuracy of the system in predicting the deterioration of respiratory status. Results: All 79 CT scans were included (median age, 62 years (interquartile range, 46-77 years); 56 (70.9%) were male. The volume of opacity was significantly higher for the increased oxygen demand group than for the nonincreased oxygen demand group (585.3 vs. 132.8 mL, p < 0.001). The sensitivity, specificity, and AUC were 76.5%, 68.2%, and 0.737, respectively, in the prediction of increased oxygen demand. Conclusion: Deep learning-based quantitative analysis of the affected lung volume in the initial CT scans of patients with COVID-19 can predict the deterioration of respiratory status to improve treatment and resource management.

Crohn disease: magnetic resonance enterocolonography features of endoscopic ulcer stages reclassified with the healing process and the relationships to prognoses.

PURPOSE: This study aimed to compare magnetic resonance enterocolonography (MREC) features among the endoscopic ulcer stages reclassified to include healing ulcers and to assess the prognoses in Crohn disease (CD). METHODS: Altogether, 89 consecutive patients with CD who had undergone MREC and ileocolonoscopy or balloon-assisted enteroscopy were retrospectively studied. Patients were reclassified into 38 patients with no deep ulcer, seven with healing deep ulcer, and 44 with active deep ulcer stage. MREC score derived from a 5-point MR classification and MR index of activity (MaRIA) were evaluated, and patients were followed-up. The primary endpoint was hospitalization. RESULTS: Healing deep ulcers had higher values in MREC score and MaRIA than no deep ulcers (p < 0.001), and lower values than active deep ulcers (p < 0.001). The 5-year cumulative rates of hospitalization for no deep ulcer, healing deep ulcer, and active deep ulcers were 24.9, 0, and 52.4% (p < 0.05), respectively. MREC score or MaRIA-positive patients had a higher 5-year cumulative rate of hospitalization than the negative patients (p < 0.01 and p < 0.05, respectively). CONCLUSION: MREC could reflect the healing stages, and the identification was revealed to be important because of the good prognosis. MREC might be useful to predict prognosis of CD.

Deep-learning approach with convolutional neural network for classification of maximum intensity projections of dynamic contrast-enhanced breast magnetic resonance imaging.

PURPOSE: We aimed to evaluate deep learning approach with convolutional neural networks (CNNs) to discriminate between benign and malignant lesions on maximum intensity projections of dynamic contrast-enhanced breast magnetic resonance imaging (MRI). METHODS: We retrospectively gathered maximum intensity projections of dynamic contrast-enhanced breast MRI of 106 benign (including 22 normal) and 180 malignant cases for training and validation data. CNN models were constructed to calculate the probability of malignancy using CNN architectures (DenseNet121, DenseNet169, InceptionResNetV2, InceptionV3, NasNetMobile, and Xception) with 500 epochs and analyzed that of 25 benign (including 12 normal) and 47 malignant cases for test data. Two human readers also interpreted these test data and scored the probability of malignancy for each case using Breast Imaging Reporting and Data System. Sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUC) were calculated. RESULTS: The CNN models showed a mean AUC of 0.830 (range, 0.750-0.895). The best model was InceptionResNetV2. This model, Reader 1, and Reader 2 had sensitivities of 74.5%, 72.3%, and 78.7%; specificities of 96.0%, 88.0%, and 80.0%; and AUCs of 0.895, 0.823, and 0.849, respectively. No significant difference arose between the CNN models and human readers (p > 0.125). CONCLUSION: Our CNN models showed comparable diagnostic performance in differentiating between benign and malignant lesions to human readers on maximum intensity projection of dynamic contrast-enhanced breast MRI.

Virtual Interpolation Images of Tumor Development and Growth on Breast Ultrasound Image Synthesis With Deep Convolutional Generative Adversarial Networks.

OBJECTIVES: We sought to generate realistic synthetic breast ultrasound images and express virtual interpolation images of tumors using a deep convolutional generative adversarial network (DCGAN). METHODS: After retrospective selection of breast ultrasound images of 528 benign masses, 529 malignant masses, and 583 normal breasts, 20 synthesized images of each were generated by the DCGAN. Fifteen virtual interpolation images of tumors were generated by changing the value of the input vector. A total of 60 synthesized images and 20 virtual interpolation images were evaluated by 2 readers, who scored them on a 5-point scale (1, very good; to 5, very poor) and then answered whether the synthesized image was benign, malignant, or normal. RESULTS: The mean score of overall quality for synthesized images was 3.05, and that of the reality of virtual interpolation images was 2.53. The readers classified the generated images with a correct answer rate of 92.5%. CONCLUSIONS: A DCGAN can generate high-quality synthetic breast ultrasound images of each pathologic tissue and has the potential to create realistic virtual interpolation images of tumor development.

The Utility of Deep Learning in Breast Ultrasonic Imaging: A Review.

Breast cancer is the most frequently diagnosed cancer in women; it poses a serious threat to women's health. Thus, early detection and proper treatment can improve patient prognosis. Breast ultrasound is one of the most commonly used modalities for diagnosing and detecting breast cancer in clinical practice. Deep learning technology has made significant progress in data extraction and analysis for medical images in recent years. Therefore, the use of deep learning for breast ultrasonic imaging in clinical practice is extremely important, as it saves time, reduces radiologist fatigue, and compensates for a lack of experience and skills in some cases. This review article discusses the basic technical knowledge and algorithms of deep learning for breast ultrasound and the application of deep learning technology in image classification, object detection, segmentation, and image synthesis. Finally, we discuss the current issues and future perspectives of deep learning technology in breast ultrasound.

Dynamic Enhancement Pattern on CT for Predicting Pancreatic Neuroendocrine Neoplasms with Low PAX6 Expression: A Retrospective Observational Study.

Paired box 6 (PAX6) is a transcription factor that plays a critical role in tumor suppression, implying that the downregulation of PAX6 promotes tumor growth and invasiveness. This study aimed to examine dynamic computed tomography (CT) features for predicting pancreatic neuroendocrine neoplasms (Pan-NENs) with low PAX6 expression. We retrospectively evaluated 51 patients with Pan-NENs without synchronous liver metastasis to assess the pathological expression of PAX6. Two radiologists analyzed preoperative dynamic CT images to determine morphological features and enhancement patterns. We compared the CT findings between low and high PAX6 expression groups. Pathological analysis identified 11 and 40 patients with low and high PAX6 expression, respectively. Iso- or hypoenhancement types in the arterial and portal phases were significantly associated with low PAX6 expression (p = 0.009; p = 0.001, respectively). Low PAX6 Pan-NENs showed a lower portal enhancement ratio than high PAX6 Pan-NENs (p = 0.044). The combination based on enhancement types (iso- or hypoenhancement during arterial and portal phases) and portal enhancement ratio (≤1.22) had 54.5% sensitivity, 92.5% specificity, and 84.3% accuracy in identifying low PAX6 Pan-NENs. Dynamic CT features, including iso- or hypoenhancement types in the arterial and portal phases and lower portal enhancement ratio may help predict Pan-NENs with low PAX6 expression.

High b-value computed diffusion-weighted imaging for differentiating bowel inflammation in Crohn's disease.

Purpose This study evaluated whether high b-value computed diffusion-weighted imaging (cDWI) can improve detection and differentiation of bowel inflammation in patients with Crohn's disease (CD). Methods Fifty-four consecutive CD patients who had undergone magnetic resonance enterography (MRE) and ileocolonoscopy (ICS) or balloon-assisted enteroscopy (BAE) were retrospectively studied; cDWI with a b-value = 1500s/mm2 (cDWI1500) was generated using DWI acquired with b-values of 0 and 800 s/mm2 (aDWI800). Overall, 366 bowel segments were evaluated. The signal intensities (SIs) of the bowel lumina were visually assessed on DWI. Bowel wall-to-iliopsoas muscle SI ratios on aDWI800 and cDWI1500 images and apparent diffusion coefficient (ADC) values were measured; visual assessments for lesion detection were performed using a 5-point Likert-like scale on plain MRE with aDWI800, plain MRE with cDWI1500, and contrast-enhanced (CE)-MRE without DWI. The area under the receiver-operating characteristic curve (AUC) was calculated to compare quantitative and qualitative assessments. Results SIs of the intraluminal fluid were shown as comparable to, or lower than background SIs on 157 (44.7 %) and 345 (98.3 %) of 351 segments on aDWI800 and cDWI1500, respectively. AUCs of SI ratios on cDWI1500 images (82.0 %, [95 % confidence interval: 76.6-87.3 %]) were greater than on aDWI800 (75.2 %, [68.2-82.3 %]; p < 0.001), and were close to the ADC values (81.5 % [76.3-86.7 %]; p = 0.76). The AUCs of CE-MRE images were largest, followed by plain MRE with cDWI1500, and plain MRE with aDWI800. Conclusions As it suppresses the SIs of intraluminal fluid and improves contrast between severe and non-severe inflammation, cDWI1500 helps with CD evaluation.

Evaluation of the Usefulness of CO-RADS for Chest CT in Patients Suspected of Having COVID-19.

The purpose of this study was to use the Coronavirus Disease 2019 (COVID-19) Reporting and Data System (CO-RADS) to evaluate the chest computed tomography (CT) images of patients suspected of having COVID-19, and to investigate its diagnostic performance and interobserver agreement. The Dutch Radiological Society developed CO-RADS as a diagnostic indicator for assessing suspicion of lung involvement of COVID-19 on a scale of 1 (very low) to 5 (very high). We investigated retrospectively 154 adult patients with clinically suspected COVID-19, between April and June 2020, who underwent chest CT and reverse transcription-polymerase chain reaction (RT-PCR). The patients' average age was 61.3 years (range, 21-93), 101 were male, and 76 were RT-PCR positive. Using CO-RADS, four radiologists evaluated the chest CT images. Sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) were calculated. Interobserver agreement was calculated using the intraclass correlation coefficient (ICC) by comparing the individual reader's score to the median of the remaining three radiologists. The average sensitivity was 87.8% (range, 80.2-93.4%), specificity was 66.4% (range, 51.3-84.5%), and AUC was 0.859 (range, 0.847-0.881); there was no significant difference between the readers (p > 0.200). In 325 (52.8%) of 616 observations, there was absolute agreement among observers. The average ICC of readers was 0.840 (range, 0.800-0.874; p < 0.001). CO-RADS is a categorical taxonomic evaluation scheme for COVID-19 pneumonia, using chest CT images, that provides outstanding performance and from substantial to almost perfect interobserver agreement for predicting COVID-19.

Feasibility of new fat suppression for breast MRI using pix2pix.

PURPOSE: To generate and evaluate fat-saturated T1-weighted (FST1W) image synthesis of breast magnetic resonance imaging (MRI) using pix2pix. MATERIALS AND METHODS: We collected pairs of noncontrast-enhanced T1-weighted an FST1W images of breast MRI for training data (2112 pairs from 15 patients), validation data (428 pairs from three patients), and test data (90 pairs from 30 patients). From the original images, 90 synthetic images were generated with 50, 100, and 200 epochs using pix2pix. Two breast radiologists evaluated the synthetic images (from 1 = excellent to 5 = very poor) for quality of fat suppression, anatomic structures, artifacts, etc. The average score was analyzed for each epoch and breast density. RESULTS: The synthetic images were scored from 2.95 to 3.60; the best was reduction in artifacts when using 100 epochs. The average overall quality scores for fat suppression were 3.63 at 50 epochs, 3.24 at 100 epochs, and 3.12 at 200 epochs. In the analysis for breast density, each score was significantly better for nondense breasts than for dense breasts; the average score was 2.88-3.18 for nondense breasts and 3.03-3.42 for dense breasts (P = 0.000-0.042). CONCLUSION: Pix2pix had the potential to generate FST1W synthesis for breast MRI.

Efficient Anomaly Detection with Generative Adversarial Network for Breast Ultrasound Imaging.

We aimed to use generative adversarial network (GAN)-based anomaly detection to diagnose images of normal tissue, benign masses, or malignant masses on breast ultrasound. We retrospectively collected 531 normal breast ultrasound images from 69 patients. Data augmentation was performed and 6372 (531 × 12) images were available for training. Efficient GAN-based anomaly detection was used to construct a computational model to detect anomalous lesions in images and calculate abnormalities as an anomaly score. Images of 51 normal tissues, 48 benign masses, and 72 malignant masses were analyzed for the test data. The sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) of this anomaly detection model were calculated. Malignant masses had significantly higher anomaly scores than benign masses (p < 0.001), and benign masses had significantly higher scores than normal tissues (p < 0.001). Our anomaly detection model had high sensitivities, specificities, and AUC values for distinguishing normal tissues from benign and malignant masses, with even greater values for distinguishing normal tissues from malignant masses. GAN-based anomaly detection shows high performance for the detection and diagnosis of anomalous lesions in breast ultrasound images.

Classification of Breast Masses on Ultrasound Shear Wave Elastography using Convolutional Neural Networks.

We aimed to use deep learning with convolutional neural networks (CNNs) to discriminate images of benign and malignant breast masses on ultrasound shear wave elastography (SWE). We retrospectively gathered 158 images of benign masses and 146 images of malignant masses as training data for SWE. A deep learning model was constructed using several CNN architectures (Xception, InceptionV3, InceptionResNetV2, DenseNet121, DenseNet169, and NASNetMobile) with 50, 100, and 200 epochs. We analyzed SWE images of 38 benign masses and 35 malignant masses as test data. Two radiologists interpreted these test data through a consensus reading using a 5-point visual color assessment (SWEc) and the mean elasticity value (in kPa) (SWEe). Sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) were calculated. The best CNN model (which was DenseNet169 with 100 epochs), SWEc, and SWEe had a sensitivity of 0.857, 0.829, and 0.914 and a specificity of 0.789, 0.737, and 0.763 respectively. The CNNs exhibited a mean AUC of 0.870 (range, 0.844-0.898), and SWEc and SWEe had an AUC of 0.821 and 0.855. The CNNs had an equal or better diagnostic performance compared with radiologist readings. DenseNet169 with 100 epochs, Xception with 50 epochs, and Xception with 100 epochs had a better diagnostic performance compared with SWEc (P = 0.018-0.037). Deep learning with CNNs exhibited equal or higher AUC compared with radiologists when discriminating benign from malignant breast masses on ultrasound SWE.

Objective evaluation for treat to target in Crohn's disease.

BACKGROUND: Crohn's disease (CD) is a chronic and destructive bowel disease; continued disease activity can lead to penetrating complications. With the recent advent of effective medications, the importance of using a treat-to-target approach to guide therapy is becoming important. METHODS: In this review, we reviewed the previous evidence for evaluating CD lesions. RESULTS: We describe ileocolonoscopy's role in assessing disease activity, as well as recent progress in modalities, such as balloon-assisted endoscopy, capsule endoscopy, magnetic resonance enterography, computed tomography enterography, and ultrasonography. Advances in modalities have changed CD assessment, with small-bowel involvement becoming more important. CONCLUSIONS: Proper optimization is necessary in clinical practice.

Breast Ultrasound Image Synthesis using Deep Convolutional Generative Adversarial Networks.

Deep convolutional generative adversarial networks (DCGANs) are newly developed tools for generating synthesized images. To determine the clinical utility of synthesized images, we generated breast ultrasound images and assessed their quality and clinical value. After retrospectively collecting 528 images of 144 benign masses and 529 images of 216 malignant masses in the breasts, synthesized images were generated using a DCGAN with 50, 100, 200, 500, and 1000 epochs. The synthesized (n = 20) and original (n = 40) images were evaluated by two radiologists, who scored them for overall quality, definition of anatomic structures, and visualization of the masses on a five-point scale. They also scored the possibility of images being original. Although there was no significant difference between the images synthesized with 1000 and 500 epochs, the latter were evaluated as being of higher quality than all other images. Moreover, 2.5%, 0%, 12.5%, 37.5%, and 22.5% of the images synthesized with 50, 100, 200, 500, and 1000 epochs, respectively, and 14% of the original images were indistinguishable from one another. Interobserver agreement was very good (|r| = 0.708-0.825, p < 0.001). Therefore, DCGAN can generate high-quality and realistic synthesized breast ultrasound images that are indistinguishable from the original images.