Search

Purpose: The research aimed to verify the applicability of low computational complexity and high diagnosis accuracy deep convolutional neural network, using MobileNetV2 to identify the presence of chest catheters and tubes on chest X-ray images. Methods: The dataset of chest X-rays collected from a teaching hospital included the endotracheal tube (ETT), the central venous catheter (CVC), and the nasogastric tube (NGT) datasets. A new method of applying dynamic image size training procedures was implemented and compared with fixed image size training. The idea is to learn more features through dynamic image size training. Transfer learning with pre-trained MobileNetV2 on ImageNet was conducted to accelerate the training process and acquire higher accuracy. Class activation mapping (CAM) was also employed to visualize artificial intelligence (AI) predictions, making AI decisions more explainable. Results: The ETT datasets included 10464 X-ray images, while the CVC and NGT datasets contained 10274 and 9610 images, respectively. The accuracies for ETT, CVC, and NGT are 99.0%, 98.4%, and 96.2% in the validation dataset, while in the testing dataset are 98.8%, 98.6%, and 96.8%, respectively. The area under the receiver operating characteristics (AUROCs) were 0.992, 0.988, and 0.980 in the ETT, CVC, and NGT testing datasets. Conclusion: MobileNetV2 with the dynamic image size achieved dedicated performance in the application of chest catheters and tubes classifications. The similarity of the accuracy between the validation and testing data suggests the good generalization capability of the model.

In this paper, we consider an integrated error concealment system for lost color frames and lost depth frames in multiview videos with depths. We first proposed a pixel-based color error-concealment method with the use of depth information. Instead of assuming that the same moving object in consecutive frames has minimal depth difference, as is done in a state-of-the-art method, a more realistic situation in which the same moving object in consecutive frames can be in different depths is considered. In the derived motion vector candidate set, we consider all the candidate motion vectors in the set, and weight the reference pixels by the depth differences to obtain the final recovered pixel. Compared with the two state-of-the-art methods, the proposed method has average peak signal-to-noise ratio gains of up to 8.73 and 3.98 dB, respectively. Second, we proposed an iterative depth frame error-concealment method. The initial recovered depth frame is obtained by depth-image-based rendering from another available view. The holes in the recovered depth frame are then filled in the proposed priority order. Preprocessing methods (depth difference compensation and inconsistent pixel removal) are performed to improve the performance. Compared with a method that uses the available motion vector in a color frame to recover the lost depth pixels, the hybrid motion vector extrapolation method, the inpainting method and the proposed method have gains of up to 4.31, 10.29, and 6.04 dB, respectively. Finally, for the situation in which the color and the depth frames are lost at the same time, our two methods jointly perform better with a gain of up to 7.79 dB.

This paper proposes four logic-chain bridging fault models, which involve one net in the combinational logic and the other net in the scan chain. Test results of logic-chain bridging faults, unlike existing scan chain fault models, depend on the previous scan inputs as well as primary inputs. A bridging pair extraction algorithm is proposed to quickly extract bridging pairs from the layout. The paper proposed two sets of structural reduction techniques so that runtime is very short. Experimental results on ISCAS benchmark circuits show that, on the average, logic-chain bridging faults can be diagnosed within an accuracy of four bridging pairs. The techniques are still applicable when there are only 10 failing patterns due to limited ATE failure memory. This paper demonstrates the feasibility to diagnose logic-chain bridging faults by software.

OBJECTIVE. Early diagnosis of prosthetic valve disorders is very important in reducing morbidity and mortality of patients with a replaced heart valve. Recently, MDCT has become an important noninvasive problem-solving tool in clinical practice. This article will introduce the scanning technique, interpretation algorithm, and image findings of many valvular and extravalvular diagnoses.CONCLUSION. With proper scanning technique and interpretation, MDCT can be a powerful technique for prosthetic heart valve evaluation.

With the increased availability of multi-detector row CT, indirect CT venography has become an important non-invasive image modality for patients with suspected deep vein thrombosis. Use of indirect CT venography can not only diagnose/exclude deep vein thrombosis, but can also determine if there are other anomalies or diseases which might contribute to the patient's symptoms. In this pictorial essay, we introduce the scanning protocol, post-processing techniques, and interpretation algorithm used in widely available 64 multi-detector row technology. We discuss several cases, including deep vein thrombosis in acute and chronic stages, anatomic variation, vena cava filter, and collateral veins. Lastly, we consider alternative diagnoses including varicose veins, infection, prosthesis failure of arthroplasty, traumatic vessel injury, and other musculoskeletal conditions. Radiologists should be familiar with the comprehensive interpretation of indirect CT venography to facilitate differential diagnosis and further treatment decision.

OBJECTIVE. CT-guided core biopsy is playing an increasing role in the diagnosis of benign disease, cellular differentiation, somatic mutation analysis, and molecular fingerprint analysis.CONCLUSION. In this article, we summarize the basic concepts, protocols, and techniques that we use for CT-guided core biopsy of lung lesions to assist radiologists in obtaining diagnostic specimens while reducing preventable complications.

The objective of our study was to evaluate whether iodixanol 320 mg I/mL (iodixanol 320), with the highest iodine concentration of dimeric nonionic contrast agents on the market, results in decreased vascular or myocardial enhancement compared with iohexol 350 mg I/mL (iohexol 350).During a 4-month period, 72 patients referred for cardiac MDCT were consecutively enrolled and randomized into two groups: iohexol 350 and iodixanol 320. The injection and scanning protocols were the same for both groups. Enhancement of the right heart, left heart, coronary arteries, and left ventricular (LV) myocardium in both the arterial and delayed phases was compared using two-tailed independent Student's t test.Enhancement in the right heart, left heart, coronary arteries, and LV myocardium in the arterial phase showed no statistical difference (p0.05) between the two groups, although the iohexol group showed slightly higher enhancement (average, 11.2 H) in all of the areas. Surprisingly, in the delayed phase, the iodixanol group displayed significantly higher (7.7 H) persistent enhancement (p0.05) in the LV myocardium.Iodixanol 320 can provide vascular enhancement in cardiac MDCT that is similar to iohexol 350. In the delayed phase, iodixanol 320 shows significantly higher delayed enhancement (7.7 H) in the LV myocardium than iohexol 350.

Using catheter coronary angiography (CAG) as reference standard, we examined the agreement of 40-detector row computed tomography (MDCT) in triaging patients into the 2 controversial strategies of managing low-risk acute coronary syndrome (ACS).Seventy-eight patients with low-risk ACS received both MDCT and CAG. Early invasive strategy was assigned for the patient if there was significant stenosis (or =50% diameter stenosis) in any of the coronary artery segments with diameter larger than 1.5 mm. The results of MDCT were compared with the CAG for agreement.The overall agreement of the early conservative/early invasive strategy assignment was 92.3%, with kappa value of 0.82 between MDCT and CAG. Only 1 patient needing early invasive strategy was missed by MDCT.Forty-detector row computed tomography is reliable in triaging patients into the 2 strategies of managing low-risk ACS.

3Department of Pathology, Taichung Veterans General Hospital, Taichung, Taiwan, R.O.C. e report on the clinical data, radiology images, and findings at surgery of the first left atrial bronchogenic cyst in the literature. Combined persistent left superior vena cava (SVC) with intracardiac bronchogenic cyst is unique. A 68-year-old woman with bronchiectasis was incidentally found on a chest CT scan to have an intracardiac bronchogenic cyst. Though it is difficult preoperatively to diagnosis this disorder, a complete differential diagnostic list for a cystic lesion inside the heart chambers should include bronchogenic cyst as a possibility. Intracardiac bronchogenic cysts are exceedingly rare with only three cases reported in the literature. Furthermore, no bronchogenic cyst inside the left atrium has been documented. Combined persistent left SVC with intracardiac bronchogenic cyst also is unique. We report such a case based on CT and MRI. Animation of the cardiac cycle from an MRI cardiac functional study and the hemodynamic change associated with the lesion are presented. The possible embryologic explanation also is discussed.

The video quality after source compression loss and communication channel loss is very important for user experience in consumer electronic. In this paper, we propose a rate-distortion optimization method considering not only source loss, but also channel loss. We modify the source Lagrangian function in the H.264 to further consider channel loss effects, measured by a packet loss visibility model. The experiment shows that compared to the H.264 standard, the proposed method can achieve better combined source-loss and channel-loss video quality, where the winning percentage in channel-loss video quality can be up to 90.63% compared to the standard.

HEVC (High Efficiency Video Coding) standard is ready to be launched for consumer electronics. With better coding efficiency, the coding complexity is high. We propose a fast CU (Coding Unit) depth decision algorithm based on SKIP mode information in the neighboring block to early terminate the CU splitting. Compared to the HEVC standard software, the proposed method can save by up to 41% of the time. The proposed method also saves by up to 16% of the time compared to a state-of-the-art work. The proposed method introduces very negligible losses in the video PSNR and bitrate.

In this paper, we propose an error concealment algorithm that considers the partition decision information from previous frame, to improve a widely used algorithm that extrapolates the motion vectors from previous frame. The result shows that objectively and subjectively, the proposed method can perform better. Also, the improvement of our method is larger in HEVC video than in AVC video, due to larger partition decision space in HEVC to account for more various sizes of moving objects in a frame.

OBJECTIVE. Pulmonary hypertension is a challenge for imagers and clinicians, with a variety of possible underlying causes, each with its own specific treatment. Although the diagnosis is based on physiologic measurements, ECG-gated MDCT can play a vital role in elucidating underlying cardiac, vascular, and pulmonary causes. CONCLUSION. A revised system for pulmonary hypertension, the Dana Point classification, can provide a template for review of the myriad causes of this complex condition.

Hypereosinophilic syndrome is a rare disease that can cause multiple-organ damage. Cardiac involvement, which presents as myocardial necrosis, infarction, or thrombus formation, is the leading cause of morbidity and death. Here, we present the case of a 17-year-old girl who had hypereosinophilic syndrome with cardiac involvement, which was evaluated by multidetector-row computed tomography. By means of arterial-phase and low-dose delayed-phase scanning, multidetector-row computed tomography clearly identified the mural thrombus and the infarcted and noninfarcted myocardium—which surpassed the performance of echocardiography in tissue characterization. Therefore, we propose that multidetector-row computed tomography could be a useful noninvasive tool for evaluating patients with cardiac involvement of hypereosinophilic syndrome.

This study used multi-detector row CT (MDCT) to evaluate the regional geometric parameters of left ventricle (LV) in patients and dogs after right ventricular apical (RVA) pacing. First, we measured and compared the global and regional wall motion parameters derived from MDCT images between three patients post RVA pacing and seven age-matched healthy subjects. The LV ejection fraction (LVEF), LV end-diastolic volume and LV end-systolic volume were measured. We also measured the regional wall thickness, wall thickening and regional wall motion in 12 different segments of the LV. Second, we performed MDCT scan on five dogs as the study group (pacing wire + RVA pacing, 2 months) and four dogs as the sham control group (pacing wire, 2 months). The global and regional geometric parameters were compared within both human and canine groups. Compared with normal subjects, patients post RVA pacing had low LVEF (60.4 ± 10.5 vs. 33.2 ± 17.6, P = 0.014), impaired regional wall thickening and regional wall motion, particularly in segments near the pacing lead. Some segments near the pacing lead were showing dyskinesia after pacing. These findings were successfully reproduced in the canine model. We found that RVA pacing can result in impaired regional wall thickening and regional wall motion, particularly in segments near the pacing lead.

OBJECTIVE. Patent arteriovenous fistula (AVF) is related to better prognosis and quality of life for patients on long-term dialysis. When AVF dysfunction is suspected, MDCT is a good noninvasive tool for evaluating the entire AVF structure and determining reversible conditions for treatment. The aim of this article is to introduce the scanning and interpretation techniques and to illustrate the conditions related to early and late fistula failures.CONCLUSION. MDCT is a fast, noninvasive, and accurate technique for diagnosing AVF complications. Radiologists familiar with these techniques can help to improve the prognosis and quality of life for hemodialysis patients.

The high accuracy of multidetector-row computed tomography (MDCT) in evaluating prosthetic valve disorders has been confirmed. This, we believe, is the 1st report of the use of MDCT to detect and evaluate left ventricular outflow tract (LVOT) pseudoaneurysms in patients who have undergone aortic valve replacement with prosthetic valves. We used MDCT to scan 21 such patients, 3 of whom had a small pseudoaneurysm in the LVOT. Each pseudoaneurysm projected away from the LVOT and had a narrow neck that was located just below the sewing ring of the prosthetic aortic valve. One pseudoaneurysm was not thrombotic, 1 was partially thrombotic, and 1 was completely thrombotic. One of these had gone undetected earlier on transthoracic echocardiography.We consider MDCT to be superior to echocardiography in the detection of LVOT pseudoaneurysms in patients who have undergone aortic valve replacement with prosthetic valves. We publicize our results in the hope that larger studies will be undertaken in order to investigate the prevalence and clinical implications of our findings.

Recently, CT pulmonary angiography (CTPA), especially performed with multi-detector row CT, has become a key imaging modality for pulmonary embolism. However, CTPA that was performed under clinical suspicion of pulmonary embolism has been shown to lead to high prevalence of alternative diagnosis, up to 25.4%. A comprehensive evaluation of pulmonary and extrapulmonary abnormalities including cardiovascular lesions is critical in proper diagnosis and patient care. Radiologists should be familiar with the comprehensive interpretation of CTPA to facilitate differential diagnosis and further treatment decision.

Sir, In the pictorial essay published in March 2009, Goo et al. [1] shared their comprehensive experience in identifying coronary artery anomalies in patients with congenital heart diseases using current multislice CT techniques. The authors showed us that haemodynamically benign anomalies may have clinical significance. For example, in patients with high-take-off coronary artery, reckless aortic clamping during cardiopulmonary bypass may reduce coronary artery flow. In our hospital, high-take-off of coronary artery is considered to be the most-easily-overlooked surgically important coronary anomaly. We would like to share our experience to emphasize the importance of this to readers. A 7-year-old boy was diagnosed with polysplenia syndrome and levotransposition of the great arteries, double outlet right ventricle, pulmonary stenosis, bilateral superior vena cava and interrupted inferior vena cava with azygous continuation after birth. Due to progressive symptoms of poor exercise tolerance and cyanosis, he was admitted for bidirectional Glenn shunt. Before surgery, MDCT was arranged. MDCT not only confirmed the diagnosis, but also revealed high-take-off of the left anterior descending coronary artery that was positioned beside the main pulmonary artery and passed anterior to the right ventricular outflow tract (Fig. 1). Since high-take-off of a coronary artery is a haemodynamically benign coronary anomaly and usually considered a normal variant [2], the operator focused on the major cardiac structures during operation. After cavopulmonary anastomoses, while attempting to close the main pulmonary artery stump, the dissection procedure resulted in transection of the left anterior descending coronary artery. In order to preserve left ventricular anterior wall perfusion, emergency coronary artery bypass graft surgery was performed using the right internal mammary artery to anastomose the middle left anterior descending coronary artery. Fortunately, the operation ran smoothly, and the patient was discharged uneventfully. No wall motion impairment was identified on follow-up echocardiography. This case highlights that high-take-off of coronary artery, considered to be a minor coronary anomaly or normal variant [2], can have critical surgical importance. The surgical implications of coronary artery anomalies can be different in different operations. For example, if the patient underwent a Blalock-Taussig shunt, the high take-off would not be a problem. We suggest that cardiac surgeons should review the high-resolution axial, coronal and sagittal thinsection images that are provided by ECG-gated MDCT [3] to simulate the surgical procedure before congenital heart operation to avoid coronary artery injury. W.-L. Tsai : I.-C. Tsai (*) Department of Radiology, Taichung Veterans General Hospital, No. 160, Section 3, Taichung Harbor Road, Taichung 407, Taiwan e-mail: sillyduck.radiology@gmail.com

We propose five logic-to-chain bridging fault models, which involve one net in the combinational logic and the other net in the scan chain. Test results of logic-to-chain bridging faults, unlike any existing fault, depend on the previous scan inputs as well as primary inputs. An accurate diagnosis technique is presented to locate logic-to-chain bridging faults. In addition, a bridging pair extraction algorithm is proposed to quickly extract bridging net pairs from the layout. Experimental results on ISCAS benchmark circuits show that, on the average, logic-to-chain bridging faults can be diagnosed within an accuracy of three bridging pairs. The technique is still applicable when only ten failing patterns are recorded on the tester.

OBJECTIVE. Hemodynamic information observed on serial images obtained with bolus tracking might help in diagnosing septal defects on CT. Our purpose was to qualitatively and quantitatively examine a new sign called gradual pulmonary artery enhancement.CONCLUSION. Gradual pulmonary artery enhancement is a newly recognized CT sign that may be helpful in evaluating septal defects.

This study compares the empiric setting and contrast-covering time (CCT) concept for i.v. contrast injection in pediatric thoracic CT aortography.A total of 113 pediatric patients referred for thoracic CT aortography were classified into four groups: group 1 (1-5 years old, CCT), group 2A (6-10 years old, CCT), group 2B (6-10 years old, empiric setting), and group 3 (11-15 years old, CCT). The CT attenuation values from the left common carotid artery to the descending aorta were recorded every 0.5 second. The quantitative bolus geometric analysis of each group included average enhancement, SD within the patient, and slope of enhancement. Groups 2A and 2B were compared to determine whether better bolus geometry could be obtained with the CCT concept than with the traditional empiric setting. Groups 1, 2A, and 3 were compared to determine whether homogeneous bolus geometry could be obtained in different age groups.More homogeneous enhancement was obtained with the CCT concept than the empiric setting with a smaller SD of enhancement (25.5 +/- 8.5 H vs 49.3 +/- 16.2 H, p0.001). Furthermore, in the three different age groups (groups 1, 2A, and 3) examined using the CCT concept, there was no significant difference in the average enhancement (415.7 +/- 83.6 H, 422.8 +/- 97.1 H, 392.0 +/- 78.5 H, respectively; all p0.05), SD of enhancement (28.5 +/- 9.8 H, 25.5 +/- 8.5 H, 28.5 +/- 14.6 H, respectively; all p0.05), or enhancement slopes (-5.6 +/- 18.0 H, -2.7 +/- 10.7 H, -5.4 +/- 12.3 H, respectively; all p0.05).The CCT concept yields more homogeneous enhancement than the empiric setting. It also can routinely obtain homogeneous bolus geometry in patients in different age groups.

In their article published online on 19 April 2005, Leschka et al. 1 shared the first experience of the new 64-slice MSCT. The adequate sample size and evaluation of performance without beta-blocker make the article of high reference value. However, we have several questions to the authors. First, the authors mentioned that the scanner was a 64-MSCT with only 32-detector rows. About the collimation, though the z-sharp technology is used, the total collimation will not be doubled.2 The actual coverage of the collimation should be 32×0.6 mm, but the authors chose to represent it as 64×0.6 mm2, which is …