Deep learning-based classification of adequate sonographic images for self-diagnosing deep vein thrombosis.

BACKGROUND: Pulmonary thromboembolism is a serious disease that often occurs in disaster victims evacuated to shelters. Deep vein thrombosis is the most common reason for pulmonary thromboembolism, and early prevention is important. Medical technicians often perform ultrasonography as part of mobile medical screenings of disaster victims but reaching all isolated and scattered shelters is difficult. Therefore, deep vein thrombosis medical screening methods that can be easily performed by anyone are needed. The purpose of this study was to develop a method to automatically identify cross-sectional images suitable for deep vein thrombosis diagnosis so disaster victims can self-assess their risk of deep vein thrombosis. METHODS: Ultrasonographic images of the popliteal vein were acquired in 20 subjects using stationary and portable ultrasound diagnostic equipment. Images were obtained by frame split from video. Images were classified as "Satisfactory," "Moderately satisfactory," and "Unsatisfactory" according to the level of popliteal vein visualization. Fine-tuning and classification were performed using ResNet101, a deep learning model. RESULTS: Acquiring images with portable ultrasound diagnostic equipment resulted in a classification accuracy of 0.76 and an area under the receiver operating characteristic curve of 0.89. Acquiring images with stationary ultrasound diagnostic equipment resulted in a classification accuracy of 0.73 and an area under the receiver operating characteristic curve of 0.88. CONCLUSION: A method for automatically identifying appropriate diagnostic cross-sectional ultrasonographic images of the popliteal vein was developed. This elemental technology is sufficiently accurate to automatically self-assess the risk of deep vein thrombosis by disaster victims.

Acute heart failure due to left common iliac arteriovenous fistula: A case of VEXAS syndrome.

We describe the case of a 78-year-old man presenting with multiple oedematous erythemas, fever, and arthralgia who subsequently developed neutrophil infiltration into the cartilage of the bilateral auricularis, consistent with relapsing polychondritis. A skin biopsy of the erythema on his right arm showed dense neutrophilic infiltration into the dermis, while a bone marrow aspirate revealed myelodysplastic syndromes with characteristic vacuoles in myeloid precursor cells. Although the patient achieved remission with high-dose oral prednisolone, the inflammatory symptoms relapsed, and he was resistant to colchicine and cyclosporine. The patient spontaneously developed left leg oedema and high-output cardiac failure caused by an arteriovenous fistula with a common iliac artery aneurysm. We successfully performed a two-stage surgery using internal iliac artery coil embolisation and endovascular aortic repair of the iliac aneurysm. We assumed the patient was suffering from large-vessel vasculitis such as giant cell arteritis or Takayasu's arteritis. We treated him with tocilizumab in addition to prednisolone, and the febrile events and elevated C-reactive protein levels improved. One year later, sequencing of ubiquitylation-initiating E1 enzyme using peripheral blood leucocytes revealed somatic variants (c.121A>C p.Met41Leu), confirming the diagnosis of vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic (VEXAS) syndrome. This case suggests that arteriovenous fistula could be a complication of VEXAS syndrome with large-vessel vasculitis, and adequate surgical intervention and prompt diagnosis are essential for rescue. Although arteriovenous fistula is a rare complication of VEXAS syndrome, physicians should be aware of this complication to ensure prompt diagnosis and timely surgical intervention.

Transcatheter coil embolization for large pulmonary arteriovenous fistulae through an artificial tricuspid ball valve.

Pulmonary arteriovenous fistulae (PAFs) occur congenitally or are acquired. A PAF can cause hypoxemia, sudden death from rupture, abscess formation, and embolism. Treatment for PAF is transcatheter embolization or surgery. Transcatheter embolization is the first choice of treatment; however, this treatment is impossible to perform if a patient has had tricuspid or pulmonary valve replacement. In this paper, we describe a case of PAFs complicated with tricuspid valve replacement with a ball valve (which had been performed 40 years earlier) that was treated with transcatheter embolization. .

Low Tube Voltage Computed Tomography Venography for Patients With Deep Vein Thrombosis of the Lower Extremities　- A Comparison With Venous Ultrasonography.

BACKGROUND: Low tube voltage computed tomography venography (CTV) can be expected to increase imaging contrast and decrease radiation exposure by using iterative reconstruction (IR). This study evaluated the diagnostic ability of low tube voltage CTV with IR for deep vein thrombosis (DVT), compared to ultrasonography (US).Methods and Results:Two experienced radiologists retrospectively reevaluated the CTV data of 55 of 318 consecutive patients suspected of having DVT or pulmonary embolism between December 2015 and April 2017. The 55 patients had undergone both low tube voltage CTV and US (within 1 day before or after CTV). The lower extremity veins were divided into 10 segments. The DVT forms were categorized into 3 types: complete, concentric, and eccentric. We analyzed the 534 overall segments (16 segments excluded in US) measured using both CTV and US. The sensitivity-specificity was overall 73.3-90.0%, for femoropopliteal, it was 90.0-93.2%, and for the calf, it was 71.1-87.2%. The diagnostic accuracy between the 'eccentric only' and 'others' groups focusing on DVT forms was compared, and significant differences were revealed, especially in the muscular vein. CONCLUSIONS: The DVT diagnostic ability above the knee was comparable between low tube voltage CTV with IR and conventional CTV, and the radiation dose was reduced. It was suggested that eccentric DVT measured by CTV tend to be a false-positive, especially in the calf muscular vein.

Impact of the genetic variants of GLCCI1 on clinical features of asthmatic patients.

BACKGROUND: Several gene variants are associated with a response to an inhaled corticosteroids (ICSs) treatment in patients with bronchial asthma. A variant of the glucocorticoid-induced transcript 1 (GLCCI1) genes has previously been associated with decreased lung function improvement upon treatment with ICSs in patients with bronchial asthma. Another report has also demonstrated that this genetic biomarker did not influence the change in flow volume in 1 second. However, no studies have considered the treatment content and the GLCCI1 variants. We were able to determine the relationship between the pulmonary function and clinical features and the variant of the GLCCI1 in Japanese asthmatic patients receiving long-term ICS treatment. MATERIALS AND METHODS: In this study, 405 patients with bronchial asthma, who were receiving ICS and living in Japan, were recruited, genotyped and underwent pulmonary function tests. To identify the GLCCI1 protein expression cells, endobronchial biopsy specimens were examined. RESULTS: We found that the pulmonary function was not significantly different in the homozygotes compared to the wild types. Also, the homozygotes increased the risk of a sustained step-up of the asthma treatment when compared to the wild type and heterozygotes. GLCCI1-positive cells were localized to the bronchial epithelial cells. The amount of GLCCI1 protein that cultured epithelial cells harboring GLCCI1 variants produced was less than the GLCCI1 wild type in the presence of a corticosteroid. CONCLUSIONS: A worsening of pulmonary function caused by GLCCI1 variants could be prevented due to recently used medications based on new action mechanisms.

Coronary computed tomography angiography using 128-slice dual-source computed tomography in patients with severe calcification.

PURPOSE: To compare coronary computed tomography angiography (CTA) and coronary angiography (CAG) with regard to luminal graphic definition of calcified segments using 128-slice dual-source computed tomography (DSCT), specifically for patients with an Agatston score >400. MATERIALS AND METHODS: Of 1148 consecutive patients who underwent coronary CTA using a 128-slice DSCT, 132 subjects had severe calcification with an Agatston score >400. Thirty-nine of the 132 patients who had undergone CAG within 3 months before or after coronary CTA were included. We investigated the distribution of calcification, and we visually evaluated significant stenosis in the calcified and all segments. Results were compared with CAG. RESULTS: The target group in this study had a very high mean Agatston score of 1771 ± 1724. Results for sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of 247 calcified vs all 325 segments were as follows: sensitivity 93.2 vs 92.2%, specificity 83.9 vs 87.5%, PPV 70.8 vs 69.6%, and NPV 96.7 vs 97.3%, respectively. CONCLUSION: 128-slice DSCT has potential for evaluation of calcified segments in the lumen, even in patients whose Agatston score exceeds 400.

Mitigation of tight junction protein dysfunction in lung microvascular endothelial cells with pitavastatin.

BACKGROUND: Statin use in individuals with chronic obstructive pulmonary disease (COPD) with coexisting cardiovascular disease is associated with a reduced risk of exacerbations. The mechanisms by which statin plays a role in the pathophysiology of COPD have not been defined. To explore the mechanisms involved, we investigated the effect of statin on endothelial cell function, especially endothelial cell tight junctions. METHOD: We primarily assessed whether pitavastatin could help mitigate the development of emphysema induced by continuous cigarette smoking (CS) exposure. We also investigated the activation of liver kinase B1 (LKB1)/AMP-activated protein kinase (AMPK) signaling, which plays a role in maintaining endothelial functions, important tight junction proteins, zonula occludens (ZO)-1 and claudin-5 expression, and lung microvascular endothelial cell permeability. RESULTS: We found that pitavastatin prevented the CS-induced decrease in angiomotin-like protein 1 (AmotL1)-positive vessels via the activation of LKB1/AMPK signaling and IFN-γ-induced hyperpermeability of cultured human lung microvascular endothelial cells by maintaining the levels of AmotL1, ZO-1, and claudin-5 expression at the tight junctions. CONCLUSION: Our results indicate that the maintenance of lung microvascular endothelial cells by pitavastatin prevents tight junction protein dysfunctions induced by CS. These findings may ultimately lead to new and novel therapeutic targets for patients with COPD.

Acute coronary syndrome: evaluation of detection capability using non-electrocardiogram-gated parenchymal phase CT imaging.

PURPOSE: To evaluate the capability to detect acute coronary syndrome (ACS) by using non-electrocardiogram-gated parenchymal phase CT imaging. MATERIALS AND METHODS: Of 962 consecutive patients who underwent emergent coronary angiography for suspected ACS, 32 with ACS who underwent CT ≤24 h before angiography and 15 without ACS who underwent CT ≤24 h before or after angiography were included. Parenchymal phase was acquired at 100-s scan delay. The presence of a myocardial perfusion defect (MPD) on the left ventricle (a decrease of >20 HU) and its capability to detect ACS were evaluated. Results were compared with laboratory findings. RESULTS: MPD was detected in 29 of 32 ACSs. The sensitivity, specificity, and positive and negative predictive values were 91 % (29/32), 93 % (14/15), 97 % (29/30), and 82 % (14/17), respectively. The sensitivities of ST- and non-ST-elevation ACSs were 89 % (16/18) and 93 % (13/14), respectively, without significant difference (P > 0.99). Of the CT-detectable ACS, non-ST-elevation on the electrocardiogram and a normal creatine kinase-myocardial band were observed in 41 % (12/29) and 24 % (7/29), respectively. CONCLUSION: ACS is highly detectable even using conventional parenchymal phase CT imaging. Therefore, even when CT is non-gating, radiologists should carefully evaluate the heart to avoid overlooking ACS.

Analysis of decrease in lung perfusion blood volume with occlusive and non-occlusive pulmonary embolisms.

PURPOSE: The aim of this study was to determine if lung perfusion blood volume (lung PBV) with non-occlusive pulmonary embolism (PE) differs quantitatively and visually from that with occlusive PE and to investigate if lung PBV with non-occlusive PE remains the same as that without PE. MATERIALS AND METHODS: Totally, 108 patients suspected of having acute PE underwent pulmonary dual-energy computed tomography angiography (DECTA) between April 2011 and January 2012. Presence of PE on DECTA was evaluated by one radiologist. Two radiologists visually evaluated the PE distribution (segmental or subsegmental) and its nature (occlusive or non-occlusive) on DECTA and classified perfusion in lung PBV as "decreased," "slightly decreased," and "preserved". Two radiologists used a lung PBV application to set a region of interest (ROI) in the center of the lesion and measured HU values of an iodine map. In the same slice as the ROI of the lesion and close to the lesion, another ROI was set in the normal perfusion area without PE, and HUs were measured. The proportion of lesions was compared between the occlusive and non-occlusive groups. HUs were compared among the occlusive, non-occlusive, and corresponding normal groups. RESULTS: Twenty-five patients had 80 segmental or subsegmental lesions. There were 37 and 43 lesions in the occlusive and non-occlusive groups, respectively. The proportion of decreased lesions was 73.0% (27/37) in the occlusive group, while that of preserved lesions in the non-occlusive group was 76.7% (33/43). There was a significant difference in the proportion of lesions (P<0.001) between the two groups. HUs of the iodine map were significantly higher in the non-occlusive group than in the occlusive group (33.8 ± 8.2 HU vs. 11.9 ± 6.1 HU, P<0.001). There was no significant difference in HUs for the entire lesion between the non-occlusive (33.8 ± 8.2 HU) and corresponding normal group (34.5 ± 6.8 HU; P=0.294). CONCLUSION: Iodine perfusion tended to be visually and quantitatively preserved in lungs with nonocclusive PE. Lung PBV is required to evaluate pulmonary blood flow.

Where is the most common site of DVT? Evaluation by CT venography.

PURPOSE: Our aim was to clarify the common site of deep venous thrombosis (DVT) in patients suspected of having pulmonary embolism using computed tomography pulmonary angiography with computed tomography venography (CTV). MATERIALS AND METHODS: We evaluated 215 patients. For all studies, 100 ml of 370 mg I/ml nonionic contrast material was administered. CTV were scanned with helical acquisition starting at 3 min in four-slice multidetector-row computed tomography (MDCT) or 5 min in 64-MDCT after the start of contrast material injection. The site of DVT was divided into iliac vein, femoral vein, popliteal vein, or calf vein. Calf vein was divided into muscular (soleal and gastrocnemius) and nonmuscular (anterior/posterior tibial and peroneal) veins. The 2 × 2 chi-square test was used. RESULTS: One hundred and thirty-seven patients showed DVT; the muscular calf vein was more prevalent than other veins (P < 0.01). CONCLUSIONS: Our study showed that the most common site of DVT was the muscular calf vein.

Correlation between the site of pulmonary embolism and the extent of deep vein thrombosis: evaluation by computed tomography pulmonary angiography and computed tomography venography.

PURPOSE: The aim of this study was to evaluate the relation between the sites of pulmonary embolism (PE) and deep vein thrombosis (DVT) by computed tomography pulmonary angiography (CTPA) and CT venography (CTV) of the pelvis and lower extremities. MATERIALS AND METHODS: We retrospectively reevaluated CTPA-CTV data sets for 227 consecutive patients suspected of having a PE. The PEs were divided into proximal (located at the lobar artery or proximal to it) and distal groups. DVTs were divided into proximal (located above the knee) and distal groups. Cohen's kappa statistic and chi-squared tests were performed. RESULTS: The incidence of PE was significantly higher in patients with a proximal DVT than with a distal DVT (P < 0.01). In patients with a proximal DVT, the incidence of proximal PE was significantly higher than that of distal PE (P < 0.05). In patients with a proximal DVT, the incidence of PE was significantly higher in patients with a right-side DVT than with a left-side DVT (P < 0.05). CONCLUSION: Proximal PEs were correlated with proximal DVTs. Patients with a proximal DVT tended to have a PE, especially with a right-proximal DVT. Hence, the presence of a right-proximal DVT has the potential for serious complications, and carefully diagnosis is required for PE and DVT.

Relationship between heart rate and optimal reconstruction phase in dual-source CT coronary angiography.

RATIONALE AND OBJECTIVES: To evaluate reconstruction image quality at the systolic and diastolic cardiac phases and determine the optimal phase for reconstruction according to heart rate when using dual-source computed tomography (CT) with 75 ms temporal resolution. MATERIALS AND METHODS: We retrospectively reviewed the CT datasets of 35 patients with regular heartbeats who underwent coronary CT angiography. Images were reconstructed in 2% steps between 32 and 78% of the beat-to-beat interval. Two experienced radiologists determined the reconstruction interval with the fewest motion artifacts and the motion score of each vessel for the systolic and diastolic phases. Subgroup analysis was performed in patients having heart rates of <70, 70-80, and >80 beats per minute (bpm). RESULTS: In the subgroup with heart rates of <70 bpm, the diastolic phase reconstruction image quality was significantly better than for the systolic phase (P < .01). In the 70-80 bpm and >80 bpm subgroups, no significant difference was observed. In the diastolic phase, the image quality of the <70 bpm subgroup was significantly better than for the >80 bpm subgroup (P < .05). In all systolic phase subgroups and other diastolic phase subgroups, no significant difference was observed. CONCLUSIONS: Using a DSCT scanner with 75 ms temporal resolution, reconstruction at the diastolic phases should be used for patients with heart rates <70 bpm. For heart rates >70 bpm, larger studies are necessary to determine whether reconstruction at the systolic, diastolic, or both phases should be used.

Multidetector computed tomography venography: optimum dose of contrast material.

PURPOSE: The aim of this study was to determine the optimum dose of contrast material for evaluating veins in the lower limbs. MATERIALS AND METHODS: A total of 134 patients who underwent multidetector computed tomography (MDCT) due to suspected deep vein thrombosis (DVT) or pulmonary embolism (PE) were included in this study. One hundred milliliters of iopamidol 100 ml, with 370 mg I/ml was administered. The degree of contrast enhancement of veins in the lower limbs was classified on a 4-point scale (grade 1, poor --> 4, excellent). Regions of interest (ROIs) were positioned in the femoral vein and the popliteal vein to measure CT numbers in these veins. Correlations between the CT number in each ROI and body weight were examined. RESULTS: The mean +/- SD body weights of patients by contrast-enhancement grade were as follows: grade 1, 86.3 +/- 10.2 kg; grade 2, 72.6 +/- 10.7 kg; grade 3, 59.7 +/- 8.7 kg; grade 4, 51.3 +/- 7.9 kg. Negative correlations were found between body weight and CT number for both the femoral vein and the popliteal vein. Grade 3 or better contrast enhancement was obtained in 79 of 81 patients (97.5%) weighing <60 kg. CONCLUSION: For patients weighing <60 kg, 100 ml of contrast material (370 mg I/ml) is considered sufficient for evaluating veins in the lower limbs.