Search

Aims The black box nature of artificial intelligence (AI) hinders the development of interpretable AI models that are applicable in clinical practice. We aimed to develop an AI model for classifying patients of reduced left ventricular ejection fraction (LVEF) from 12-lead electrocardiograms (ECG) with the decision-interpretability. Methods and results We acquired paired ECG and echocardiography datasets from the central and co-operative institutions. For the central institution dataset, a random forest model was trained to identify patients with reduced LVEF among 29 907 ECGs. Shapley additive explanations were applied to 7196 ECGs. To extract the model’s decision criteria, the calculated Shapley additive explanations values were clustered for 192 non-paced rhythm patients in which reduced LVEF was predicted. Although the extracted criteria were different for each cluster, these criteria generally comprised a combination of six ECG findings: negative T-wave inversion in I/V5–6 leads, low voltage in I/II/V4–6 leads, Q wave in V3–6 leads, ventricular activation time prolongation in I/V5–6 leads, S-wave prolongation in V2–3 leads, and corrected QT interval prolongation. Similarly, for the co-operative institution dataset, the extracted criteria comprised a combination of the same six ECG findings. Furthermore, the accuracy of seven cardiologists’ ECG readings improved significantly after watching a video explaining the interpretation of these criteria (before, 62.9% ± 3.9% vs. after, 73.9% ± 2.4%; P = 0.02). Conclusion We visually interpreted the model’s decision criteria to evaluate its validity, thereby developing a model that provided the decision-interpretability required for clinical application.

The cost-effectiveness of sodium-glucose cotransporter 2 (SGLT2) inhibitors for chronic kidney disease (CKD) has not been evaluated in Japan, so we analyzed the cost-effectiveness of dapagliflozin, an SGLT2 inhibitor, for CKD stages 3a and 3b.Methods and Results: We used the Markov model for CKD to assess the costs and benefits associated with and without dapagliflozin from a health system perspective. We estimated the incremental cost-effectiveness ratio (ICER), expressed as per quality-adjusted life-years (QALYs). An ICER5 million Japanese yen (JPY)/QALY was judged to be cost-effective. The effect of dapagliflozin on renal and cardiovascular events was based on published clinical trials. In patients with CKD stage 3a, the ICER of dapagliflozin over standard treatment was 4.03 million JPY/QALY gained. With a cost-effectiveness threshold of 5 million JPY/QALY gained, the cost-effectiveness probability of dapagliflozin over standard treatment was 52.6%. In patients with CKD stage 3b, the ICER of dapagliflozin over standard treatment was 0.12 million JPY/QALY gained. The cost-effectiveness probability of dapagliflozin over standard treatment was 75.2%.The results seemed to show acceptable cost-effectiveness when dapagliflozin was used for CKD stage 3b. On the other hand, cost-effectiveness of dapagliflozin for CKD stage 3a was ambiguous, and further validation is needed.

Left ventricular dilatation (LVD) and left ventricular hypertrophy (LVH) are risk factors for heart failure, and their detection improves heart failure screening. This study aimed to investigate the ability of deep learning to detect LVD and LVH from a 12-lead electrocardiogram (ECG). Using ECG and echocardiographic data, we developed deep learning and machine learning models to detect LVD and LVH. We also examined conventional ECG criteria for the diagnosis of LVH. We calculated the area under the receiver operating characteristic (AUROC) curve, sensitivity, specificity, and accuracy of each model and compared the performance of the models. We analyzed data for 18,954 patients (mean age (standard deviation): 64.2 (16.5) years, men: 56.7%). For the detection of LVD, the value (95% confidence interval) of the AUROC was 0.810 (0.801-0.819) for the deep learning model, and this was significantly higher than that of the logistic regression and random forest methods (P0.001). The AUROCs for the logistic regression and random forest methods (machine learning models) were 0.770 (0.761-0.779) and 0.757 (0.747-0.767), respectively. For the detection of LVH, the AUROC was 0.784 (0.777-0.791) for the deep learning model, and this was significantly higher than that of the logistic regression and random forest methods and conventional ECG criteria (P0.001). The AUROCs for the logistic regression and random forest methods were 0.758 (0.751-0.765) and 0.716 (0.708-0.724), respectively. This study suggests that deep learning is a useful method to detect LVD and LVH from 12-lead ECGs.

Asymptomatic or silent atrial fibrillation (AF) has long been a clinical problem due to the incidence of ischemic stroke. A method is needed to predict the development of silent AF before the occurrence of ischemic stroke. This study was focused on the symptoms of AF, especially palpitation, in pacemaker patients. We assessed the hypothesis that absence of palpitation during rapid ventricular pacing could be a predictor of future onset AF being asymptomatic.In this study, we assessed the presence of symptoms during RV pacing and AF symptoms on 145 pacemaker patients at the outpatient clinic by VVI pacing at 120 ppm. The relationship between symptoms during RV pacing and symptom during AF was assessed. The predictive value of absence of symptom during RV pacing on AF being asymptomatic was assessed.Of 145 patients, 74 had previous AF episode. Among the AF patients, absence of symptom during VVI pacing was associated with AF being asymptomatic.Of 145 patients, 71 had no previous AF events. There were 14 patients who had new-onset AF or atrial flutter (AFL) after the device implantation. Four of the 14 patients (28.6%) were symptomatic during first AF/AFL episode, and 10 (71.4%) were asymptomatic during first-onset AF. All ten patients who were asymptomatic during cardiac pacing test were asymptomatic during their initial episodes of AF as well.This study showed that absence of symptoms during rapid ventricular pacing was associated with first-onset AF being asymptomatic.

BACKGROUND Aortic regurgitation (AR) is a common heart disease, with a relatively high prevalence of 4.9% in the Framingham Heart Study. Because the prevalence increases with advancing age, an upward shift in the age distribution may increase the burden of AR. To provide an effective screening method for AR, we developed a deep learning-based artificial intelligence algorithm for the diagnosis of significant AR using electrocardiography (ECG). METHODS Our dataset comprised 29,859 paired data of ECG and echocardiography, including 412 AR cases, from January 2015 to December 2019. This dataset was divided into training, validation, and test datasets. We developed a multi-input neural network model, which comprised a two-dimensional convolutional neural network (2D-CNN) using raw ECG data and a fully connected deep neural network (FC-DNN) using ECG features, and compared its performance with the performances of a 2D-CNN model and other machine learning models. In addition, we used gradient-weighted class activation mapping (Grad-CAM) to identify which parts of ECG waveforms had the most effect on algorithm decision making. RESULTS The area under the receiver operating characteristic curve of the multi-input model (0.802; 95% CI, 0.762-0.837) was significantly greater than that of the 2D-CNN model alone (0.734; 95% CI, 0.679-0.783; p

Background and Aims Radiofrequency ablation (RFA) is a widely accepted, minimally invasive treatment for hepatocellular carcinoma (HCC). This study aimed to develop a machine learning (ML) model to predict the risk of HCC recurrence after RFA treatment for individual patients. Methods We included a total of 1778 treatment-naive HCC patients who underwent RFA. The cumulative probability of overall recurrence after the initial RFA treatment was 78.9% and 88.0% at 5 and 10 years, respectively. We developed a conventional Cox proportional hazard model, and six ML models – including the deep learning-based DeepSurv model. Model performance was evaluated using Harrel's c-index, and was validated externally using the split-sample method. Results The gradient boosting decision tree (GBDT) model achieved the best performance with a c-index of 0.67 from external validation, and it showed a high discriminative ability in stratifying the external validation sample into two, three, and four different risk groups (p Conclusions We developed a novel ML model for the personalized risk prediction of HCC recurrence after RFA treatment. The current model may lead to the personalization of effective follow-up strategies after RFA treatment according to the risk stratification of HCC recurrence.

Microvascular thrombosis is a typical symptom of COVID-19 and shows similarities to thrombosis. Using a microfluidic imaging flow cytometer, we measured the blood of 181 COVID-19 samples and 101 non-COVID-19 thrombosis samples, resulting in a total of 6.3 million bright-field images. We trained a convolutional neural network to distinguish single platelets, platelet aggregates, and white blood cells and performed classical image analysis for each subpopulation individually. Based on derived single-cell features for each population, we trained machine learning models for classification between COVID-19 and non-COVID-19 thrombosis, resulting in a patient testing accuracy of 75%. This result indicates that platelet formation differs between COVID-19 and non-COVID-19 thrombosis. All analysis steps were optimized for efficiency and implemented in an easy-to-use plugin for the image viewer napari, allowing the entire analysis to be performed within seconds on mid-range computers, which could be used for real-time diagnosis.

Background: The transradial approach (TRA) is a less invasive treatment approach for aortoiliac (AI) lesions than the transfemoral or transbrachial approach. However, this approach has some problems for complex lesions, including the procedural difficulty. We used a bidirectional approach combining the TRA with a sheathless technique for femoral artery (FA) puncture to treat complex AI lesions, as a minimally-invasive approach that is almost equivalent to a single-puncture method. Here, we describe representative cases with AI chronic total occlusion in which the combination of TRA and sheathless technique for FA puncture was extremely useful. Case presentation: Case 1 was a 71-year-old man with intermittent claudication (IC). Control angiography showed total occlusion of the left common iliac artery (CIA) ostium to distal external iliac artery (EIA). We used an antegrade TRA, and advanced a guidewire into the subintimal space under intravascular ultrasound (IVUS) guidance. With the retrograde wire passing through a micro-puncture and IVUS-guided parallel wiring, we achieved wire externalization. We successfully deployed two nitinol stents to the entire lesion. Case 2 was a 63-year-old man with IC. Control angiography revealed total occlusion of the right CIA ostium to common femoral artery (CFA) with severe calcification. The antegrade wire reached the distal EIA using the TRA, but was prevented from progressing by the calcified CFA occlusion. A 21-G metal needle was used to penetrate the CFA calcification through the distal true lumen and the wire was inserted into the EIA for wire externalization. We successfully deployed three nitinol stents in the CIA to EIA, and dilated a drug-coated balloon in the CFA with hemostasis of the distal puncture site. In both cases, the retrograde puncture site was hemostatic during the procedure and postoperative bed rest was not required. Conclusions: TRA combined with a sheathless technique from the FA has the potential to treat AI complex lesions in a less invasive manner.

Background Venoarterial extracorporeal membrane oxygenation (VA-ECMO) can be undertaken surgically, but there are various problems with this method. Methods and Case report We removed the arterial cannula of VA-ECMO percutaneously by combining intravascular balloon dilatation and the Perclose ProGlide™ (PP) closure device in a short amount of time and the extent of bleeding was extremely small. Simultaneously, the venous cannula was removed by suturing and manual compression. We report a series of cases of percutaneous removal of VA-ECMO using intravascular balloon dilatation and PP. Conclusions By using this method, VA-ECMO removal was possible in a very short amount of time with minimal invasiveness.

New versions of balloon-expandable and self-expandable valves for transcatheter aortic valve replacement (TAVR) have been developed, but few studies have examined the outcomes associated with these devices using national-level data. This study aimed to elucidate the clinical and economic outcomes of TAVR for aortic stenosis in Japan through an analysis of real-world data.This retrospective cohort study was performed using data from patients with aortic stenosis who had undergone transfemoral TAVR with Edwards SAPIEN 3, Medtronic CoreValve, or Medtronic Evolut R valves throughout Japan from April 2016 to March 2018. Pacemaker implantation, mortality, and health expenditure were examined for each valve type during hospitalization and at 1 month, 3 months, 6 months, and 1 year. Generalized linear regression models and Cox proportional hazards models were used to examine the associations between the valve types and outcomes.We analyzed 7244 TAVR cases (SAPIEN 3: 5276, CoreValve: 418, and Evolut R: 1550) across 145 hospitals. The adjusted 1-year expenditures for SAPIEN 3, CoreValve, and Evolut R were $79 402, $76 125, and $75 527, respectively; SAPIEN 3 was significantly more expensive than the other valves (P.05). The pacemaker implantation hazard ratios (95% confidence intervals) for CoreValve and Evolut R were significantly higher (P.001) than SAPIEN 3 at 2.61 (2.07-3.27) and 1.80 (1.53-2.12), respectively. The mortality hazard ratios (95% confidence intervals) for CoreValve and Evolut R were not significant at 1.11 (0.84-1.46) and 1.22 (0.97-1.54), respectively.SAPIEN 3 users had generally lower pacemaker implantation and mortality but higher expenditures than CoreValve and Evolut R users.

The increasing number of percutaneous endovascular procedures in highly anticoagulated patients has increased the possibility of iatrogenic femoral artery pseudoaneurysm (IFAP). Ultrasound (US)-guided percutaneous thrombin injection is one of the feasible treatments; however, there are concerns about complications such as peripheral embolization. This study was performed to examine the efficacy and safety of treatment of IFAPs using a combination of percutaneous thrombin injection and intravascular balloon inflation. In this retrospective, single-center study, we analyzed 11 patients who developed and were treated for IFAPs from January 2017 through April 2020. The patients were treated with endovascular therapy (EVT) with percutaneous thrombin injection. The technique utilized fluoroscopic guidance to place a balloon at the neck of the IFAP, and the balloon was then inflated to prevent the inflow of blood to the aneurysm. We then performed US-guided thrombin injection. The mean age was 72.36 ± 10.43 years; mean body mass index (BMI) was 25.25 ± 3.18. All patients had hypertension, 72.7% were undergoing hemodialysis, and 54.5% used oral anticoagulant drugs. The mean aneurysm size was 24.34 ± 13.54 mm. The approach was transfemoral in ten patients and transradial in one patient. All procedures were successful, and there were no complications. The mean thrombin dose was 677.3 ± 410.7 IU; the total hemostatic time was 45.4 ± 24.9 min. In conclusion, the combination of percutaneous thrombin injection and endovascular balloon inflation was feasible and safe for the treatment of IFAPs. This technique may contribute to the treatment of IFAPs.

Background: Nationwide data on transcatheter aortic valve implantation (TAVI) and surgical aortic valve replacement (SAVR) in Japan are scarce. Methods and Results: Using a nationwide inpatient database, we analyzed patients undergoing TAVI (n=8,338) or SAVR (n=16,298) due to aortic stenosis between 2014 and 2017. The annual number of TAVI increased rapidly from 2014 to 2017, particularly in older patients. In-hospital deaths were lower and the length of hospital stay was shorter for patients undergoing TAVI than SAVR. Conclusions: TAVI has been penetrating in Japan as an alternative therapeutic option for aortic stenosis and is associated with acceptable clinical outcomes.

Porcelain aorta, defined as extensive calcification of the ascending aorta or aortic arch, is a reported risk factor for embolic stroke during cardiac surgery. However, the time course of the progression of aortic calcification leading to porcelain aorta has not been elucidated. We herein describe a 70-year-old woman who was followed up for systemic lupus erythematosus and antiphospholipid syndrome for approximately 20 years. A routine computed tomography scan revealed progression of ascending aortic calcification to porcelain aorta. The calcification was absent during the preceding 12 years, partial 6 years later, and total after another 3 years. Computed tomography also demonstrated aortic and mitral valve calcification in the development of porcelain aorta. During the 3 years prior to the last admission, annual echocardiography examinations showed progression of calcific aortic stenosis with symptoms. The patient was admitted to our institution for aortic valve replacement. Considering the high risk of perioperative stroke associated with porcelain aorta, transcatheter aortic valve implantation was performed. Postoperative transthoracic echocardiography revealed improvement of the aortic stenosis with no symptoms. The present case revealed aortic calcific progression to porcelain aorta during an approximately 10-year period with deterioration of aortic stenosis within a short time. The aortic and valvular calcification could be attributed to the inflammatory process of systemic lupus erythematosus and antiphospholipid syndrome. The presence of aortic and mitral annular calcification in the serial imaging can provide information on aortic and valvular atherosclerotic progression, which may be modifiable by early steroid-lowering therapy.

This study aimed to clarify the usefulness of the Ikari-curve left (IL) guiding catheter for balloon pulmonary angioplasty (BPA).The current BPA strategy for chronic thromboembolic pulmonary hypertension is dilation of as many branches as possible to normalize hemodynamics and oxygenation. The shape of the guiding catheter is a major factor in achieving this. However, conventional guiding catheters are difficult to introduce into particular branches. The IL guiding catheter may be suitable; however, its utility remains unclear.We retrospectively analyzed 202 consecutive BPA sessions of 40 patients from November 2016 to October 2019 and divided these sessions into two groups: the IL group where the IL guiding catheter was used and the non-IL group where other catheters were utilized. The occurrence of lung injury was determined by the presence of bloody sputum. We compared the rates of successful introduction into target vessels and assessed for the occurrence of lung injury.The average age of enrolled patients was 60.3 ± 14.4 years, with females comprising 65%. There were 99 sessions in the IL group. The median treated branches per session differed between the 2 groups (IL group: 15 versus non-IL group: 10, P < 0.05). The occurrence of lung injury was lower in the IL group (4.0% versus 11.7%, P = 0.07). The IL group had more successful vessel insertions than the non-IL group (78.8% versus 42.7%, P < 0.01).The IL guiding catheter may be introduced into branches that cannot be accessed by conventional guiding catheters.

Background There are still cases that are difficult to treat for femoropopliteal chronic total occlusion (CTO). The Outback® Elite catheter is effective re-entry device to treat such kind of difficult cases, however, it might be difficult to use the Outback® Elite catheter antegradely in cases with severely calcified lesions. In this case, we performed EVT using the Outback Elite® catheter via the retrograde popliteal approach. Case presentation We report a case of a 77-year-old male with end-stage renal disease who presented with pain and cyanosis of his left foot. Control angiography showed total occlusion from the middle of the left superficial femoral artery to the proximal portion of the popliteal artery. The CTO lesion was severely calcified, which prevented the antegrade advancement of any guidewire. Retrograde popliteal puncture was performed with the patient in the supine position. After intentional retrograde subintimal wiring, the Outback® Elite catheter was advanced via the retrograde approach after the identification of a suitable re-entry site using intravascular ultrasound. After wire crossing, one nitinol stent was deployed and sufficient antegrade flow was achieved without any complications. Conclusions Using Outback® Elite from retrograde should be considered in cases where antegrade advancement fails and bidirectional wiring cannot pass through the CTO lesion.

The conventional treatment goal for chronic thromboembolic pulmonary hypertension (CTEPH) is improving hemodynamic status, particularly mean pulmonary artery pressure (mPAP); however, some patients complain of impaired quality of life (QOL) caused by dyspnea on effort, even after hemodynamic status has fully improved. It remains unclear whether treatment for CTEPH can improve patients' QOL or whether quantitative measures of QOL correlate with hemodynamic status. This study quantified QOL among Japanese CTEPH patients and investigated whether balloon pulmonary angioplasty (BPA) improved QOL.We calculated the QOL scores of 45 CTEPH patients using the European Quality of Life Five Dimension scale (EQ-5D). The mean QOL score among 12 of those 45 not involved in any therapy except supplemental oxygen at the time of initial measurement was 0.673 ± 0.251. QOL was measured before and after BPA for 17 patients, and the mean significantly improved (QOL from 0.741 ± 0.195 to 0.802 ± 0.160, p 0.05; mPAP from 33.0 ± 8.4 to 23.4 ± 4.2 mmHg, p 0.05; pulmonary vascular resistance from 441.4 ± 214.2 to 268.4 ± 85.6 dyne/s/cmBPA improved QOL in patients with CTEPH mainly by improving "usual activities," besides the improvement in mPAP.

Awareness, treatment, and control of hypertension are of the utmost importance in conquering stroke and cardiovascular disease. To reduce the global burden of hypertension, the Japanese Society of Hypertension (JSH) established the "JSH Future Plan" based on an increasing need to transform the strategy for combating hypertension. In addition to energizing conventional approaches in basic, translational, and clinical research, the application of rapidly evolving digital health technologies and artificial intelligence to hypertension healthcare and research (digital hypertension) holds promise for providing further insights into the pathophysiology and therapeutic targets and implementing predictive, personalized, and preemptive approaches in clinical practice. With great potential to revolutionize the landscape of hypertension, digital hypertension has some technical, legal, ethical, social, and financial issues to overcome. Given the multidisciplinary framework, digital hypertension requires comprehensive and strategic collaboration among industry, academia, and government to move forward toward the goal of "Future Medicine".

Background Balloon pulmonary angioplasty (BPA) has been performed in patients with inoperable chronic thromboembolic pulmonary hypertension (CTEPH). However, some patients have inoperable CTEPH despite having lesions suitable for surgical treatment. The effectiveness of BPA in such cases is unclear. The aim of this study was to clarify the effectiveness of BPA in these cases. Methods We retrospectively investigated patients with inoperable CTEPH and divided them into two groups: BPA-suitable and BPA-unsuitable groups based on the findings of pulmonary angiography, computed tomography, and perfusion scintigraphy. The BPA-unsuitable group included patients whose lesions are suitable for surgical treatment but who did not undergo the procedure for any specified reason. We analyzed the hemodynamic, respiratory, and functional status of the patients before and after BPA. Results Forty-three consecutive patients with inoperable CTEPH (age, 62.6 ± 13.5 years; 31 women) were included; all of them underwent BPA. There were 10 patients in the BPA-unsuitable group. In all patients, the mean pulmonary artery pressure, pulmonary vascular resistance, arterial oxygen saturation level, and 6-min walking distance significantly improved after BPA (mean pulmonary artery pressure, from 43.3 ± 7.8 mmHg to 23.9 ± 4.7 mmHg; pulmonary vascular resistance, from 924.1 ± 462.2 dynes/s/cm−5 to 319.7 ± 163.8 dynes/s/cm−5; arterial oxygen saturation level, from 89.3 ± 4.3% to 93.4 ± 3.3%; 6-min walking distance, from 370.0 ± 107.4 m to 443.8 ± 101.4 m). Notably, none of the parameters significantly differed between the groups after BPA. Importantly, the amount of lung bleeding did not differ between them. However, several sessions were required in the BPA-unsuitable group (BPA-unsuitable group: six sessions vs. BPA-suitable group: four sessions). Conclusions BPA safely improved the hemodynamic and functional statuses of the patients with CTEPH who are judged as inoperable for any reason despite lesion being suitable for surgical treatment. However, numerous BPA sessions were required in these patients.

Background: Although endovascular therapy is used to treat chronic limb-threatening ischemia, long chronic total occlusion (CTO) is still challenging to treat. Especially in patients with poor run-off below-the-knee (BTK) arteries, it is difficult to perform a retrograde approach, and even guidewire passage may be difficult.Case presentation: We treated two cases of chronic limb-threatening ischemia using our novel extreme antegrade guidewire crossing technique by AnteOwl WR intravascular ultrasound (IVUS)-guided parallel wiring to a BTK artery (EXCAVATOR technique). Case 1 was a 70-year-old man with ulceration of the right toe. The AnteOwl WR IVUS was intentionally advanced into the subintimal space of the posterior tibial artery, and the totally intraplaque route was advanced by IVUS-guided parallel wiring that was successfully passed from the lateral plantar aspect to the true lumen of the digital artery. Case 2 was a 76-year-old woman with rest pain and cyanosis of the right lower limb. Angiography showed total occlusion from the superficial femoral artery to BTK arteries. AnteOwl WR IVUS-guided parallel wiring was repeatedly performed until the distal true lumen of the peroneal artery was reached, and revascularization was successfully achieved via the antegrade approach alone.Conclusions: With its excellent crossable performance, good image quality, and high navigational ability within the CTO, the AnteOwl WR can be used to pass parallel wiring into the distal true lumen for BTK CTO.

Repeated hospitalization for heart failure (HF) is a strong predictor of mortality among HF patients. While recent cardiac electrical implantable devices (CIEDs) can detect worsening HF through remote monitoring1,2, there is no early detection system for HF progression in patients at home without a CIED. We therefore developed an artificial intelligence-based HF detection system that uses single-lead electrocardiograms (ECGs) recorded at home. Our convolutional neural network (CNN) model calculated a novel HF-index from the estimated NYHA grades as a quantitative indicator of HF severity. Retrospective data revealed a strong correlation between HF-indexes and plasma BNP levels (R=0.91).A prospective clinical study confirmed the accuracy of the HF severity judged from the estimated HF-index using a portable single-lead ECG monitor at home.We have thus successfully constructed a novel, at-home HF monitoring system for a portable single-lead ECG device, which enables early detection and early medical intervention in HF.

Many patients with atrial septal defects (ASD) are asymptomatic and undiagnosed during the first few decades of life, but have overt heart failure, arrhythmias, cerebral infarction, and increased mortality in adults with advancing age. To provide a non-invasive, easy-to-use, and effective method for detecting ASD, we aimed to develop and validate a deep learning-based algorithm to diagnose ASD on chest radiographs. The ASD dataset was created from 173 chest radiographs of 74 patients with ASD and 170 chest radiographs of 100 patients without ASD. Convolutional neural network models (VGG16, ResNet50, DenseNet121, and Xception) for diagnosing ASD were pretrained using two different datasets, the large-scale real-world ImageNet dataset and the ChestX-ray14 dataset released by National Institutes of Health, followed by a round of training using the training set of the ASD dataset. Model performance was evaluated by five-fold stratified cross-validation. The best performance in ImageNet pretraining was achieved by ResNet50 model, and the cross-validation area under the curve (AUC) was 0.95, with sensitivity of 0.86, specificity of 0.87, and overall accuracy of 0.87. The best performance in ChestX-ray pretraining was achieved by Xception, and the cross-validation AUC was 0.93, with sensitivity of 0.85, specificity of 0.85, and overall accuracy of 0.85. The diagnostic performances of these models were comparable to those of cardiologists. Gradient-weighted Class Activation Mapping showed that the ImageNet-pretrained model focused on bilateral hilar regions, while the ChestX-ray14-pretrained model focused on areas around cardiac silhouette and lower lung fields. Our deep learning-based algorithms made a diagnosis of ASD on the input chest radiographs with high accuracy, and had potential to help clinicians make accurate diagnosis of ASD from routine chest radiography, leading to improvement of prognosis and quality of life in patients with ASD.

Paclitaxel-eluting technologies improve the clinical outcome of femoropopliteal (FP) occlusive disease. Several studies reported efficacy of the high-dose (nominal paclitaxel density of 3.5 μg/mmThis was a single-center, retrospective study. From June 2018 to November 2020, we compared 41 patients (43 lesions) treated with high-dose DCB and 36 patients (37 lesions) treated with DES. The study period was defined as the period after DCB and DES became available simultaneously at our hospital, when all surviving patients had at least 1 year of follow-up. The primary endpoint was 12-month primary patency. The secondary endpoints were 12-month freedom from: (1) clinically driven target lesion revascularization (CD-TLR), and (2) re-occlusion.Baseline clinical data were comparable between the two groups. Reference vessel diameter was smaller in the DCB group. The mean lesion and occlusion lengths were about the same in both groups. The subintimal angioplasty and bailout stent rate was 0% in the DCB group. The Kaplan-Meier estimate for 12-month primary patency was 92.0% in the DCB group and 87.2% in the DES group (p = 0.47). Freedom from CD-TLR also did not differ significantly between the two groups. The 12-month freedom from re-occlusion rate tended to be higher in the DCB group than in the DES group.High-dose DCB with successful lesion preparation showed 12-month clinical outcomes comparable with DES for CTO of the SFA, even without bailout stents.

Deep learning models can be applied to electrocardiograms (ECGs) to detect left ventricular (LV) dysfunction. We hypothesized that applying a deep learning model may improve the diagnostic accuracy of cardiologists in predicting LV dysfunction from ECGs. We acquired 37,103 paired ECG and echocardiography data records of patients who underwent echocardiography between January 2015 and December 2019. We trained a convolutional neural network to identify the data records of patients with LV dysfunction (ejection fraction < 40%) using a dataset of 23,801 ECGs. When tested on an independent set of 7,196 ECGs, we found the area under the receiver operating characteristic curve was 0.945 (95% confidence interval: 0.936-0.954). When 7 cardiologists interpreted 50 randomly selected ECGs from the test dataset of 7,196 ECGs, their accuracy for predicting LV dysfunction was 78.0% ± 6.0%. By referring to the model's output, the cardiologist accuracy improved to 88.0% ± 3.7%, which indicates that model support significantly improved the cardiologist diagnostic accuracy (P = 0.02). A sensitivity map demonstrated that the model focused on the QRS complex when detecting LV dysfunction on ECGs. We developed a deep learning model that can detect LV dysfunction on ECGs with high accuracy. Furthermore, we demonstrated that support from a deep learning model can help cardiologists to identify LV dysfunction on ECGs.

Introduction: The cost and/or cost-effectiveness for inpatient management according to the gender of attending physicians remain to be elucidated. Hypothesis: The treatment by female cardiologists may be cost-effective. Methods: The hospitalization costs were extracted from the Diagnosis Procedure Combination (DPC) -based payment system. Using the dataset of 7,457 hospitalized patients with cardiovascular diseases in our hospital from 2012 to 2018, we compared the actual cost of inpatient management by female cardiologists with that by male cardiologists. Next, we estimated the cost-effectiveness of inpatient management according to the gender of attending cardiologists. Results: The cost of initial hospitalization per patient was similar between the patients treated by a female middle-grade cardiologist and those treated by a male middle-grade cardiologist ($17,527 ± 14,158, vs. $17,358 ± 15,183, P=0.69). As an analysis on cost-effectiveness, the incremental cost of hospitalization managed by male middle-grade cardiologists was $67 per patient as compared with that by female middle-grade cardiologists. Concordantly, the evaluation on incremental cost-effectiveness ratio per quality-adjusted life year gained showed that the inpatient management by female cardiologists was dominant over that by male cardiologists (less cost and better effect). Conclusions: Inpatient management by female cardiologists was more cost-effective as compared with that by male cardiologists. Physicians’ gender might have a considerable effect on the medical economics.

In recent years, there has been considerable research on the use of artificial intelligence to estimate age and disease status from medical images. However, age estimation from chest X-ray (CXR) images has not been well studied and the clinical significance of estimated age has not been fully determined.To address this, we trained a deep neural network (DNN) model using more than 100,000 CXRs to estimate the patients' age solely from CXRs. We applied our DNN to CXRs of 1562 consecutive hospitalized heart failure patients, and 3586 patients admitted to the intensive care unit with cardiovascular disease.The DNN's estimated age (X-ray age) showed a strong significant correlation with chronological age on the hold-out test data and independent test data. Elevated X-ray age is associated with worse clinical outcomes (heart failure readmission and all-cause death) for heart failure. Additionally, elevated X-ray age was associated with a worse prognosis in 3586 patients admitted to the intensive care unit with cardiovascular disease.Our results suggest that X-ray age can serve as a useful indicator of cardiovascular abnormalities, which will help clinicians to predict, prevent and manage cardiovascular diseases.Chest X-ray is one of the most widely used medical imaging tests worldwide to diagnose and manage heart and lung diseases. In this study, we developed a computer-based tool to predict patients’ age from chest X-rays. The tool precisely estimated patients’ age from chest X-rays. Furthermore, in patients with heart failure and those admitted to the intensive care unit for cardiovascular disease, elevated X-ray age estimated by our tool was associated with poor clinical outcomes, including readmission for heart failure or death from any cause. With further testing, our tool may help clinicians to predict outcomes in patients with heart disease based on a simple chest X-ray.