Cardiac Diagnosis Research Articles

Cardiac Amyloidosis Imaging, Part 2: Quantification and Technical Considerations.

99mTc-pyrophosphate imaging has been around for a long time. In the 1970s, it was used to image recent myocardial infarction. However, it has recently been recognized for its value in detecting cardiac amyloidosis, leading to widespread use across the United States. Increased use led to considerable procedure variability. As the evidence base to support formal guidelines was being developed, experts from several professional medical societies issued imaging and interpretation recommendations titled "ASNC/AHA/ASE/EANM/HFSA/ISA/SCMR/SNMMI Expert Consensus Recommendations for Multimodality Imaging in Cardiac Amyloidosis: part 1 of 2-Evidence Base and Standardized Methods of Imaging." To reach a consensus on a protocol that would benefit the bulk of laboratories, the experts considered several parameters and radiotracer kinetics. The most critical parameters concerned injection-to-imaging delay and planar imaging versus SPECT. Accordingly, the standardized protocol recommends the injection of 370-740 MBq (10-20 mCi) of 99mTc-pyrophosphate with imaging 3 h later. Planar images of the chest are acquired in the anterior and lateral views accompanied by SPECT images. Both the planar and the SPECT images are used to semiquantitatively grade the degree of myocardial uptake compared with the amount of uptake in the ribs using a 0-3 scale. A grade of 2 or 3 on the SPECT images is considered positive for cardiac amyloidosis. The planar images are used to calculate a heart-to-contralateral-lung ratio. A ratio greater than 1.3 at 3 h helps to confirm the diagnosis of cardiac amyloid if the SPECT images have positive findings. This article is part of a 3-part series in this issue of the Journal of Nuclear Medicine Technology Part 1 details the etiology of cardiac amyloidosis and 99mTc-pyrophosphate imaging acquisition parameters. Part 2, this article, describes the procedure evolution over 50 y, image processing, and quantification. It further discusses radiotracer kinetics and 2 important technical considerations: injection-to-imaging delay and planar imaging versus SPECT. Part 3 covers study interpretation along with cardiac amyloidosis diagnosis and treatment.

Subcutaneous mechano-electrocardiogram (MECG) sensor for complementary cardiac diagnosis

Since the heart pumps out the blood through the excitation-contraction coupling, simultaneous monitoring of the electrical and mechanical characteristics is beneficial for comprehensive diagnosis of cardiac disorders. Currently, these characteristics are monitored separately with electrocardiogram (ECG) and medical imaging techniques. This work presents a fully implantable device named mechano-electrocardiogram (MECG) sensor that can measure mechanocardiogram (MCG) and ECG together. The key to the success is fabrication of permeable electrodes on a single low-modulus porous nanofiber mat, which helps immediate adhesion of the sensor on the tissue. A strain-insensitive electrode is used as the ECG electrode and a strain-sensitive electrode is used for MCG. The MECG device is implanted subcutaneously in the skin above the heart of the rat. Through a vasopressor (phenylephrine) injection test, the MECG signals indicate that the MCG amplitude is related with blood pressure and the ECG peak interval is more related with heart rate. These results confirm that the MECG device is clinically meaningful for continuous and comprehensive monitoring of the electrical and mechanical characteristics of the heart.

Toward Right Ventricle Segmentation in Cardiac MRIs via Feature Multiplexing and Multiscale Weighted Convolution.

Cardiovascular diseases are the leading cause of mortality, and accurate segmentation of ventricular regions in cardiac magnetic resonance images (MRIs) is crucial for diagnosing and treating these diseases. However, fully automated and accurate right ventricle (RV) segmentation remains challenging due to the irregular cavities with ambiguous boundaries and mutably crescentic structures with relatively small targets of the RV regions in MRIs. In this article, a triple-path segmentation model, called FMMsWC, is proposed by introducing two novel image feature encoding modules, i.e., the feature multiplexing (FM) and multiscale weighted convolution (MsWC) modules, for the RV segmentation in MRIs. Considerable validation and comparative experiments were conducted on two benchmark datasets, i.e., the MICCAI2017 Automated Cardiac Diagnosis Challenge (ACDC), and the Multi-Centre, Multi-Vendor & Multi-Disease Cardiac Image Segmentation Challenge (M&MS) datasets. The FMMsWC outperforms state-of-the-art approaches, and its performance can approach that of the manual segmentation results by clinical experts, facilitating accurate cardiac index measurement for the rapid assessment of cardiac function and aiding diagnosis and treatment of cardiovascular diseases, which has great potential for clinical applications.

20-Year Audit of Community Issued Automated External Defibrillators, Auckland City Hospital 2002–2022

Early community defibrillation has been shown to improve survival in patients with cardiac arrest. Providing home automated external defibrillators (AEDs) to patients who do not meet guidelines for an implantable cardioverter defibrillator (ICD) implant, have significant contraindications for an ICD implant, or do not wish to have an ICD implanted is a service delivered by some hospitals. This was a retrospective audit of patients issued an AED by Auckland City Hospital 2002–2022. A telephonic survey of active patients was undertaken. Over the 20-year period, 34 patients were issued AEDs. Mean age was 18.5 years with 58.5% aged ≤15 years at date of issue; 58.8% were NZ European and 14.7% Māori. The predominant cardiac diagnosis was Long QT syndrome (41.2%) followed by hypertrophic cardiomyopathy/hypertrophic obstructive cardiomyopathy (23.5%). Twenty-one patients still have AEDs. Of four patients who had multiple prior cardiac implantable electronic device (CIED) implants and lead extractions, one proceeded to a heart transplant, and one died. Six patients proceeded to CIED implants: three ICDs and three implantable loop recorders (ILRs). No surveyed patients required AED therapy. All surveyed patients/caregivers reported that home AED gives them peace of mind. Issuing of home AEDs is not standard clinical practice, but for specific patients who do not meet guidelines, are unwilling to have or have contraindications to an ICD, it provides peace of mind. No surveyed patients required AED therapy.

Cardiac screening: physician-led or a score based approach?

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): Research innovation and development trust Beating Hearts Malta. Introduction Sudden cardiac death (SCD) is an important cause of morbidity in young individuals. Cardiac screening may help identify high risk subjects and enable physicians to implement various early preventive strategies. However, geographical limitations and physician related expenses hinder the feasibility of widespread screening. Objective The objective of this study was to compare a physician-led assessment with a score-based approach. Methodology Students attending Form 5 (2017/2018 scholastic year) were invited to undergo cardiac screening. The screening protocol consisted of a questionnaire, ECG and an onsite consultation with a physician (Approach 1). The questionnaire included symptomatology, family history and athletic ability. Subjects were referred based on symptoms or red flags in the family history or ECG. Questions from the validated Sports Cardiology British Columbia (SCBC) questionnaire were also included in the questionnaire.[1] The SCBC score was calculated retrospectively and a score ≥7 would have required physician referral. Those with red flags in the family history or ECG would also have been referred (Approach 2). The referral, diagnostic rates and costs were compared for both approaches. Results 2672 students gave consent to undergo cardiac screening (mean 15.0±0.3 years, 50.4% female, 95.85 Caucasian). Patients who were under follow-up for known heart disease were excluded (n=9). 9 subjects were diagnosed with a condition linked to SCD (n=5 WPW, n=2 coronary anomalies, n=1 HCM, n=1 LQTS). A physician led screening method (Approach 1) led to 109 (4.1%) referrals. The commonest reason for referral was an abnormal ECG (n=99, 3.7%) in isolation or in combination with symptoms/family history (Pie Chart). ECG was far more sensitive (88.9% [95% CI, 50.7% to 99.4%]) compared to symptoms (22.2% [95% CI, 3.9% to 59.8%]). Symptoms and ECG had comparable specificity (99.3% [95% CI, 98.9% to 99.5%] vs 96.5% [95% CI 95.7% 97.2%]). Approach 2 would have led to more referrals (14.3% (n=382), p<0.001)) compared to a physician-led approach. 294 (n=11.0%) had a SCBC questionnaire score of ≥7, most (n=277, 10.4%) did not have other red flags (family history or ECG). The sensitivity of this approach was however comparable to ECG (88.9% [95% CI, 50.7% to 99.4%]). Specificity was also high (89.4% [95% CI, 88.2% to 90.5%]). One individual would have been missed, resulting in a lower diagnostic yield (2.2% vs 8.3%, p=0.005). Having a two-tier approach (SCBC and ECG initially) led to a 49.4% cost reduction. The total cost of the screening program would drop to €70,079.95, equivalent to €26.32/individual screened. The cost per cardiac diagnosis linked to SCD would drop to €7,786.66 per cardiac diagnosis. Conclusion A two-tier approach leads to substantial reductions in cost, with no negative impact on sensitivity and specificity. The impact of score-based questionnaires should be evaluated in larger cohorts.

Diagnostic yield in national team football players

Abstract Funding Acknowledgements Type of funding sources: Private company. Main funding source(s): Malta Football Association. Introduction Sudden cardiac death (SCD) is an important cause of morbidity in competitive athletes. Pre-participation evaluation (PPE) is mandatory for most sporting organisations. The ECG plays a very important diagnostic role and is included in most screening algorithms. PPE for athletes ≥14 years registered with the National Football Association is now mandatory. Study objectives: The objective of this study was to assess the diagnostic yield of PPE in a group of elite Maltese football players competing at national level. Methodology Players selected at national level (years 2020-2022) underwent PPE (history, physical examination, and resting ECG). Senior team players also underwent echocardiography (biannually) as per regulations. Data was tabulated and analyzed in Microsoft Excel. Results 300 athletes (69% male) were selected in this timeframe. 10.6% (n=32) required referral for secondary evaluation, equating to 1 in 11 athletes screened. Referrals were more frequent in male vs female athletes (22.1% vs 10.2%, p=0.005). Most referrals were from the men’s senior’s squad (22.5%), followed by the men’s U17 squad (21.7%) and 20% of the men’s U15 team. The ECG was the commonest reason for referral to a sports cardiologist (n=23, 71.8%). The commonest abnormal ECG patterns included anterior TWI (n=9, 3.0%), inferior/lateral TWI (n=6, 2.0%) and WPW (n=2, 0.7%). 108 (36.0%) had echocardiography as a first line screening modality. This was abnormal in a fifth of those referred (n=7, 21.9%), 4 of these had a normal ECG. 3 (1.0%) athletes were diagnosed with a condition linked to SCD (n=2 WPW, n=1 HCM). Another 5 (1.7%) athletes were diagnosed with a cardiac disorder (n=1 left ventricular outflow tract ectopy, n=1 dilated aortic root, n=1 aortic regurgitation, n=1 atrial septal defect, n=1 mitral regurgitation). Echocardiography doubled the diagnostic yield (1.3% vs 2.7%), with 3 athletes having otherwise a normal history, physical examination and ECG. 1 in 38 athletes (2.7%) were given a cardiac diagnosis, 1 in 100 athletes (1.0%) were labelled with a condition linked to SCD. Conclusion The referral rate is comparable to that reported in international screening programs, higher in males. The ECG was the commonest reason for referral. The overall diagnostic yield equates to 2.7%, while 1.0% were given a diagnosis linked to SCD. The higher diagnostic yield in this population highlights the importance of raising awareness about heart disease and PPE in young athletes.

Does optimized daytime/nighttime programming reduce insertable cardiac monitor event burden?

Abstract Funding Acknowledgements Type of funding sources: Private company. Main funding source(s): Boston Scientific. Background Insertable Cardiac Monitors (ICM) are widely used for long term cardiac arrhythmia monitoring and diagnosis. ICMs can generate episodes (and associated alerts) that are non-actionable from a clinical standpoint leading to alert fatigue. In particular, non-actionable bradycardia and pause events may occur at nighttime due to reduced sympathetic tone during sleep. We have used stored auto-detected events to better understand optimal programming with nighttime settings to eliminate potentially clinically non-actionable episodes and improve clinic workflow. Objective Evaluate the reduction in clinic episode review burden in bradycardia (brady) and pause episodes with a new daytime/nighttime programming optimization in LUX-Dx ICM. Methods We analyzed data from 2,555 patients implanted with LUX-Dx ICM that included 132,928 detected brady and pause episodes stored remotely. We evaluated optimized settings with separately programmed nocturnal (11p-7a) and nominal daytime (7a-11p) rates and durations as compared to full-time nominal settings (24-hour nominal). Optimized nocturnal programming was set to trigger event storage on a heart rate setting of ≤30 bpm as well as a duration setting of ≥5s for nocturnal brady and pause episodes. The nominal day time and 24-hour settings for brady and pause episodes were HR ≤40bpm and duration ≥3s. The episode burden was analyzed retrospectively using stored episodes comparing optimized versus full-time nominal programming. Results A simulation using the optimized settings showed that 98% and 90% of the nocturnal bradycardia (n = 1909) and pause episodes (n = 1198) had a HR >30bpm, and duration <5s respectively. As a result, the optimized settings for nocturnal episodes reduced 75% and 56% of bradycardia and pause episodes, respectively. (See Figure 1). At a patient level, 63% and 25% of patients had a substantial reduction (>=50%) in brady and pause episodes, respectively. Conclusion The optimized programming settings, which simply adjusts for nocturnal physiology, may substantially improve clinical workflow by reducing potentially clinically non-actionable episodes and associated alerts detected by the LUX-Dx ICM. More clinical correlative studies are needed to understand patient risk.

Characteristics and correlates of disease-related knowledge and exercise self-efficacy among cardiac patients attending virtual cardiac rehabilitation during the first COVID-19 lockdown in Peru

BackgroundGovernment responses and restrictions due to the COVID-19 pandemic (e.g., limits to non-urgent health care services, including non-urgent outpatient appointments) led to the suspension of center-based (in-person) cardiac rehabilitation (CR), with many programs switching to virtual delivery. This study aimed to understand the characteristics and correlates of disease-related knowledge and exercise self-efficacy in a group of patients attending a virtual CR program during the first COVID-19 lockdown in Peru. MethodsIn this prospective observational study, 240 patients receiving virtual CR care (exercise instructions and patient education) between August/2020 and December/2021 completed questionnaires pre- and post-CR assessing disease-related knowledge (CADE-Q SV questionnaire) and self-efficacy (SE; Bandura's Exercise SE scale). Paired t tests were used to investigate changes pre/post-CR and Pearson correlation coefficients were used to determine the association between knowledge/SE and patients’ characteristics. ResultsParticipants were mainly comprised of men, with a cardiac diagnosis of stable coronary artery disease, who underwent percutaneous coronary intervention or had a known diagnosis of hypertension and with at least one cardiovascular risk factor (95.8%). Mean total knowledge scores improved significantly at post-CR (12.9 ± 2.4 to 15.6 ± 2.0/20; p<0.001), as well as in 4/5 knowledge areas (cardiovascular risk factors, exercise, nutrition, and psychosocial risk; p<0.001). Mean SE scores improved significantly at post-CR (1.9 ± 0.9 to 3.0 ± 0.9/5; p = 0.01). Post-CR knowledge and SE were significantly correlated with cardiac diagnosis and surgical procedures (r = 0.17, p = 0.02 and r = 0.27, p = 0.02, respectively). ConclusionsThe virtual CR program improved disease-related knowledge and SE of cardiac patients during the first months of the COVID-19 pandemic. Post-CR outcomes were correlated with cardiac diagnosis and surgical procedures and more research with other characteristics is warrantied.

Detection of Arrhythmia Heartbeats from ECG Signal Using Wavelet Transform-Based CNN Model

In India, over 25,000 people have died from cardiovascular annually over the past 4 years , and over 28,000 in the previous 3 years. Most of the deaths nowadays are mainly due to cardiovascular diseases (CVD). Arrhythmia is the leading cause of cardiovascular mortality. Arrhythmia is a condition in which the heartbeat is abnormally fast or slow. The current detection method for diseases is analyzing by the electrocardiogram (ECG), a medical monitoring technique that records heart activity. Since actuations in ECG signals are so slight that they cannot be seen by the human eye, the identification of cardiac arrhythmias is one of the most difficult undertakings. Unfortunately, it takes a lot of medical time and money to find professionals to examine a large amount of ECG data . As a result, machine learning-based methods have become increasingly prevalent for recognizing ECG features. In this work, we classify five different heartbeats using the MIT-BIH arrhythmia database . Wavelet self-adaptive thresholding methods are used to first denoise the ECG signal. Then, an efficient 12-layer deep 1D Convolutional Neural Network (CNN) is introduced for better features extraction, and finally, SoftMax and machine learning classifiers are applied to classify the heartbeats. The proposed method achieved an average accuracy of 99.40%, precision of 98.78%, recall of 98.78%, and F1 score of 98.74%, which clearly show that it outperforms with the exiting model . Architecture of proposed work is simple but effective in remote cardiac diagnosis paradigm that can be implemented on e-health devices.

Medical and End-of-Life Decision-Making Preferences in Adolescents and Young Adults with Advanced Heart Disease and Their Parents

Despite high disease morbidity and mortality and complex treatment decisions, little is known about the medical and end-of-life decision-making preferences of adolescents and young adults (AYA) with advanced heart disease. AYA decision-making involvement is associated with important outcomes in other chronic illness groups. To characterize decision-making preferences of AYAs with advanced heart disease and their parents and determine factors associated with these preferences. Cross-sectional survey between July 2018 and April 2021 at a single-center heart failure/transplant service at a Midwestern US children's hospital. Participants were AYAs aged 12 to 24 years with heart failure, listed for heart transplantation, or posttransplant with life-limiting complications and a parent/caregiver. Data were analyzed from May 2021 to June 2022. Single-item measure of medical decision-making preferences, MyCHATT, and Lyon Family-Centered Advance Care Planning Survey. Fifty-six of 63 eligible patients enrolled in the study (88.9%) with 53 AYA-parent dyads. Median (IQR) patient age was 17.8 (15.8-19.0) years; 34 (64.2%) patients were male, and 40 patients (75.5%) identified as White and 13 patients (24.5%) identified as members of a racial or ethnic minority group or multiracial. The greatest proportion of AYA participants (24 of 53 participants [45.3%]) indicated a preference for active, patient-led decision-making specific to heart disease management, while the greatest proportion of parents (18 of 51 participants [35.3%]) preferred they and physician(s) make shared medical decisions on behalf of their AYA, representing AYA-parent decision-making discordance (χ2 = 11.7; P = .01). Most AYA participants stated a preference to discuss adverse effects or risks of treatment (46 of 53 participants [86.8%]), procedural and/or surgical details (45 of 53 participants [84.9%]), impact of condition on daily activities (48 of 53 participants [90.6%]), and their prognosis (42 of 53 participants [79.2%]). More than half of AYAs preferred to be involved in end-of-life decisions if very ill (30 of 53 participants [56.6%]). Longer time since cardiac diagnosis (r = 0.32; P = .02) and worse functional status (mean [SD] 4.3 [1.4] in New York Heart Association class III or IV vs 2.8 [1.8] in New York Heart Association class I or II; t-value = 2.7; P = .01) were associated with a preference for more active, patient-led decision-making. In this survey study, most AYAs with advanced heart disease favored active roles in medical decision-making. Interventions and educational efforts targeting clinicians, AYAs with heart disease, and their caregivers are needed to ensure they are meeting the decision-making and communication preferences of this patient population with complex disease and treatment courses.

Slide deep reinforcement learning networks: Application for left ventricle segmentation

Automatic segmentation of the left ventricle (LV) in four-chamber view images is critical for computer-aided cardiac disease diagnosis. The complex structure of the cardiac image and the encoder-decoder networks may cause coarse segmentation results. High-accuracy LV segmentation is still a challenge with existing automatic LV segmentation methods. In this paper, we propose a slide deep reinforcement learning segmentation network for pixelwise LV segmentation. The main architecture of the slide reinforcement learning networks consists of a slider item combined state, a group of morphology transforming actions and an agent network. The specifically designed reinforcement learning state comprises an image item and a slider item, which contains both original image information and network act information. The reinforcement learning actions proposed in this paper enable accurate and fast formulation of the binary segment result for each frame by controlling the length and location of the slider. Additionally, the confidence branch proposed in our experiment provides a continuous frame series environment, and the identification algorithm avoids losing the segmentation target. The segmentation result reveals that the proposed method outperforms FCN, SegNet, U-Net and TransUnet. The IoU improved by 23.01%, 15.4%, 11.24% and 6.9%. Additionally, we demonstrate how the proposed method can be used as a semisupervised method, which is more convenient for the image annotation process.

Reducing segmentation failures in cardiac MRI via late feature fusion and GAN-based augmentation

Cardiac magnetic resonance (CMR) image segmentation is an integral step in the analysis of cardiac function and diagnosis of heart related diseases. While recent deep learning-based approaches in automatic segmentation have shown great promise to alleviate the need for manual segmentation, most of these are not applicable to realistic clinical scenarios. This is largely due to training on mainly homogeneous datasets, without variation in acquisition, which typically occurs in multi-vendor and multi-site settings, as well as pathological data. Such approaches frequently exhibit a degradation in prediction performance, particularly on outlier cases commonly associated with difficult pathologies, artifacts and extensive changes in tissue shape and appearance. In this work, we present a model aimed at segmenting all three cardiac structures in a multi-center, multi-disease and multi-view scenario. We propose a pipeline, addressing different challenges with segmentation of such heterogeneous data, consisting of heart region detection, augmentation through image synthesis and a late-fusion segmentation approach. Extensive experiments and analysis demonstrate the ability of the proposed approach to tackle the presence of outlier cases during both training and testing, allowing for better adaptation to unseen and difficult examples. Overall, we show that the effective reduction of segmentation failures on outlier cases has a positive impact on not only the average segmentation performance, but also on the estimation of clinical parameters, leading to a better consistency in derived metrics.

Prenatal Congenital Heart Disease—It Takes a Multidisciplinary Village

Prenatal diagnosis of congenital heart disease (CHD) allows for thoughtful multidisciplinary planning about location, timing, and need for medical interventions at birth. We sought to assess the accuracy of our prenatal cardiac diagnosis, and postnatal needs for patients with CHD utilizing a multidisciplinary approach. We performed a retrospective chart review of fetal CHD patients between 1/1/18 and 4/30/19. Maternal and infant charts were reviewed for delivery planning, subspecialty care needs, genetic evaluation, prenatal and postnatal cardiac diagnoses, need for prostaglandin (PGE) and neonatal cardiac intervention. 82 maternal-fetal dyads met inclusion criteria during the study period and delivered at a median of 38w2d gestation. 32 (39%) dyads had CHD and other anomalies or genetic abnormalities. All dyads met with a genetic counselor and neonatologist. 11 patients delivered at outside hospitals as planned (all with isolated CHD not requiring neonatal intervention), and 5 chose a palliative delivery. 30 patients were counseled to expect a neonatal cardiac intervention and 25 (83%) underwent an intervention within the expected time period. No neonates required an uncounseled cardiac intervention. 29 patients planned for PGE at birth and 31 received PGE. Of the 79 postnatal echocardiograms, 60 (76%) were entirely consistent with the fetal diagnosis. A multidisciplinary approach to the prenatal diagnosis of CHD in maternal-fetal dyads is optimal and utilizing this method we were able to accurately predict postnatal physiology and ensure that patients delivered in the correct location with an appropriate supportive structure in place.

Graphene Biointerface for Cardiac Arrhythmia Diagnosis and Treatment.

Heart rhythm disorders, known as arrhythmias, cause significant morbidity and are one of the leading causes of mortality. Cardiac arrhythmias are frequently treated by implantable devices, such as pacemakers and defibrillators, or by ablation therapy guided by electroanatomical mapping. Both implantable and ablation therapies require sophisticated biointerfaces for electrophysiological measurements of electrograms and delivery of therapeutic stimulation or ablation energy. In this work, a graphene biointerface for in vivo cardiac electrophysiology is reported for the first time. Leveraging sub-micrometer-thick tissue-conformable graphene arrays, sensing and stimulation of the open mammalian heart are demonstrated both in vitro and in vivo. Furthermore, the graphene biointerface treatment of atrioventricular block (the kind of arrhythmia where the electrical conduction from the atria to the ventricles is interrupted) is demonstrated. The graphene arrays show effective electrochemical properties, namely interface impedance down to 40Ω cm2 at 1kHz, charge storage capacity up to 63.7mCcm-2 , and charge injection capacity up to 704µCcm-2 . Transparency of the graphene structures allows for simultaneous optical mapping of cardiac action potentials, calcium transients, and optogenetic stimulation while performing electrical measurements and stimulation. The report presents evidence of the significant potential of graphene biointerfaces for advanced cardiac electrophysiology and arrhythmia therapy.

A Segmentation Method Based on PDNet for Chest X-rays with Targets in Different Positions and Directions

To assess the impact of the relative displacement between machines and subjects, the machine angle and the fine-tuning of the subject posture on the segmentation accuracy of chest X-rays, this paper proposes a Position and Direction Network (PDNet) for chest X-rays with different angles and positions that provides more comprehensive information for cardiac image diagnosis and guided surgery. The implementation of PDnet was as follows: First, the extended database image was sent to a traditional segmentation network for training to prove that the network does not have linear invariant characteristics. Then, we evaluated the performance of the mask in the middle layers of the network and added a weight mask that identifies the position and direction of the object in the middle layer, thus improving the accuracy of segmenting targets at different positions and angles. Finally, the active-shape model (ASM) was used to postprocess the network segmentation results, allowing the model to be effectively applied to 2014 × 2014 or higher definition chest X-rays. The experimental comparison of LinkNet, ResNet, U-Net, and DeepLap networks before and after the improvement shows that its segmentation accuracy (MIoU) are 5%, 6%, 20%, and 13% better. Their differences of losses are 11.24%, 21.96%, 18.53%, and 13.43% and F-scores also show the improved networks are more stable.

Interrater Reliability of the 2015 Pediatric Acute Lung Injury Consensus Conference Criteria for Pediatric ARDS

Diagnostic guidelines for pediatric acute respiratory distress syndrome (PARDS) were developed at the 2015 Pediatric Acute Lung Injury Consensus Conference (PALICC). Although this was an improvement in creating pediatric-specific diagnostic criteria, there remains potential for variability in identification of PARDS. What is the inter-rater reliability of the 2015 PALICC criteria for diagnosing moderate-severe PARDS? What clinical criteria and patient factors are associated with diagnostic disagreements? Patients with acute hypoxic respiratory failure admitted from 2016-2021 who received invasive mechanical ventilation were retrospectively reviewed by two PICU physicians. Reviewers evaluated whether the patient met the 2015 PALICC definition of moderate-severe PARDS and rated their diagnostic confidence. Inter-rater reliability was measured using Gwet's Agreement Coefficient (AC1). 37 of 191 encounters had a diagnostic disagreement. Inter-rater reliability was "substantial" (AC1=0.74, 95% CI [0.65 - 0.83]). Disagreements were due to different interpretations of chest radiographs (56.8%), ambiguity in origin of pulmonary edema (37.8%), or lack of clarity if patient's current condition was significantly different from baseline (27.0%). Disagreement was more likely in patients who were chronically ventilated (OR 4.66, 95% CI [2.16-10.08], p<0.001), had a primary cardiac admission diagnosis (OR 3.36, 95% CI [1.18 - 9.53], p=0.02), or underwent cardiothoracic surgery during the admission (OR 4.90, 95% CI [1.60 - 15.00], p=0.005). Reviewers were at least moderately confident in their decision 73% of the time, however were less likely to be confident if the patient had cardiac disease or chronic respiratory failure. The inter-rater reliability of the 2015 PALICC criteria for diagnosing moderate-severe PARDS in this cohort was substantial, with diagnostic disagreements commonly due to differences in chest radiograph interpretations. Patients with cardiac disease or chronic respiratory failure were more vulnerable to diagnostic disagreements. More guidance is needed on interpreting chest radiographs and diagnosing PARDS in these subgroups.

Learning contextual representations with copula function for medical image segmentation

Long-range contextual information modeling plays an important role in medical image segmentation. Existing approaches often ignore local continuous information or the attention missing. In this paper, we propose a novel approach named CouplaNet for efficiently modeling dense 3D contextual dependencies. CouplaNet consists of two core modules, namely Relationship Modelling (RM) and Contextual Aggregation (CA). RM leverages the GCN to model the relationships of elements and yields the long-range contextual representation for each dimension individually. Then CA is introduced to perform the aggregation for the ensemble of the representation of each dimension according to the coupla property in the Sklar’s theorem. We conducted extensive experiments across six medical image segmentation tasks, including nodule segmentation, skin lesion segmentation, liver tumor segmentation, optic disc and cup segmentation, thyroid nodule segmentation, and automated cardiac diagnosis. Through these experiments, we were able to thoroughly evaluate our contributions and demonstrate the effectiveness of our approach.

Impaired Maternal-Fetal Environment and Risk for Preoperative Focal White Matter Injury in Neonates With Complex Congenital Heart Disease.

Background Infants with congenital heart disease (CHD) are at risk for white matter injury (WMI) before neonatal heart surgery. Better knowledge of the causes of preoperative WMI may provide insights into interventions that improve neurodevelopmental outcomes in these patients. Methods and Results A prospective single-center study of preoperative WMI in neonates with CHD recorded data on primary cardiac diagnosis, maternal-fetal environment (MFE), delivery type, subject anthropometrics, and preoperative care. Total maturation score and WMI were assessed, and stepwise logistic regression modeling selected risk factors for WMI. Among subjects with severe CHD (n=183) who received a preoperative brain magnetic resonance imaging, WMI occurred in 40 (21.9%) patients. WMI prevalence (21.4%-22.1%) and mean volumes (119.7-160.4 mm3) were similar across CHD diagnoses. Stepwise logistic regression selected impaired MFE (odds ratio [OR], 2.85 [95% CI, 1.29-6.30]), male sex (OR, 2.27 [95% CI, 1.03-5.36]), and older age at surgery/magnetic resonance imaging (OR, 1.20 per day [95% CI, 1.03-1.41]) as risk factors for preoperative WMI and higher total maturation score values (OR, 0.65 per unit increase [95% CI, 0.43-0.95]) as protective. A quarter (24.6%; n=45) of subjects had ≥1 components of impaired MFE (gestational diabetes [n=12; 6.6%], gestational hypertension [n=11; 6.0%], preeclampsia [n=2; 1.1%], tobacco use [n=9; 4.9%], hypothyroidism [n=6; 3.3%], and other [n=16; 8.7%]). In a subset of 138 subjects, an exploratory analysis of additional MFE-related factors disclosed other potential risk factors for WMI. Conclusions This study is the first to identify impaired MFE as an important risk factor for preoperative WMI. Vulnerability to preoperative WMI was shared across CHD diagnoses.

Association of physician malpractice claims rates with admissions for low-risk chest pain

BackgroundChest pain accounts for 5% of all emergency department visits and accounts for the highest malpractice payout against emergency physicians. To clarify the impact of defensive medicine, we assessed whether admission rates of low-risk chest pain patients are associated with malpractice claims rates. MethodsA cross-sectional time-series analysis of state-year level malpractice claims rates, admission rates for low-risk chest pain (LRCP; requiring ED physician discretion), and admission rates for acute myocardial infarction (AMI; requiring minimal physician judgment for admission, used as a control) from 2008 to 2017 was performed. Admission rates were derived from Optum's deidentified Clinformatics Data Mart Database. LRCP visits were defined by primary ICD-9 or ICD-10 codes of 786.5, R07.9, or R07.89; length of stay of 2 or fewer days; and no previous major cardiac diagnosis and AMI visits with ICD-9 or ICD-10 codes 410, I21.3, or I121.9. Malpractice claims rates (MPCRs) were derived from the National Practitioner Database (NPD). The association between state-year level MPCR and admission rates for LRCP and AMI was estimated using state fixed-effects models. Standardized costs were inflation adjusted and are expressed in US dollar rate as of 2019. ResultsThere were 40,482,813 ED visits during the 10-year study period, of which 2,275,757 (5.6%) were for chest pain, and 1,163,881 met LRCP criteria. Mean age of LRCP patients was 67.8 years, 60.9% were female, and 16.6% were hospitalized, at a mean cost of $17,120. During the same period, 75,266 (0.2%) visits were for AMI, with 87% admitted. The MPCR by state-year varied widely, from 2.6 to 8.6 claims per 100,000 population. A state fixed-effects model showed that an additional physician malpractice claim per 100,000 population was associated with a 3.66% (95% CI 2.02%–5.30%) increase in the admission rate of LRCP. An analogous model showed no association between MPCR and admission rates for AMI (−1.52%, 95% CI −4.06% to 1.02%). ConclusionHigher MPCRs are associated with increased admission rates for LRCP, at substantial cost, which may be attributable to defensive medicine in the ED.

Infants less than or equal to 2.5 kg have increased mortality and worse motor neurodevelopmental outcomes at 2 years of age after Norwood–Sano palliation

ObjectivesIn infants with single-ventricle congenital heart disease, prematurity and low weight at the time of the Norwood operation are risk factors for mortality. Reports assessing outcomes (including neurodevelopment) post Norwood palliation in infants ≤2.5 kg are limited. MethodsAll infants who underwent a Norwood–Sano procedure between 2004 and 2019 were identified. Infants ≤2.5 kg at the time of the operation (cases) were matched 3:1 with infants >3.0 kg (comparisons) for year of surgery and cardiac diagnosis. Demographic and perioperative characteristics, survival, and functional and neurodevelopmental outcomes were compared. ResultsTwenty-seven cases (mean ± standard deviation: weight 2.2 ± 0.3 kg and age 15.6 ± 14.1 days at surgery) and 81 comparisons (3.5 ± 0.4 kg and age 10.9 ± 7.9 days at surgery) were identified. Post-Norwood, cases had a longer time to lactate ≤2 mmol/L (33.1 ± 27.5 vs 17.9 ± 12.2 hours, P < .001), longer duration of ventilation (30.5 ± 24.5 vs 18.6 ± 17.5 days, P = .005), greater need for dialysis (48.1% vs 19.8%, P = .007), and greater need for extracorporeal membrane oxygenation support (29.6% vs 12.3%, P = .004). Cases had significantly greater postoperative (in-hospital) (25.9% vs 1.2%, P < .001) and 2-year (59.2% vs 11.1%, P < .001) mortality. Neurodevelopmental assessment showed the following for cases versus comparisons, respectively: cognitive delay (18.2% vs 7.9%, P = .272), language delay (18.2% vs 11.1%, P = .505), and motor delay (27.3% vs 14.3%, P = .013). ConclusionsInfants ≤2.5 kg at Norwood–Sano palliation have significantly increased postoperative morbidity and mortality up to 2-year follow-up. Neurodevelopmental motor outcomes were worse in these infants. Additional studies are warranted to assess the outcome of alternative medical and interventional treatment plans in this patient population.