Deep learning supported echocardiogram analysis: A comprehensive review.

An echocardiogram is a sophisticated ultrasound imaging technique employed to diagnose heart conditions. The transthoracic echocardiogram, one of the most prevalent types, is instrumental in evaluating significant cardiac diseases. However, interpreting its results heavily relies on the clinician's expertise. In this context, artificial intelligence has emerged as a vital tool for helping clinicians. This study critically analyzes key state-of-the-art research that uses deep learning techniques to automate transthoracic echocardiogram analysis and support clinical judgments. We have systematically organized and categorized articles that proffer solutions for view classification, enhancement of image quality and dataset, segmentation and identification of cardiac structures, detection of cardiac function abnormalities, and quantification of cardiac functions. We compared the performance of various deep learning approaches within each category, identifying the most promising methods. Additionally, we highlight limitations in current research and explore promising avenues for future exploration. These include addressing generalizability issues, incorporating novel AI approaches, and tackling the analysis of rare cardiac diseases.

Automatic Diagnostic Tool for Detection of Regional Wall Motion Abnormality from Echocardiogram.

The echocardiogram is an ultrasound imaging modality, employed to assess cardiac abnormalities. The Regional Wall Motion Abnormality (RWMA) is the occurrence of abnormal or absent contractility of a region of the heart muscle. Conventional assessment of RWMA is based on visual interpretation of endocardial excursion and myocardial thickening from the echocardiogram videos. Wall motion assessment accuracy depends on the experience of the sonographer. Current automated methods highly depend on the preprocessing steps such as segmentation of ventricle part or manually finding systole and diastole frames from an echocardiogram. Additionally, state-of-the-art methods majorly make use of images rather than videos, which specifically lack the usage of temporal information associated with an echocardiogram. The deep learning models used, employ highly complex networks with billions of trainable parameters. Further, the existing models used on video data add to the computational intensity because of the high frame rates of echocardiogram videos. We developed a novel deep learning architecture EC3D-Net (Echo-Cardio 3D Net), which captures the temporal information for identifying regional wall motion abnormality from echocardiogram. We demonstrate that EC3D-Net can extract temporal information from even raw echocardiogram videos, at low frame rates, employing minimal training parameter-based deep architecture. EC3D-Net achieves both an overall F1-Score and an Area Under Curve (AUC) score of 0.82. Further, we were able to reduce time for training and trainable parameters by 50% through minimizing frames per second. We also show the EC3D-Net is an interpretable model, thereby helping physicians understand our model prediction. RWMA detection from echocardiogram videos is a challenging process and our results demonstrate that we could achieve the state-of-the-art results even while using minimal parameters and time by our EC3D-Net. The proposed network outperforms both complex deep networks as well as fusion methods generally used in video classification.

Anomalous left atrial chorda and its association with mitral regurgitation.

Anomalous left atrial chorda is associated with mitral regurgitation. A young woman presenting for mitral valve repair with the diagnosis of mid-segment (A2) of anterior mitral leaflet prolapse causing severe mitral regurgitation. Transesophageal echocardiography examination in pre-bypass period showed an anomalous chorda attaching A2 to the left atrial roof, tethering the anterior mitral leaflet toward the atrial wall. Surgical findings confirmed the abnormally attached chordae and an absence of normal chorda of A2 segment. The anomalous chorda was resected and neo-chordae placed between the A2 segment and papillary muscles and annulus strengthened with an annuloplasty ring.

Comparison of efficacy of two different doses of tranexamic acid in prevention of post operative blood loss in patients with congenital cyanotic heart disease undergoing cardiac surgery.

Background: The optimal dose of tranexamic acid in minimizing perioperative bleeding is uncertain. We compared efficacy of two different doses of tranexamic acid in reducing post-operative blood loss and its side effects in patients with congenital cyanotic heart disease undergoing cardiac surgery. Settings and Design: Prospective observational study at a pediatric cardiac center in South India. Methods: Consecutive cyanotic patients undergoing cardiac surgery were divided into groups I and II to receive either 10 mg/kg or 25 mg/kg of tranexamic acid administered as triple dose regime after induction, during cardiopulmonary bypass, and after protamine. Post-operative blood loss at 24 hours, blood component utilization, incidence of renal dysfunction and seizures were compared. Results: Totally, 124 patients were recruited, 62 in each group. The pre-operative variables and cardiopulmonary bypass time were comparable. Patients receiving 25 mg/kg had lower post-operative blood loss compared to patients in lower dose group (8.04 ± 8.89 vs 12.41 ± 19.23 ml/kg/24 hours, P = 0.03). There was no difference in the transfused volume of packed red cells (9.21 ± 7.13 ml/kg vs 12.41 ± 9.23 ml/kg, P = 0.712), fresh frozen plasma (13.91 ± 13.38 ml/kg vs 11.02 ± 8.04 ml/kg, P = 0.19), platelets (9.03 ± 6.76 ml/kg vs 10.90 ± 6.9 ml/kg, P = 0.14) or cryoprecipitate (0.66 ± 0.59 ml/kg vs 0.53 ± 0.54 ml/kg, P = 0.5) in group II and I, respectively. Two patients developed renal dysfunction secondary to low cardiac output in lower dose group. There were no seizures. Conclusions: Tranexamic acid administered at a dose of 25 mg/kg as triple dose regime is associated with lower post-operative blood loss compared to a lower dose of 10 mg/kg in cyanotic patients undergoing cardiac surgery without causing major adverse effects.

Surgical Challenges in Retrieval of an Embolized Transcatheter Valve From the Aorta.

Embolization of a transcatheter aortic valve is a rare complication. Surgery is required if percutaneous retrieval is not possible. We present a case of embolization of an Evolut R device (Medtronic, Minneapolis, MN) into the ascending aorta. This device, owing to its taller profile, presents unique surgical challenges in retrieval with respect to arterial access and cross clamping of the aorta. Prior knowledge of the profile of the various devices is critical to ensure a safe retrieval and smooth conduct of the operation.

Waiting for Godot: A cross sectional survey based analysis of the hydroxychloroquine prophylaxis strategy against COVID-19 in India.

Background: India currently has the second largest burden of infections due to COVID-19. Health Care Worker (HCW) shortages are endemic to Indian healthcare. It should therefore be a huge priority to protect this precious resource as a critical component of the systemic response to this pandemic. Advisories from the Indian Council of Medical Research (ICMR) have focused on using hydroxychloroquine prophylaxis against COVID-19 in at risk HCW. This prophylaxis strategy has no evidence. In further jeopardy there appear to insubstantial attempts to build this evidence as well. In this connection, we commissioned a survey within our Institution to estimate the penetration of hydroxychloroquine (HCQ) use and use this to statistically model the impact of current ongoing studies in India. We also briefly review the literature on HCQ prophylaxis for COVID-19. Design and methods: A structured survey designed using RedCAP application was disseminated among healthcare professionals employed at an academic referral tertiary care centre via online social media platforms. The survey was kept open for the entire month of June 2020. The survey was additionally used to statistically model the size of studies required to comprehensively address the efficacy of HCQ in this setting. Results: 522 responses were received, of which 4 were incomplete. The ICMR strategy of 4 or more doses of HCQ was complete only in 15% of HCW in our survey. The majority of respondents were doctors (238, 46%). Amongst all category of responders, only 12% (n=63) received the full course. A majority of those who initiated the chemoprophylaxis with HCQ turned out to be medical professionals (59/63) with neither nurse nor other categories of healthcare workers accessing the medication. The respondents of our institutional survey did not report any life-threatening side effects. Presuming efficacy as per ICMR modelling for new registry trial on the lines of the published case control study, equal allocation between cases and controls and assuming a RR of 1.3.6, the power of such a study would be very low for n=2000 for event rates from 2.5-12.5%. Conclusion: We report the low penetration of HCQ chemoprophylaxis among the healthcare workers of our institution. We highlight the inherent drawbacks in the study design of current national COVID related trial based on the statistical modelling of our survey results and published literature, and thereby emphasis the need of evidence-based strategies contributing to research policy at national level.