Researchers at Stanford Release A Large Pediatric Echocardiography Video Dataset for Computer Vision Research

The most important frontline diagnostic tool for heart disease in the United States is echocardiography since it is portable, effective, and non-invasive while still giving high-quality imagery. The most often used and easily accessible imaging modality for evaluating and treating congenital and acquired heart disease in children is pediatric echocardiography, also known as cardiac ultrasound. A wide range of pediatric disorders, including those managed by arrhythmia, heart failure, post-surgical ventricular function, genetic abnormalities, and acquired heart disease, require assessment of left ventricular function to track disease development and focus treatment.

Researchers also presented a new huge video dataset of echocardiograms for computer vision research in addition to the deep learning model. Patients with a wide range of sizes and ages (0–18 years) are included in the database (43% are female). The 7,643 annotated echocardiography films in the EchoNet-Peds database, along with the measurements, tracings, and calculations made by human experts, serve as a starting point for research on heart motion and chamber sizes.

As proven by EchoNet-Dynamic, machine learning has shown tremendous promise in adults, particularly in its capacity to enhance the evaluation of left ventricular function. However, the difficulty in gathering big datasets and the rarity of congenital and acquired cardiac disease sometimes restrict research in kids. Children’s environments differ from adults’ in several ways, including a wider range of size, age, heart rate, and patient participation, all of which are known to affect the quality of the image. In addition to the absence of open databases that enable cooperation, guarantee reproducibility, and build infrastructure to support generalizability without bias, they have somewhat constrained machine learning in pediatric cardiology. Furthermore, whether models that have only been trained on adult data will be suitable for juvenile patients is debatable. By training and verifying a model only on pediatric data and adding echocardiographic images relevant to the pediatric community for the assessment of left ventricular function, researchers set out to expand on the work of EchoNet-Dynamic with EchoNet-Peds. Researchers offer the EchoNet-Peds dataset of 7,643 echocardiography videos, covering the range of typical echocardiography lab imaging acquisition conditions, with corresponding labeled measurements such as ejection fraction, left ventricular volume at end-systole and end-diastole, and human expert tracings of the left ventricle as a help for researching machine learning methods to assess cardiac function. Researchers also demonstrate the model’s ability to classify videos using a 3-dimensional convolutional neural network design. This model serves as a standard for future collaboration, comparison, and the development of task-specific architectural designs. It is used to semantically segment the left ventricle, evaluate ejection fraction to judge human performance, and segment the left ventricle. This is the largest tagged pediatric medical video collection that has been made openly available to researchers and medical specialists.

Dataset

  • Echocardiogram Movies: An average full resting echocardiogram exam comprises several videos and pictures showing the heart from various perspectives using various image-acquisition methods. The dataset includes 4,424 parasternal short-axis echocardiography recordings and 3,176 apical-4-chamber echocardiography videos from patients who received imaging procedures between 2014 and 2021 as part of routine clinical care at Lucile Packard Children’s Hospital at Stanford. Each video was cropped and masked to exclude text and material outside of the scanning sector. The generated images were then cubic interpolated and downsampled into uniform 112×112 pixel films.
  • Measures: Each research is linked to clinical measurements and calculations made by a registered sonographer and confirmed by a skilled physician echocardiographer by the usual clinical process and the video itself. The left ventricular ejection fraction used to diagnose cardiomyopathy, evaluate eligibility for specific chemotherapy treatments, and establish the indication for medical devices, is a key indicator of cardiac function. The “Bullet method” or the “5/6 Area Length Method,” which derives the left ventricular volumes from the apical and parasternal short axis views, is the ejection fraction method used for the pediatric dataset. The ratio of left ventricular end-systolic volume (ESV) to left ventricular end-diastolic volume (EDV), which is calculated as (EDV – ESV) / EDV, is the ejection fraction, which is given as a percentage.

The way doctors identify and track children’s cardiac issues has the potential to change with the application of AI in medical imaging. EchoNet-Pediatric can lessen the strain on cardiologists and increase the precision and consistency of diagnoses by automating the image processing process. This can assist in ensuring that kids get the best care available and can enhance patient outcomes.

However, it’s critical to remember that the application of AI in medicine is still in its infancy, and further study is required to identify its advantages and disadvantages. It’s also crucial to ensure AI systems like EchoNet-Pediatric are used ethically and responsibly and are subject to the proper oversight and regulation.

Challenges

  • Ensuring the accuracy and dependability of the results is one of the major difficulties with employing AI in medical imaging. Deep learning models like EchoNet-Pediatric may need to be corrected, especially if they were trained on a small or skewed dataset or applied outside the context in which they were developed. This means that AI systems like EchoNet-Pediatric must undergo substantial validation and testing before they can be deployed in a clinical setting.
  • Another difficulty is ensuring AI is used in medical imaging ethically and legally. For instance, there are worries about data ownership and privacy and the possibility that AI systems could be utilized in ways that discriminate against specific groups or have unforeseen repercussions. Clear rules and regulations must be established to control the usage and development of AI systems like EchoNet-Pediatric to guarantee their ethical and responsible use.

Key factors:

  • The first pediatric-trained AI model to evaluate ventricular function is EchoNet-Peds.
  • Similar to human specialists, EchoNet-Peds calculates ejection fractions with accuracy.
  • Many pediatric ages and sizes can use EchoNet-Peds with speed and consistency.
  • Using pediatric training data, EchoNet-Peds performed better than adult models.

EchoNet-Pediatric, which has the potential to enhance the diagnosis and treatment of pediatric cardiac diseases dramatically, is a fascinating example of how AI might be applied to medical imaging. To guarantee that these systems are used responsibly and ethically, it is crucial to approach AI in medicine cautiously.


Check out the Paper and Project Page. All Credit For This Research Goes To the Researchers on This Project. Also, don’t forget to join our 14k+ ML SubRedditDiscord Channel, and Email Newsletter, where we share the latest AI research news, cool AI projects, and more.

Dhanshree Shenwai is a Computer Science Engineer and has a good experience in FinTech companies covering Financial, Cards & Payments and Banking domain with keen interest in applications of AI. She is enthusiastic about exploring new technologies and advancements in today’s evolving world making everyone's life easy.

🐝 Join the Fastest Growing AI Research Newsletter Read by Researchers from Google + NVIDIA + Meta + Stanford + MIT + Microsoft and many others...