{"id":3390,"date":"2025-10-06T09:35:00","date_gmt":"2025-10-06T14:35:00","guid":{"rendered":"https:\/\/pchp.ca\/?p=3390"},"modified":"2026-04-13T09:40:43","modified_gmt":"2026-04-13T14:40:43","slug":"machine-learning-platform-developed-helps-scale-high-impact-ai-models-at-sickkids-and-chu-sainte-justine","status":"publish","type":"post","link":"https:\/\/pchp.ca\/en\/news\/machine-learning-platform-developed-helps-scale-high-impact-ai-models-at-sickkids-and-chu-sainte-justine\/","title":{"rendered":"Machine learning platform developed helps scale high-impact AI models at SickKids and CHU Sainte-Justine"},"content":{"rendered":"\n

True innovation doesn\u2019t stop at discovery, it scales.  <\/p>\n\n\n\n

At The Hospital for Sick Children (SickKids), that belief shapes the Precision Child Health<\/a> movement\u2014an initiative to diagnose faster, treat smarter and predict better to deliver individualized care. It means going beyond isolated breakthroughs, to design systems that bring precision care within reach for all children, no matter their location.  <\/p>\n\n\n\n

One example is through a machine learning platform called AtriumDB. Developed in-house at SickKids and being deployed at institutions around the world, the platform enables the real-time ingestion and analysis of physiological data, allowing machine learning models to be trained, validated, and safely deployed across institutions without exchanging raw patient data.  <\/p>\n\n\n\n

Recently the platform was deployed at the CHU Sainte-Justine, in Montr\u00e9al, as part of a project enabled, in part, by the Precision Child Health Partnership<\/a>. Now with a shared data infrastructure the two organizations can work together to faster and more effectively scale high-impact AI models.  <\/p>\n\n\n\n

Detecting arrhythmia in real time to treat faster <\/h2>\n\n\n\n
\n

One of the first models to leverage the joint infrastructure is a real-time arrhythmia detection tool developed at SickKids. Designed to identify junctional ectopic tachycardia (JET), a critical but often missed paediatric arrhythmia, the model continuously monitors bedside ECG feeds and alerts clinical teams when it detects concerning rhythms.<\/p>\n\n\n\n

\u201cJET is one of the more common arrythmias, but it can also be hard to detect for clinicians not specifically trained in diagnosing them,\u201d says Azedah Assadi, Cardiac Critical Care Unit Lead Nurse Practitioner, SickKids. \u201cBut when we\u2019re delayed in diagnosing and treating it, patients are at risk of becoming sicker, especially if they already have an underlying heart condition.\u201d <\/p>\n<\/div>

\"JET<\/figure><\/div>\n\n\n\n

\u201cThe good thing is, if we catch it early, it\u2019s easy to address. This model helps by alerting earlier, so teams can react faster to avoid a negative outcome for patients.\u201d<\/p>\n\n\n\n

The model is in a silent trial phase at SickKids and now entering a silent trial phase at the CHU Sainte-Justine, where it will run in the background without influencing care, allowing researchers to validate its accuracy quickly and effectively.<\/p>\n\n\n\n

Scaling to benefit children around the world <\/h2>\n\n\n\n

Beyond the arrhythmia model, having a scalable shared data environment means future models\u2014whether developed at SickKids, the CHU Sainte-Justine or one of the 10 other sites that now have the same environment\u2014can be trained, validated, and deployed in other clinical settings to benefit patients and families.<\/p>\n\n\n\n

\"Drs.<\/figure>
\n

\u201cWe can now develop decision-support systems that not only help us care with greater precision, but also strengthen teamwork across institutions and, ultimately, give children a better chance at recovery,\u201d says Micha\u00ebel Sauthier, Paediatric Intensivist and bioinformatician at CHU Sainte-Justine, and co-lead of the joint project.<\/p>\n\n\n\n

\u201cWe want our work to make a difference,\u201d says Dr. Steven Schwartz, Chief of Critical Care Medicine at SickKids and co-lead. \u201cThe most effective way to do that is to make sure it goes beyond the walls of our own institution. This platform helps us show quickly what\u2019s working, so the next institution can feel confident, and then the next and so on.”<\/p>\n<\/div><\/div>\n\n\n\n

Robert Greer, Technical Director of SKAI<\/a> and lead engineer of the platform, adds: \u201cWe have an amazing wealth of knowledge and talent here at SickKids. It\u2019s our imperative to disseminate that knowledge as widely as possible to improve the lives of children everywhere.\u201d  <\/p>\n\n\n\n

The implementation of the platform at the CHU Sainte-Justine was supported by the Azrieli Precision Child Health Platform, a key pillar of the partnership (PCHP) between SickKids and the CHU Sainte-Justine. The joint work of the PCHP is funded by a $50-million-dollar transformational gift by The Azrieli Foundation, with key contributions of the Pathy Family Foundation and the Breakthrough Fund, along with the generous support of SickKids and CHU Sainte- Justine donors.<\/em> <\/p>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

True innovation doesn\u2019t stop at discovery, it scales.   At The Hospital for Sick Children (SickKids), that belief shapes the Precision Child…<\/p>\n","protected":false},"author":5,"featured_media":3391,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1],"tags":[28,30,26],"class_list":["post-3390","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized","tag-artificial-intelligence","tag-cardiac-health","tag-machine-learning"],"acf":[],"_links":{"self":[{"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/posts\/3390","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/comments?post=3390"}],"version-history":[{"count":3,"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/posts\/3390\/revisions"}],"predecessor-version":[{"id":3395,"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/posts\/3390\/revisions\/3395"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/media\/3391"}],"wp:attachment":[{"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/media?parent=3390"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/categories?post=3390"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/pchp.ca\/en\/wp-json\/wp\/v2\/tags?post=3390"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}