Cloud-based AI being used to improve health outcomes and processes

When it comes to advancing the health of Canadians, there’s no question that artificial intelligence (AI) and machine learning will be driving care innovation in 2020. From personalizing the patient health journey to accelerating precision medicine, AI and machine learning tools are supporting clinicians with data-driven insights to improve health outcomes and the patient experience.

Cloud computing is a key enabler, creating an environment in which it’s easier to experiment with new innovative solutions. By simplifying the path to building and deploying advanced AI, analytics, and machine learning tools, we’re creating a future for Canadian healthcare that thrives on the intelligent use of data to improve care.

A great example of AI supporting physicians can be found at Vancouver General Hospital (VGH). Working with researchers at University of British Columbia, VGH used AWS technology to develop a new machine learning model that speeds diagnosis of pneumonia on chest X-rays and reduces time to treatment.

This AI tool is brilliantly embedded into the clinical workflow. When a patient comes to the VGH Emergency Department with symptoms of pneumonia, the ensuing chest X-ray is immediately analyzed by the AI tool even before a radiologist sees the study.

If the algorithm suspects pneumonia, it flags and escalates the study to the top of the radiologist’s work list. This means the radiologist reviews the films much earlier than if it were simply placed in the worklist in the order it was taken. Quicker review means quicker diagnosis and faster treatment for the patient.

This system was built using two solutions from Amazon Web Services. The first, Amazon Comprehend Medical, makes it easy to use machine learning to extract relevant medical information from unstructured text, like medical charts or doctor’s notes. The second solution, Amazon SageMaker, provides all of the components used for machine learning in a single toolset so models get to production faster, with less effort and at lower cost.

We see AI being applied in three areas to advance care: predicting patient health events, personalizing the health journey and promoting interoperability.

Predicting patient health events: We’re seeing a renaissance in healthcare as organizations leverage machine learning technologies to uncover new ways to enhance patient care, improve health outcomes, and ultimately save lives.

As the industry shifts towards value-based care, AI and machine learning, paired with data interoperability, will improve patient outcomes while driving operational efficiency to lower the overall cost of care. By enabling data liquidity securely and supporting healthcare providers with predictive machine learning models, clinicians will be more readily able to use technology to forecast clinical events like strokes, cancer or heart attacks, opening the door to early intervention.

For example, Cerner can detect Congestive Heart Failure about 15 months ahead of a clinical manifestation – all done with the power of AWS’s machine learning services. Pairing this predictive tool with real-time integration to individual health records can support provider decision making in real-time. Future projects will look to improve pre-procedural decision and interventions for chest pain using a cardiovascular prediction model.

Personalizing the health journey: For many healthcare organizations, creating a frictionless and more personalized experience for patients is top of the list. We are living in a consumer-centric world where our best experience anywhere is what we expect everywhere. For example, Aidoc’s always-on, AI-based, decision-support software analyzes CT scans on AWS to flag acute abnormalities, prioritize urgent studies, and expedite patient care.

To date, Aidoc has analyzed more than 3.2 million cases at more than 300 medical facilities around the world. At one major U.S. medical center, the Aidoc solution reduced patient ED visits by an average of 59 minutes and overall hospitalization time by 18 hours.

Promoting interoperability in healthcare: Most electronic health record systems (EHRs) do not follow patients on their care journey beyond the hospital or clinic walls. As a result, only a portion of healthcare data is available at any point of care, resulting in a fragmented view of a patient’s health history. One of the biggest barriers right now is that most health and patient data is stored in an unstructured medical format, and identifying this information is a manual and time-consuming process. AI has the power to break down this barrier, improving the patient experience. With access to all available information, advanced analytics and machine learning can enhance medical and scientific insights tied to patient outcomes in an accurate, scalable, secure, and timely manner.