Synth.AI
A Trusted AI Co-Pilot for Clinicians

Our comprehensive suite of AI clinical decision support and workflow tools assists and expands the capabilities of clinicians to consistently detect abnormalities in radiology images from various modalities. Synth.AI provides predictive insight, increased workflow efficiency, and superior outcomes for patients and healthcare providers. With Synth.AI as a trusted partner in diagnosis and reporting, every clinician, regardless of environment, can provide world-class care.

Scalable Intelligence for Frontline Clinicians

We expand the capabilities of clinicians by providing critical information at the point of care, regardless of the clinical environment. This instant second opinion is distilled from the combined knowledge of a vast, ever-growing network of radiologists through the power of machine learning.

Revolutionary Quality Assurance Capabilities

Our tools will cover the entire patient journey, providing critical insight from patient acquisition to reporting and follow-ups, to enabling better clinician training, oversight, and patient monitoring.

Intuitive Integration

Synth.AI integrates seamlessly into existing clinical information systems with no disruption to workflow; outputs are available within any standard radiology viewer. Synth.AI is integrated with Synth.OS, and we’ve partnered with leading PACS manufacturers to bring these solutions to healthcare providers around the world.

Unparalleled AI Transparency

We put the power in the hands of clinicians with our Confidence Dial, allowing dynamic control of algorithmic sensitivity. Physicians are provided critical information at the right time, so they can decide what’s important.

Advanced Diagnostic Support and Workflow Tools

Synth.AI weaves together image analysis and natural language processing to create a broad set of vendor-agnostic machine learning tools that can be easily integrated into any workflow without the need for a separate workstation.

Responsible AI

Our tools are built on robust scientific principles, utilizing professionally annotated, demographically relevant data, and are validated using the highest industry standards. Synthesis has established a stringent, ISO 13485 certified process for creating, testing, and certifying every tool we create.

Synth.AI Chest (formerly XrAI)

The first AI tool for radiology authorized by Health Canada as a Certified Medical Device.

Synth.AI Chest is a clinical decision support and quality-control tool that uses machine learning to classify and highlight regions of abnormality in the lungs and pleural spaces. It provides immediate analysis of chest radiographs, enhancing the clinician’s ability to diagnose quickly and accurately.

How it works:

PACS-integrated operation – Synth.AI Chest is integrated directly into the PACS viewer, automatically identifying regions of interest in the image.

Confidence Dial – A unique feature of Synth.AI Chest is the Confidence Dial, which places control in the hands of clinicians. The Confidence Dial allows clinicians to control the output of the algorithm using an intuitive interface, provisioning unprecedented transparency into the performance of the algorithm.

Geographically and demographically diverse data – Our algorithm is built upon longitudinal, multi-source, population-wide data.

Synth.AI FollowUp

A ground-breaking, intelligent follow-up management tool.

Synth.AI FollowUp is a quality-assurance tool that automatically detects recommendations for follow-up in radiologists’ reports. It provides an easy, intuitive way to minimize lost follow-up recommendations, ensuring clinics can optimize operations and patients receive the care they need.

How it works:

Natural Language Understanding (NLU) – Synth.AI FollowUp uses neural-net based models for understanding the context of the text being processed. NLU is able to analyze and understand the entire radiology report and produce accurate results that help clinicians provide timely, efficient care

Intelligent follow-up detection – Combining binary classification and entity-extraction models for follow-up identification, Synth.AI FollowUp detects whether a radiologist recommended a follow-up, the type of exam, the part of the body concerned, and the recommended date for the exam.

Streamlined workflows – Synth.AI FollowUp is designed to seamlessly integrate into existing PACS/RIS workflows, allowing clinics to identify follow-up recommendations and coordinate scheduling all in one place to maximize operational efficiency.

Improved patient trust – By minimizing scheduling and follow-up errors, clinics can reduce the risks of delayed patient diagnosis and build lasting, trusted relationships with their patients.