Annalise Enterprise

Annalise-AI Pty LtdAustralia
2023 People’s Choice Award
Vote for this Project
Voting closes at 23:59 EST (EST time: UTC-5) on 5 May 2023.


Production / Professional


Aengus Tran, CEO Nic Carr, Head of Product. Michelle Gardener, Product Manager. Marc Nothrop, UX Research & Design. Clinical, AI, Engineering, Testing/Validation etc. teams


Medical imaging (radiology) has long been a critical diagnostic tool, playing an ever-growing role in establishing which treatment approach best serves each patient. 

The ongoing growth in the sheer number of medical images challenges radiologists and healthcare providers to read more scans during each workday, all while maintaining the accuracy, quality, and completeness of their results. The result? Well-resourced medical systems face significant imaging backlogs. Under-resourced systems with low clinician-to-patient ratios may miss opportunities for early disease detection. 

AI can assist radiologists in critical triage and diagnostics, acting as “a second set of eyes” in high-pressure, stressful conditions, helping achieve the best patient outcomes. Our advanced machine learning algorithms and software interface fits into their workflow, helping them treat critical patients sooner.

At, our vision is to empower medical systems with comprehensive medical imaging AI, helping radiologists provide higher quality, more accurate, and efficient diagnoses, creating a more equitable healthcare future for every human. 

We trained our AI models on almost 1 million unique, de-identified patient studies labeled by ~150 consultant radiologists, producing hundreds of millions of individual data points. AI tools serve to improve the accuracy of image interpretation, leading to better patient outcomes and improved workflow efficiencies in a world of ever-increasing demand for radiological services. 

Proudly clinician-led, our patient-first approach comes from a deep understanding of the challenges faced in medical imaging. We fully understand the importance of our work and deeply care about the outcomes we work to support.

Project Description

Founded in 2020, started researching, designing and building our first AI model for chest X-ray targeting multiple clinical findings instead of the conventional single-finding approach. In 2021, we passed our first clinical trials, released our first regulatory-approved product, and were busy researching and designing the much more complex AI for CT brain.

Through iterative research and design, hundreds of hours of observing, interviewing and testing with hundreds of international clinicians, we understood their work’s incredible time pressure, unique operating environment, and fatiguing characteristics, including “eye miles” exerted scanning large visual fields and interacting with small, high-density interfaces.

Ergonomics, usability, efficiency and responsiveness are paramount; a tool used constantly must be lightweight, low-friction, effective and even pleasing to sustain use and positively impact this critical work.

We want clinicians only to interact enough to inform diagnosis — the less interaction, the faster to critical information, the sooner to treat the next patient.

  • The interface is carefully designed for progressive disclosure, expanding to reveal more detail as needed.
  • Results are immediately available when clinicians open a study, expanded or collapsed to reduce biasing. Users can quickly scan the list and focus on the most concerning findings.
  • Users quickly point at findings to expand the UI, highlighting the abnormality on the patient image without any clicks or other direct manipulation.
  • To speed viewing, we auto-straighten and use the best clinical view (Axial, Coronal, Sagittal) on the critical slice showing the most significant extent of the abnormality.
  • Each abnormality preview image is auto-adjusted according to the specific disease marker; something radiologists must constantly do as they search for abnormalities.
  • Users can scroll the full 3D scan, seeing the highlighted region grow and shrink, illustrating its extent in the body and revealing how an abnormality (like a mass or a bleed) impacts other anatomy.
  • Our Slice Scrolling UI previews the extent of a finding, giving a sense of size and spread. For more clinically salient abnormalities, Clinicians can scroll to see precise contour changes throughout the body.
  • With Clinical trial/feedback mode users give feedback on AI findings and visualizations and can audit/tune AI sensitivity thresholds.

Annalise detects 124 CXR and 130 CTB findings and is the only solution analysing up to three images, including lateral images (improving accuracy). The AI algorithm was trained on one of the world’s largest chest X-ray datasets (821,681 images, 284,649 patients, five data sets from Australia, Europe and the USA) and 212,000 CTB studies, then trained/validated by 148 radiologists who hand-labeled each study, generating 280m CXR and 240m CTB labels.

A major peer-reviewed study published in Lancet Digital Health (Jul-21) demonstrated that Annalise CXR improved radiologist ability to perceive 102 chest X-ray findings and was statistically non-inferior for 19. No findings showed decreased accuracy.

An observational study (BMJ, Dec-21) demonstrated improvement in radiologist reporting in hospital and community clinic settings. Of the 2,972 cases reviewed, 92 had significant report changes, 43 had changed patient management, and 29 had further imaging recommendations.


The Interaction Awards are an initiative of the Interaction Design Association (IxDA), a global community of more than 120,000 individuals worldwide dedicated to the professional practice of Interaction Design. Find out how to join your local group and get involved at

© 2012 - 2023 Interaction Design Association