AI / Computer Vision

By using a combination of Artificial Intelligence and Computer Vision (where AI trains computers to interpret and understand the visual world using digital images from cameras and videos) machines can accurately identify and classify objects — and then react to what they see with Computer Vision.

Computer Vision is a subset of AI that enables computers to see and make sense of the world. It is the AI application that allows a computer to learn to analyse information from photos or video to thermal and infrared data, amongst other sources, and then make decisions or come to a clear understanding of the environment or situation based on that information.

Such a system has to be capable of processing and interpreting visual information so that it can be used for patterns and object recognition, and for adapting 2D images from our 3D world into 3D information. Yet in just under a decade, the accuracy of object identification rates has increased from 50% TO 99%. This makes computer vision more accurate than humans at reacting rapidly to visual data.


AI/Computer Vision and Melanoma Cancer

The process begins with a curated set of images or video data, known as training data. This is used to help the machines learn certain things about a certain topic; think different stages of development of a lesion/s.

Any training data will require within it sample images of lesions at various stage of growth and those characteristics. Each image will be tagged with metadata indicating the correct answer – in this case, permitted or prohibited.

A neural network will process the visual data, using pattern recognition to identify the many different components of an image. Its outputs, or ‘answers’ as to what is and what is not a lesion, are fed back into the system allowing it to learn and improve in accuracy and learn from the images it receives.

  • Phase 1 Triage – With thousands of images to review, we built a model to triage cases. The first step is to process images and analyse constructs such as the boundaries of a tumour or lesions using a convolutional neural network (CNN).

  • Phase 2 Augmented – Image data is combined with patient data such as their age, medical treatment, family history, and risk factors such as diet and lifestyle. Our model reviews this visual and contextual information to determine the probability that the scan shows cancer.

  • Phase 3 Recommended – We then built a recommendation engine that suggests treatments and interventions plotting them against risk and likely outcomes. For example, a patient with multiple existing lesions might be served better with a monitoring and palliative intervention than invasive surgery.

  • Phase 4 Learn – In order to ensure the system continues to perform at outstanding levels, the hospital then tracks the outcomes of the recommendation engine, creating a feedback loop that helps it to learn and further improve outputs at both the triage and recommendation phases.



  •  Fully automated evaluation (faster than the human eye).
  •  Increase speed and accuracy of evaluations, e.g. lesion/moles and identify different types of lesions.
  •  The beginnings of a skin cancer pathway / bespoke triage system to utilise/maximise time and resources / reduce unnecessary referrals from Primary Care to Secondary Care.
  • Maximise Dermatologist time and resource helping to saving time and money.
  •  The more data delivered to the platform the more intelligent and accurate the AI’s become in prevention and prediction.
  •  Ability to add additional demographic data sets to AI’s – socio economic, skin types, gender, geographical location, diet etc and potential correlation to specific cancers/lesions/moles to identify malignant and pre-malignant lesions if identified early enough.

Our Computer Vision is supported by Predictive Analytics where the use of data, statistical algorithms and machine learning techniques identify the likelihood of future outcomes based on historical data.
The goal is to go beyond knowing what has happened to providing a best assessment of what will happen in the future.