How can AI change the industry? Faster diagnoses and improved treatment options

By Vassia Barba contact

- Last updated on GMT

(Image: Getty/PhonlamaiPhoto)
(Image: Getty/PhonlamaiPhoto)

Related tags: Artificial intelligence, AI, ImaginAB, Life sciences

Artificial intelligence tools will allow the pharmaceutical companies to make predictions about diseases more effectively than before, says ImaginAB’s CEO.

Ian Wilson spoke about the use of AI during the panel discussion “Technologies and Treatments - What’s on the horizon?”, at the 16th Anglonordic Life Science Conference.

ImaginAB, a company focusing on positron-emission tomography (PET) molecular imaging, has developed antibody fragments aiming for safe, fast and detailed illumination of the targeted cancer cells by the radioisotope.

On the path to create precise images of disease activity, ImaginAB’s technology requires fast and efficient data management and, in order to achieve this, the company uses artificial intelligence (AI) and ‘machine learning’ tools.

“We collect multiple data sets from images, biopsies and health records. However, creating a database and storing it in the cloud is easy; what we need to do is look for trends into the complex data sets,”​ Wilson told us.

This is where AI tools can contribute. “Machine learning allows us to rapidly question and explore those trends. Through automatic data comparison we get the chance to identify new things and test hypotheses,”​ he said.

According to Wilson, the implementation of AI in whole body PET molecular imaging not only reduces the image processing times but also allows to produce images of a higher quality.

As a result, the patient gets scanned for less time and is exposed to less radioactivity. This leads to improved patient comfort, more reliable and faster diagnoses, and improved treatment decisions.

ImaginAB is currently testing in Phase II clinical trials for its ‘89ZrCD8’ PET tracer and, as Wilson told us, AI tools are in use to determine how effective the tracer is for monitoring clusters of differentiation 8 (CD8) T cells, in patients treated with immunotherapies.

Nevertheless, Wilson acknowledged that the results of AI tools can only be as good as the data provided.

Bad data in, bad comparisons out. If the experiments aren’t robust, the machine won’t learn. So, the key in trusting computer algorithms is being very certain that the data used is robust.

Users need to keep in mind that AI teaches itself, making use of the data we provide it​. AI always knows what it’s looking for and that is the main factor we should benefit from​,” he concluded.

Related news

Related products

show more

The battle of performance in pharma production

The battle of performance in pharma production

Telstar Life Science Solutions | 26-Mar-2019 | Case Study

Telstar case study. Discover how a versatile freeze-dryer automatic loading & unloading systemDiscover how a versatile freeze-dryer automatic loading...

From Candidate to Clinic- European Capabilities

From Candidate to Clinic- European Capabilities

Catalent Pharma Solutions | 22-Oct-2018 | Technical / White Paper

Catalent’s Nottingham, UK facility focuses on early stage development of small molecule drug candidates from the bench to clinic. This facility offers...

Fill Release at the Right Time in the Right Place

Fill Release at the Right Time in the Right Place

Gelita AG | 19-Sep-2018 | Technical / White Paper

From rapid release in the stomach for e.g. analgesics to enteric release for fills with unpleasant aftertaste – GELITA delivers the respective gelatine...

Related suppliers

Follow us


View more