Highlights
• Ixico presented an artificial intelligence tool at CTAD 2026 that supports radiologists in detecting cerebral microbleeds.
• The tool detects around 90% of microbleeds per participant scan, with consistent results across studies and techniques.
• A two-stage algorithm first locates candidate lesions, then separates true microbleeds from look-alike features, called mimics.
• The intended use is vascular safety monitoring in Clinical Trials of anti-amyloid treatments for Alzheimer's disease.
Ixico (LSE:IXI) took a new artificial intelligence tool to CTAD 2026 with a narrowly defined job: helping radiologists find cerebral microbleeds, known as CMBs, on MRI scans. The setting it targets is specific too, namely the safety monitoring that runs alongside anti-amyloid Alzheimer's trials.
What the Tool Does
The tool is designed as a support for radiologists. Its task is to identify cerebral microbleeds on MRI images, a Job that forms part of vascular safety monitoring in anti-amyloid Alzheimer's studies.
Ixico (LSE:IXI) chose CTAD 2026 as the venue to present the technology, and the announcement sets out the tool's design and detection results together.
Inside the Two-Stage Algorithm
The software works in two distinct steps. In the first, it locates candidate lesions across the scan, flagging areas that could be microbleeds. In the second, it classifies those candidates, separating true CMBs from mimics, which are features that resemble microbleeds without being them.
Splitting the work this way means the system casts a wide net before narrowing it down. The classification stage is where the tool decides which candidates should be presented as genuine findings.
Detection Performance
According to the company, the tool detects around 90% of microbleeds per participant scan. Ixico (LSE:IXI) also reports that this performance was consistent across studies and across imaging techniques.
Consistency matters in this context because trials can draw on scans gathered in different studies and with different imaging approaches. The reported results cover that variation rather than a single dataset or method.
The Aim: Time and Consistency
Ixico states that the tool aims to reduce the time radiologists spend on this work and to lower inter-reader variability, meaning the differences that can arise when separate readers assess the same scan. Both aims relate to the practical workload of reviewing images in large trials.
The announcement does not disclose regulatory clearance, customers or commercial availability for the tool. The update is centred on the technology, its design and its reported detection performance as presented at CTAD 2026.
Understanding Cerebral Microbleeds
Cerebral microbleeds are small areas of bleeding in the brain that appear as tiny dark spots on certain types of MRI scan. In studies of anti-amyloid treatments for Alzheimer's disease, imaging is used to monitor patients for vascular changes, and microbleeds form part of that safety picture. Finding and counting them consistently across many scans is a detailed task for radiologists, which is why a tool that supports the process has practical interest.
Detection and Consistency
Ixico reports that the tool detects around 90% of microbleeds per participant scan and that the performance was consistent across studies and imaging techniques. A two-stage design first finds candidate lesions and then classifies them, separating true microbleeds from mimics that resemble them. Inter-reader variability describes the differences that can arise when separate readers assess the same scan, and reducing it is one of the stated aims, alongside reducing the time radiologists spend on the work. The announcement does not give regulatory clearance, customer or commercial availability information, so the update is best read as a description of the technology and its reported results. The technology was presented at CTAD 2026, an international conference on Alzheimer's disease clinical trials. Presentations at scientific meetings give developers a way to share design and performance information with the clinical community. Readers following Ixico can therefore look for later announcements describing how the tool is used, further performance data and any commercial or regulatory developments. The reported detection rate and the two-stage design are the central facts in this update, and the company's aims of saving time and improving consistency describe why the tool has been developed for large trials.






Please wait processing your request...