Nanorobots and swarm intelligence may soon detect cancer from inside the bloodstream
A research group based at the University of the Balearic Islands (UIB) has developed a theoretical system that uses swarms of nanorobots to detect cancer biomarkers circulating in the blood. The study, published in the journal Computers in Biology and Medicine, was announced on 24 September 2026 by UIB and marks a potential shift toward less invasive diagnostic methods.
The work comes from the Motibo research group — formally known as Topological Models for the Treatment of Diffuse Information — in collaboration with the Health Research Institute of the Balearic Islands (Idisba). The system combines nanosensors with mathematical models to identify substances that indicate the presence of cancerous pathologies, according to the researchers.
How the proposed system would work
In the model, tiny devices at nanometric scale would move through the bloodstream, detecting biomarkers associated with cancer. The information gathered by these nanorobots would then transmit to an external device for analysis.
The key innovation lies in how the data is processed. Rather than relying on a fixed threshold — a limitation of earlier diagnostic models — the researchers developed an algorithm inspired by collective intelligence, similar to how insect swarms behave. This approach allows the system to better distinguish between normal and pathological situations, even when there is biological variability or background noise.
A "fusion centre" processes the accumulated data using an algorithm based on Markov chains — a statistical method that models the probability of moving from one state to another. The result, according to the study, is a decision-making method more robust and stable than previous models.
What this means for patients in Portugal
For now, this remains a theoretical framework validated only through computational simulations. But the implications for residents in Portugal are worth noting: if eventually deployed, such technology could reduce the need for invasive biopsies and enable earlier detection of cancer.
Early diagnosis is already a priority in Portugal's healthcare system, where waiting times for oncology appointments and diagnostic tests have long been a concern. A less invasive method that works through a simple blood sample could ease pressure on hospital imaging departments and spare patients the discomfort and risk of tissue biopsies.
Portugal already participates in multinational clinical trial initiatives, including the European FAST-EU pilot project, which streamlines trial authorisations across member states. The country's legal framework for clinical trials was updated in March 2026 under Law No. 9/2026, aligning national rules with European Union standards.
Context: liquid biopsy is already advancing
The proposed nanorobot system builds on a field already making strides in oncology: liquid biopsy. This technique analyses circulating tumour DNA in blood samples and has shown promise in detecting cancer earlier than traditional methods.
Recent studies indicate liquid biopsy can achieve sensitivity above 80% even when tumour burden is low, with specificity reaching 99.55% — meaning very few false positives. The method has identified four times more cancer cases than conventional screening in some research, reducing the proportion of tumours discovered at stage 4 by 14%.
Artificial intelligence is increasingly integrated with liquid biopsy to interpret complex molecular signatures. Medical AI in 2026 already assists radiologists in reading scans, with computer vision algorithms reducing report times from hours to minutes for certain examinations.
What happens next
The Motibo and Idisba researchers emphasise that the model remains theoretical and requires further validation. No specific biomarkers have been named in the published findings, and no clinical trials involving human patients have yet been announced for this particular system.
Portuguese institutions are active in broader efforts to validate AI and digital health solutions. Projects such as MedAIVeritas — a consortium developing methodologies to verify clinical AI safety and efficacy — illustrate the infrastructure already in place to test new diagnostic technologies once they move beyond the simulation stage.
The next step for the UIB team would likely involve partnerships with hospitals or research institutes to test the model with real patient data. For now, the study demonstrates what may be possible when nanotechnology and advanced mathematics converge.
