“We’re trying to mimic the mammalian sense of smell artificially. We’ve now developed an algorithm that can distinguish ovarian cancer from endometrial cancer and healthy control groups, using data from an electronic nose,” says Donatella Puglisi, associate professor at Linköping University.
In ovarian cancer, symptoms are often vague and similar to those of other more common diseases. This type of cancer is therefore detected at a late stage of development, when survival outcomes are poor. 
Olov Planthaber
“More and more people are being diagnosed with cancer, especially young adults, and this is alarming. If screening were more accessible, both in terms of cost and location, it would be possible to improve early diagnosis. Our approach could facilitate the adoption of new screening protocols and the development of new diagnostic methods, improving survival rates, quality of life, and overall clinical outcomes” says Donatella Puglisi.
AI opens new doors
Electronic nose technology has been around for about 60 years. The prototype used by the researchers has 32 sensors that react to various volatile substances emitted from the sample being examined. Each form of cancer emits different volatile substances, thus different cancers “smell” differently.
Olov Planthaber
The sensors are of a relatively simple model and are available on the market. But with the dramatic development of machine learning and AI in recent years, established technology can be used in new ways.
Current healthcare cancer screening by blood test involves searching for a number of biomarkers that are unique to the form of cancer suspected. However, test analysis is slow and often not very accurate.
“Unlike in breast cancer, there is currently no reliable ovarian cancer screening method. These tests are often based on a single biomarker and lack the precision required to detect the disease at an early stage. Our method is therefore far ahead not only in terms of accuracy but also in the ability to identify early disease,” says Jens Eriksson, associate professor at LiU and CTO at VOC Diagnostics AB, the company developing the electronic nose.
High accuracy
The method developed by the researchers does not need the identification of a specific biomarker. Instead, the electronic nose picks up a large variety of volatile substances emitted from blood plasma samples. The data are then analysed using advanced machine-learning models to identify patterns specific to, in this case, ovarian cancer. The models are trained on known samples from a biobank. The tool has 97 per cent accuracy.
Olov Planthaber
“It’s a simple test that takes 10 minutes and gives a clear result. Our method can test many people at a low cost and is much more accurate than what’s on the market today. This study is a pilot, but we hope it will be used as part of cancer screening within three years. Right now, we’ve focused on detecting cancer, but the applications are endless,” says Jens Eriksson.
The study was primarily funded by Vinnova, Formas, the Swedish Energy Agency, and the Swedish Research Council. Data processing and computations were made possible through the National Academic Infrastructure for Supercomputing in Sweden (NAISS).
Article: Biomarker-Agnostic Detection of Ovarian Cancer from Blood Plasma Using a Machine Learning-Driven Electronic Nose, Ivan Shtepliuk, Lingyin Meng, Christer Borgfeldt, Jens Eriksson, Donatella Puglisi, Advanced intelligent systems, published online 6 January 2026, DOI: 10.1002/aisy.202500838
Olov Planthaber
