Research and development of technologies for early non-invasive diagnosis of cancer

Cancer is one of the leading causes of death worldwide. Cancers can be detected or completely prevented through consistent preventive check-ups and screening. These help to find pre-cancerous stages of the disease or cancers in the early stages when treatments are most effective.

Our company is engaged in the research and development of non-invasive early diagnostics to reduce cancer mortality.

The detection method

It focuses on the analysis of volatile organic compounds (VOCs), which are continuously formed in the human body as intermediates or products of cellular metabolic pathways and thus reflect the health status of the individual.

VOCs are emitted through breath, urine, skin, saliva and other biological matrices. Changes in cellular metabolism associated with inflammation, necrosis, disease progression, or changes in the microflora lead to significant shifts in the concentrations of released volatiles.

Thus, the profiles of VOCs are unique and specific to different diseases (providing a holistic picture/VOC fingerprint of a particular disease).

PROJECT SPEnCER

The aim of the project was to use new technologies and artificial intelligence for early, non-invasive and immediate diagnosis of lung and pancreatic cancer from urine and breath samples.

The research was carried out in collaboration with the Jessenius Medical Faculty in Martin, Comenius University in Bratislava and the R&D company Hofitech.

The project “Research of spectroscopic methods for early, non-invasive real-time identification of selected diseases from gas condensate released by lungs and skin” was co-funded by the European Union.

VOC analysis

The analysis and identification of VOCs is carried out by GC-IMS (gas chromatography coupled to ion mobility spectrometry). The technology combines the high resolution of GC for pre-separation and the high sensitivity of IMS. Gas chromatography is used to separate a mixture of substances into individual components according to their solubility and interaction with a stationary phase coating on the wall of the chromatographic column. Subsequently, ion mobility spectrometry separates the ions according to their size and shape in the gas phase.

The method has the advantages of rapid response, high sensitivity, simple operation with minimal sample preparation and low cost, and can therefore also be applied in a clinical setting.

Evaluation by artificial intelligence

We use cutting-edge machine learning algorithms and neural network architectures that are specifically designed to identify and learn from the complex VOC patterns associated with a particular cancer type.

Our method does not focus on tracking specific biomarkers, but on the entire GC-IMS spectrum (VOC profile), which consists of 4000×3000 data points. Although this process requires more computational time and mathematical corrections during neural network training, it is a lossless approach.

Vyhodnotenie dát získaných zo vzorky moču od zdravého jedinca pomocou umelej inteligencie

Evaluating data obtained from a urine sample from a healthy individual using artificial intelligence

Evaluating data obtained from a urine sample from a cancer patient using artificial intelligence

Development of new technologies for cancer diagnostics

Our ambition is to develop a non-invasive diagnostic tool that is able to identify the early stage of cancer based on a urine or breath sample alone.

Our devices combine innovative sensor-based technologies and advanced machine learning algorithms and neural network architectures to recognize specific cancer VOC profiles with high accuracy.

This method offers fast response, high sensitivity, simple operation and low cost.

Due to its many advantages, our future plan is to use this technology for early cancer diagnosis, screening of the general population, as well as for monitoring treatment and cancer recurrence.