Sepsis is a life-threatening whole-body inflammatory reaction caused by a severe infection (e.g. bacteria, virus). With mortality rates around 35%, sepsis is responsible for 11 million deaths worldwide every year, being the largest killer of children (more than 5 million annually). Furthermore, sepsis survivors commonly suffer long-term health damage with a diminished quality of life and persistent high risk of mortality. The window of opportunity for sepsis management is in hours: the chance of survival drops by 7.6% each hour of disease progression until an appropriate treatment is started. Early and accurate sepsis detection and stratification is essential for enhancing survival rates. This is challenging, due to: i) complex diagnostic criteria requiring multiple screenings for differential diagnosis, (ii) time-consuming methods for identifying the bacterial causes of sepsis. Even worse, these conventional immunoassay techniques are performed in centralized laboratories, causing extra delays due to specimen transfers. Point-of-care (PoC) testing performs the analysis at bedside, enabling earlier diagnosis and facilitating therapy assessment. Currently available PoC platforms, however, do not offer multiplexed capabilities, i.e. they detect a single biomarker and do not identify the bacterial cause of the infection, nor do they offer any decision support system for disease classification.
AMBROSIA aims to transform integrated plasmo-photonic refractive index sensors into a disruptive solution for sepsis diagnosis at the point of care that will offer multiplexed (within a single test) quantification of multiple protein biomarkers and bacteria within a few minutes providing also real-time disease stage classification enabling a rapid and precise decision making for therapy and medical actuation. To realize its ambitious goals, AMBROSIA targets to build upon:
• The adoption of best-in-class integrated plasmo-photonic devices synergizing aluminum (Al) plasmonics and silicon nitride (SiN) photonics. Incorporating Bragg-grating decorated plasmonic stripes envisions to boost sensitivity to unprecedented values of 130000 nm/RIU.
• The co-integration of the plasmo-photonic sensing technology with microelectronics and microfluidics into integrated self-contained disposable sensing chips including on-chip lasers and photodiodes by means of the high-throughput and cost-effective micro-transfer technology (μTP), pioneering the field of manually pluggable disposable sensor chips.
• Optically-enabled artificial intelligence (AI) to provide a real-time identification and classification of disease severity. This will enable rapid knowledge-based critical decision-making and back-checking against misdiagnosis and medical mistakes, allowing for critical time savings and accelerating the pathway towards the optimal medical treatment. AMBROSIA’s disease classification will rely on the adoption of state-of-the-art sepsis medical protocols and their translation into Deep Learning models that will be then deployed on hardware via low-power photonic Deep Neural Networks with electro-optic activations, facilitating in this way rapid and low-energy decisions.
