Description du projet:
This project aims at developing a new portable device based on electrochemical
sensors, pharmacokinetic models, and state-of-the art machine learning lgorithms.
This new device will be the very first ever-conceived single-device capable of using
real-time data dealing with molecular concentrations in the patient' blood and patient' metadata to provide robust clinical recommendations for dosage adjustments in precision oncology. In pharmacological treatments, the scope of precision medicine is about tailoring a precise delivery of therapeutic compounds neatly adapted to the patients' characteristics. This mainly means adapting the dose to the exact disease-profile of each individual patient as well as to his metabolic behaviour right during treatment administration. This last is the most difficult to establish since it relies on genetic predisposition, epigenetic influences, and acquired characteristics (phenotype, senescence, comorbidity, comedication), all patient features that have complex interactions among them as well as with the patients' environment and life-style. Following the time-evolution of drugs concentration in patients' blood is also more difficult in case of therapeutic cocktails which, on one hand, are much more efficient and effective in cancer therapy than each single therapeutic agent while, on the other hand, are more difficult to monitor-in-time because different drugs strongly-interact in both patients metabolism and drug sensors. Therefore, the main objective is here to establish a new tool for precision oncology by developing a micro-fabricated system that integrates: electrochemical sensors for measuring several anticancer drugs commonly used in chemotherapy; all the electronic circuits required for sensors reading as well as for data elaboration and communication; local intelligence based on machine learning algorithms as implemented on hardware; dedicated fluidics to separate serum from patients' blood and to drive the serum to the sensors; a user interface to interact with the system by using any mobile device (tablets, smart-phones, smart-watches, etc.). The electrochemical sensors will be micro-fabricated on a dedicated platform realized in silicon or in ceramic. The electronic circuitry will be realized with off-the-shelf components; the algorithms for local intelligence will be, implemented on hardware with off-the-shelf computing systems (e.g., FPGA or Edge-TPU) once optimized and robust enough for the application. Data fusion and data mining will be used in optimizing the algorithms by considering data from sensors and other source of information relying to the patient' characteristics (age, weight, classification as metaboliser, etc.), and to the drugs delivery (name, administered dose, time of administration, etc.). The drug monitoring will be tested with lab samples in the presence of other spiked interfering metabolites, and further validated with patient's samples by comparing with independent lab measurements. The originality dealing with the sensors part of the platform is also
assured by the project need to design and realize an electrochemical biosensor for
busulfan, sensor that has never been reported in literature so far. Moreover, the final platform will be capable to provide the simultaneous detection of busulfan,
methotrexate, 5-fluorouracil, cyclophosphamide, ifosfamide, and etoposide with a
single multi-panel device. This also has never been reported in literature so far.
Research team within HES-SO:
Thoma Yann
, Rigamonti Roberto
Partenaires académiques: ReDS
Durée du projet:
01.09.2022 - 31.08.2026
Montant global du projet: 5'000 CHF
State: Completed
Help
SWITCH edu-ID
Administration
