Descripción del proyecto
Ordenadores para diagnosticar como médicos humanos
El potencial de la inteligencia artificial en la asistencia sanitaria es cada vez mayor. En lo que se refiere los diagnósticos médicos, los ordenadores pueden tener el mismo éxito que los médicos humanos. El proyecto financiado con fondos europeos MrDoc ha desarrollando una plataforma de IA de aprendizaje semisupervisado que es capaz de analizar e interpretar bases de datos médicos. Ha diseñado un proceso que imita la imaginación creativa humana para detectar y diagnosticar rápidamente y con un alto nivel de precisión algunas enfermedades no transmisibles, como la enfermedad cardiovascular y la diabetes, por medio de parámetros biométricos (tensión arterial, variabilidad de la frecuencia cardíaca, hemoglobina, glucosa en sangre). El proyecto está preparándose para su comercialización con el objetivo de vender y licenciar su solución a tres grupos destinatarios: pacientes, desarrolladores de «software» y herramientas de «hardware» (así como aplicaciones) y empresas farmacéuticas.
Objetivo
Non-communicable diseases such as cardiovascular diseases, diabetes, are by far the leading cause of death in the world and a growing burden for patients, healthcare providers and local economies. Despite many NCDs conditions like cardiac arrhythmia, diabetes, hypertension can be cured with early detection, they don’t often show symptoms. During their medical check-up, medical practitioners (GP) can’t be accurate as specific examinations (e.g. EGCs, blood tests), resulting in a growing number of errors or false negative/positive, which represent for Healthcare systems and additional financial burden. People are usually discouraged from doing specific examination due to long waiting time, invasiveness of medical tests and additional costs.Even if technological advancements have led to AI based easy-to-use solutions able to contribute positively to easy and early detection of diseases and pre-diseases condition, they come along with many significant limitations, such as the need to train on huge amounts of labelled data and difficulties in managing inputs that are noisy, incomplete or simply different from the original dataset (such data generated from a smartphone camera).This results in limited accuracy or significant costs and time consume for labelling of data. We have developed a platform based on a semi-supervised learning AI, able to analyse and interpret medical dataset through a process that mimics human creative imagination and, in a very short timeframe, detect and diagnose some NCDs and biometric parameters (blood pressure, Heart rate variability, haemoglobin, blood glucose) from “dirty” signals, generated by consumer electronics devices (smartphones, closed circuit cameras, etc.), with a high level of accuracy overcoming existing limitations.We aim at selling and licence our solution to 3 main targets: - final consumers/patients, - producers/owners of software and hardware tools (as well as Apps) in Health sector, Pharmaceutical companies.
Ámbito científico
Not validated
Not validated
- medical and health sciencesclinical medicinecardiologycardiovascular diseasescardiac arrhythmia
- natural sciencescomputer and information sciencesartificial intelligencemachine learningsemisupervised learning
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringsensorsoptical sensors
- medical and health sciencesclinical medicineendocrinologydiabetes
- social scienceseducational sciencespedagogyactive learning
Programa(s)
Convocatoria de propuestas
Consulte otros proyectos de esta convocatoriaConvocatoria de subcontratación
H2020-SMEInst-2018-2020-1
Régimen de financiación
SME-1 - SME instrument phase 1Coordinador
00146 ROMA
Italia
Organización definida por ella misma como pequeña y mediana empresa (pyme) en el momento de la firma del acuerdo de subvención.