Over the next three and a half years, the AI4Lungs project will develop AI tools and computational models to optimise patient categorisation, improving diagnoses and treatments. In this sense, the researchers will integrate medical data, images and innovative inputs from digital stethoscopes and promote two pilots; the one taking place in Portugal – more precisely at the Centro Hospitalar Vila Nova de Gaia/Espinho (CHVNGE) – will also feature the introduction of a diagnostic technique called liquid biopsy (developed by i3s – Institute for Research and Innovation in Health) as the first line of diagnosis. This is the first time that INESC TEC leads a large-scale European project (€6.9M) entirely focused on lung diseases.
The AI4Lungs project emerges as an innovative and crucial way to address the challenges associated with the proliferation of respiratory diseases, whether related to cancer or interstitial diseases and inflammatory conditions (viral and bacterial pneumonia, including COVID-19), facilitating the clinical processing of patients with any symptom of respiratory disease, reducing costs, and optimising the allocation of resources.
Since it focuses on developing new AI tools, using different types of data collected in clinical practice, e.g., lung auscultation, X-ray, CT, clinical analysis, demographic and clinical data, the project will become vital to support physicians during the diagnosis and treatment processes, by trying to guide them in said processes – which can often be time consuming and require several exams.
“This project focuses on the lung as a whole”, explained Hélder Oliveira, professor at the Faculty of Sciences of the University of Porto and INESC TEC researcher leading the project, stressing that “the idea is that, from the moment a patient has an appointment, the system can support the entire stratification process and understand what the disease is – and if it is necessary to do more tests to diagnose it”.
In practical terms, the project operates according to two dimensions. On the one hand, by developing technologies to improve the devices that are used in diagnosis and, on the other, by integrating all the new tools in a single platform capable of monitoring the entire clinical decision, towards making the process faster and less exhaustive in terms of costs and complementary exams.
“This platform, which we call digital-twin, will have access to data from the project’s partner hospitals and replicate the entire clinical process, suggesting recommendations to healthcare professionals without intervening with normal clinical practice”, mentioned Duarte Dias, INESC TEC researcher. According to him, the innovative aspect of the project stems from the aggregation of various technologies that until now were explored individually.
The project goes even further by using two technologies that have never been used in similar projects. One is digital auscultation – carried out by INESC TEC -, and the other is to introduce i3s liquid biopsy into these decision support systems. According to Hélder Oliveira, in cases of cancer, for example, “liquid biopsy is often used in more advanced cancers; we want to introduce it to understand the type of cancer, instead of opting for dozens of exams”.
“As the lung has different traits, sometimes professionals spend a lot of time making the diagnosis of a patient, with many tests that are not particularly necessary. With this project, we will try to overcome the use of unnecessary exams while reducing the time since the patient arrives at the appointment, until the moment we diagnose his/her disease”, said Duarte Dias.
Already accustomed to working with liquid biopsy, CHVNGE emerges as an extremely important partner, since it will allow to test the decision support platform – the same applies to Mor Research, Rabin Medical Center, in Israel.
“There are several systems that support physicians at various levels; INESC TEC has been developing several technologies in this sense, particularly in terms of X-ray and lung CT”, stated the researchers. What does not exist, they say, “is an approach where these systems are directed towards stratification as a whole, nor the integration of these various systems. This new contribution of the project, together with the whole concept of digital-twin, is the major innovative element, which we seek to advance”.
AI4Lungs is the first healthcare project led by INESC TEC, with a budget of €6.9M. The project features 17 other partners, two of them Portuguese: CHVNGE and I3s. AI4Lungs started on January 1, and will last 42 months.
News, current topics, curiosities and so much more about INESC TEC and its community!
