The world is becoming increasingly “high tech”. The most recent technical advances in science have opened up the scope for technology to expand, and today we have systems such as Artificial Intelligence (AI), an innovation that can be used to improve the way we live in society. The possibility of analyzing an incalculable amount of data makes AI useful for multiple applications. Transformative solutions based on this system are already part of contemporary everyday life and help people choose the best routes on traffic applications, personalize their internet searches, monitor economic fluctuations, shop online, and even take care of their health.
Focusing on this reality, through the practical study of Artificial Intelligence, researchers at the Universidade Federal Fluminense are also trying to develop solutions in the medical field. Professor Débora Muchaluat Saade, from the Computing Institute, coordinates three projects in the area: HealthNet, about advanced and secure networks and systems applied to health; CAPES PrInt – IA, an internationalization program from the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), about artificial intelligence applied to brain signals; and eHealth Rio, a research and innovation network in digital health applied to chronic diseases.
Multimodal applications in healthcare can be employed as a tool in psychotherapy sessions. This allows, for example, presenting certain images and capturing the patient’s facial expressions, producing a report related to these reactions.
Débora believes that the future of Artificial Intelligence in Brazil and other developing countries can revolutionize public health, increasing the effectiveness of patient care, without raising the cost of the system as a whole. “Some time from now, I imagine that we will use sensors on a large scale for continuous monitoring of our health, whether they are wearable or implanted in our bodies. This will enable the prevention of diseases, especially chronic ones, and their early diagnosis. To process all this collected data, AI models and techniques will be indispensable and increasingly used,” she points out.
According to the coordinator, the research and its subprojects are developed at UFF’s MediaCom Lab. “The team is multidisciplinary and includes professors, technicians, as well as undergraduate, masters, doctoral and post-doctoral students from the Computing Institute, in partnership with the School of Engineering, the Biomedical Institute, the School of Medicine, the School of Nursing, and the Antônio Pedro University Hospital (HUAP). Each subproject is led by one of the participating scientists, who guides the development work of the proposed solutions,” she says.
Professor Letícia de Oliveira, from the Biomedical Institute, integrates, together with Débora Muchaluat, the Capes Print-IA Project, which proposes innovations in the mental health area. She explains that mental disorders are chronic and incapacitating diseases that usually start to develop early, with pharmacological and psychotherapeutic treatments that are still not very efficient for most cases. “One of the great challenges of psychiatry today is the early detection of signs that pose a risk for mental disorders. The sooner the symptoms are identified, the sooner the patient can receive the right treatment to delay or stop the onset of the disorder. In this project, we bring great contributions to psychiatry, because when Artificial Intelligence is applied to neuroimaging, it has the potential to capture subtle changes in the pattern of brain activation and perceive changes long before the full onset of the disease.”
The diagnostic support system for dementia, Alzheimer’s disease, and Mild Cognitive Impairment (MCI) that we have created is one such solution. It can be used to give a second opinion to the specialist physician or a first indication to the non-specialist physician in these diseases.
Débora Muchaluat Saade
Computer Science undergraduate Pedro Valentim is part of the subproject that is based on the implementation of multiple modes of interaction with Digital TVs. “The multimodal applications in health can be used as a tool in psychotherapy sessions. This allows, for example, presenting certain images and capturing the patient’s facial expressions, producing a report related to these reactions. I work in the voice recognition, facial expressions, and gestures stages, and I feel fulfilled to participate in state-of-the-art research in the Artificial Intelligence area in Brazil.”
AI at UFF: Developing the future of health care
The HealthNet, CAPES Print – AI, and eHealth Rio projects aim to create proposals that put AI at the service of medical advances. Débora reports that the teams are developing monitoring and diagnostic mechanisms that make the detection and treatment of chronic diseases more efficient. They are also bringing new ideas in neuroimaging analysis using AI techniques. In addition, they are researching ways to identify mental disorders early using physiological signal analysis.
“Our efforts revolve around the creation and deployment of solutions such as clinical decision support systems; multimedia cognitive exercises with sensory effects to aid therapies; image processing and analysis for breast and thyroid tumor detection; application of AI techniques to identify brain patterns that best discriminate emotional states and predict psychiatric symptoms; new communication protocols for wireless body networks; inter-consultation screen support systems, which aim to exchange information between physicians for diagnostic or therapeutic aid; as well as techniques for building a medical record system,” she explains.
In the work we do in the health area, the patient’s well-being is always a priority. Speeding up treatment through early diagnosis is bringing more quality of life to those who are suffering.
Débora Muchaluat Saade
The researcher highlights that the performance of the three projects already has concrete results. “The diagnostic support system for dementia, Alzheimer’s disease, and mild cognitive impairment (MCI) that we created are one of these solutions. It can be used to give a second opinion to the specialist doctor or a first indication to the non-specialist doctor for these diseases. We are even planning a test phase of the system in the routine clinic of the Reference Center for Elderly Health Care (CRASI/UFF), at the Mequinho campus, under the coordination of professor Yolanda Boechat.”
In the same line of neurodegenerative diseases, Débora explains that practical tests of games developed in a virtual reality environment and on digital TV are carried out within the projects. “We started with elderly people who participate in the Memory Workshop, coordinated by Professor Rosimere Santana, from the College of Nursing, and Project Incluir, coordinated by Professor José Raphael Bokehi, from the Institute of Computing. In addition, we intend to build and equip a new multisensory therapy room using multimedia content to aid treatment through cognitive stimuli at CRASI/UFF.”
Yolanda Boechat, a professor at the School of Medicine and coordinator of the Reference Center for Elderly Health Care at HUAP (CRASI/UFF), points out, first of all, that contrary to what one might think, the elderly like new challenges with technology. “In this moment of the pandemic, for example, technological devices bring distances closer and leave the elderly connected to their families. Artificial intelligence also appears as a facilitation for remote medical care, facilitating health surveillance. Here at CRASI, we hope to have access to new forms of assessment that will make diagnoses faster and earlier through the implementation of artificial intelligence-based systems. In this way, we will be able to institute preventive therapies and preserve the patient’s quality of life.”
According to Débora Saade, in research applied to mental disorders, a subproject is being developed with the Biomedical Institute. “We seek the discovery of new markers based on physiological signals for early diagnosis of these disorders, such as post-traumatic stress disorder. Currently, AI techniques are being used to analyze signals acquired in controlled experiments. In the future, we envision testing with patients in outpatient clinics.
For the UFF researcher, Artificial Intelligence has advanced considerably in recent years with the possibility of developing intelligent models. “Within the work, we develop in the health area, the patient’s well-being is always a priority. To speed up treatment through an early diagnosis is to bring more quality of life to those who are suffering. The challenges for the development of innovative solutions in medicine are many and motivate us to move forward with research, always encouraging partnerships between technology and health scientists,” she concludes.