Improve data collection and create Connection between different systems It is a fundamental task to ensure efficiency Artificial Intelligence (AI) and its feasibility in the area Health. That's what the experts who participated in the panel discussed about the challenges and benefits of using AI Health and Wellbeing Summitcarried out by EstadoThis Monday, the 14th.
AC Camargo Cancer Center CEO Victor Piana highlights that “there is no point in using AI with bad data”. There is a reference to read exams using AC Camargo technology and organize priority order for cancer patients. According to Piana, it is necessary to create a “single mass data” in cooperation between the various health institutions and the SUS.
“Any artificial intelligence model is only as good as the data that feeds it. Small, bad or biased data will produce bad or biased artificial intelligence,” said Francisco Gaia, director of strategy and innovation at MSD Brazil. Health and Wellbeing Summit.
Diego Bahía Fontana, General Coordinator of the Dissemination and Integration of Health Data and Information at the Ministry of Health, said the government is currently working on integrating regionalized systems.
According to him, artificial intelligence has increasingly entered the government agenda. Unlike other developing countries, Brazil has a large financial contribution through SUS and the ability to guarantee a standardized national-level infrastructure, as an advantage in technology implementation.
“We already have the compute capacity with the public cloud. We already have it (Data collection), but in a fragmented manner. Today's effort is to consolidate this. In the Brazilian AI project, short-term projects focus on detecting anomalies in databases”, says Fontana.
The potential of AI in healthcare
Experts highlight the potential for the healthcare sector Prevention of morbidity and complications in surgeryBy analyzing the patient's history and comparing it with data from people with similar characteristics who have a particular disease or problem.
“We started in primary care to collect early quality data, and based on this data, we started asking questions, for example, understanding what are the predictors of patients who developed sepsis in the ICU,” said director Carlos Roberto Ribeiro de Carvalho. Digital Health at the Hospital Das Clínicas of the Faculty of Medicine of the University of São Paulo (HCFMUSP).
“From blood tests and imaging, these are important data to understand what these predictors are for this population of progression to sepsis,” says the expert.
There is a possibility Vaccine development The previously drawn models and Rapid detection of tumors By computer reading of tomography scans.
“People think artificial intelligence is still in its infancy or in the future. But today there are vaccines that have already been developed using artificial intelligence. We have open source vaccine platforms (A term for open source software that can be accessed and used by more people)”, says Gaia.
Last week, the Nobel Prizes Physics and chemistry were given to researchers in the field of artificial intelligence, one of them AlphafoldThe technology responsible for accelerating the development of a vaccine against Covid-19. Computational modeling is capable of predicting protein structures, solving biochemical problems that have remained unsolved for 50 years.
AI is fast, but data collection takes time
Gaia argues that the task should be to improve the quality of data collection and collect this information in a qualified database. Before the actual use of artificial intelligence. “At MST we have invested two years in data through WhatsApp, email and website and now we can communicate with doctors in a relevant manner,” he says.
Hospital das Clínicas realized this after applying artificial intelligence to intensive care unit (ICU) systems, says Carvalho. The initial data so the analyzes were really useful.
He argues that monitoring a patient's lifestyle, comorbidities, and simple health data is essential to understand the potential for complications and risks in the more serious health conditions they may face. as well as mapping predictors of diseases and complications in general.
Areas of health that have previously started doing data collection homework – one of the best ways to do this, readable by AI technology – are today seeing more accurate results, such as radiology, an area that diagnoses, controls and treats diseases like cancer.
“Collecting data is a big step forward in using artificial intelligence, and it's no surprise that radiology has advanced so much. Today, AI already helps detect the most subtle details in mammograms. For example, it does not compete with a radiologist when it comes to diagnosing cancer. “, says Piana. “AI will not replace the doctor, it will replace the doctor who does not know how to use AI.”
“Health and Wellbeing Summit – Has the Future of Health Arrived? Today October 14th from 8am to 6:30pm Shopping in São Paulo takes place at the Events Space at JK Iguatemi. To register, access this. connection.
