By Adrian Arnaud, partner at Neurotech, and Paulo Melo, co-founder of PickCells.
Almost 70% of the types of cancers today are treatable. The cure depends, in these cases, on early diagnosis. Through cognitive intelligence, supercomputers from five major global players produce medical reports in minutes, with an assertiveness of almost 70%, while the best medical teams, after days of analysis, reach up to 60%. In Brazil, smaller companies that use Artificial Intelligence in diagnostic medicine reach 94% of precision.
In addition to the development of their own technologies, companies from Brazilian technological hubs, such as Porto Digital, from Pernambuco, create solutions using AI aimed at the health problems that are characteristic of the country, such as the development of tests for neglected tropical diseases, such as leishmaniasis, schistosomiasis and tuberculosis, which, in addition to being more assertive, are cheaper than traditional ones, which increases the number of people who can benefit.
AI also ends up with distances, which meets other typical needs of the country. Many hospitals do not have their own laboratories, while some states do not even have specialists in areas such as radiology, which forces patients to travel great distances to obtain reports of their exams. In the case of X-rays, CT scans and MRI scans, which do not require tissue samples, they can be transmitted electronically so that a cloud system identifies, in minutes, patterns, which pathologies and medical specialty are related, for a doctor to produce the report. The solutions thus increase the productivity of healthcare professionals.
AI also allows hospitals to predict the consumption of supplies, beds, equipment and personnel. With the use of Big Data and neural networks, the history of consumption and factors that affect it are identified, which varies from one institution to another, according to the public, location and the way they operate. In this way, hospitals can forecast the demand for personnel and resources, particularly in areas of greater criticality, and their disbursements in advance. The analysis of outputs of thousands and thousands of items in years, carried out in minutes by the AI, is an unfeasible task for humans.
Another big problem for hospitals and clinics is when patients don't go to the scheduled appointments, the so-called show. The gaps left in the schedules result in both loss of revenue, due to missed appointments, as well as personnel expenses, due to the mobilized professionals. This leads several institutions not to work with appointments, which causes other setbacks, such as the arrival of a number of people much higher than the volume of services that can be carried out. The analysis of the patient's history through public scheduling bases, makes it possible to predict, with up to 90% of precision, which patients will or will not go to consultations.
Another benefit is cost savings and resource optimization. Already widely used by the financial market to detect fraud, AI can be used for the same purpose in the health area. Before reimbursing clinics and hospitals for treatments carried out with their policyholders, operators analyze the compatibility between procedures and their costs with materials, equipment and professionals. The work mobilizes teams of analysts who face, in some companies, tens of thousands of cases per month. Amid this volume, payments are released for materials that have never been used. Prostheses, due to the high values, are priority targets for fraudsters.
The adequate analysis of gigantic volumes of data demands the use of Big Data and AI, without which the verification, still subject to many flaws, involves a significant number of professionals. Gaining scale and, therefore, hiring more staff, would make the business unfeasible. The set of information sent by hospitals to the operators can generate cost profiles for each type of treatment, which can be fraud and highly complex cases.
On the other hand, the reimbursements disallowed by operators end up generating problems in the cash flow of hospitals. Until the costs of certain procedures are proven, the amounts are not released. In cases like this, which can result from errors in filling out forms, the AI can also trace the profile of costs that are more likely to be questioned, which allows corrections to be made before the processes are forwarded. In this way, instead of receiving the amounts overdue for months or having to enter funds, the institutions guarantee receipt on the first request.
Amid these clashes between operators and hospitals, doctors and other health professionals have their payments postponed due to disallowances, which make them resort to loans from hospitals. Institutions, such as banks, charge interest on these transactions. With the experience in using AI in credit risk analysis, it is possible to identify the probability that the operator's repayment will be delayed, in how long, or not, which allows these loans to have interest appropriate to the risk of each procedure, the that favors both involved.
The presence of major international players is favorable for health managers to understand the gains that artificial intelligence can provide to the sector, which has been intensifying in recent years. There are, however, cultural factors, characteristics of each region, that need to be understood locally. This aspect, as well as the mastery of technology, is already made available by Brazilian companies.
In addition to the development of their own technologies, companies from Brazilian technological hubs, such as Porto Digital, from Pernambuco, create solutions using AI aimed at the health problems that are characteristic of the country, such as the development of tests for neglected tropical diseases, such as leishmaniasis, schistosomiasis and tuberculosis, which, in addition to being more assertive, are cheaper than traditional ones, which increases the number of people who can benefit.
AI also ends up with distances, which meets other typical needs of the country. Many hospitals do not have their own laboratories, while some states do not even have specialists in areas such as radiology, which forces patients to travel great distances to obtain reports of their exams. In the case of X-rays, CT scans and MRI scans, which do not require tissue samples, they can be transmitted electronically so that a cloud system identifies, in minutes, patterns, which pathologies and medical specialty are related, for a doctor to produce the report. The solutions thus increase the productivity of healthcare professionals.
AI also allows hospitals to predict the consumption of supplies, beds, equipment and personnel. With the use of Big Data and neural networks, the history of consumption and factors that affect it are identified, which varies from one institution to another, according to the public, location and the way they operate. In this way, hospitals can forecast the demand for personnel and resources, particularly in areas of greater criticality, and their disbursements in advance. The analysis of outputs of thousands and thousands of items in years, carried out in minutes by the AI, is an unfeasible task for humans.
Another big problem for hospitals and clinics is when patients don't go to the scheduled appointments, the so-called show. The gaps left in the schedules result in both loss of revenue, due to missed appointments, as well as personnel expenses, due to the mobilized professionals. This leads several institutions not to work with appointments, which causes other setbacks, such as the arrival of a number of people much higher than the volume of services that can be carried out. The analysis of the patient's history through public scheduling bases, makes it possible to predict, with up to 90% of precision, which patients will or will not go to consultations.
Another benefit is cost savings and resource optimization. Already widely used by the financial market to detect fraud, AI can be used for the same purpose in the health area. Before reimbursing clinics and hospitals for treatments carried out with their policyholders, operators analyze the compatibility between procedures and their costs with materials, equipment and professionals. The work mobilizes teams of analysts who face, in some companies, tens of thousands of cases per month. Amid this volume, payments are released for materials that have never been used. Prostheses, due to the high values, are priority targets for fraudsters.
The adequate analysis of gigantic volumes of data demands the use of Big Data and AI, without which the verification, still subject to many flaws, involves a significant number of professionals. Gaining scale and, therefore, hiring more staff, would make the business unfeasible. The set of information sent by hospitals to the operators can generate cost profiles for each type of treatment, which can be fraud and highly complex cases.
On the other hand, the reimbursements disallowed by operators end up generating problems in the cash flow of hospitals. Until the costs of certain procedures are proven, the amounts are not released. In cases like this, which can result from errors in filling out forms, the AI can also trace the profile of costs that are more likely to be questioned, which allows corrections to be made before the processes are forwarded. In this way, instead of receiving the amounts overdue for months or having to enter funds, the institutions guarantee receipt on the first request.
Amid these clashes between operators and hospitals, doctors and other health professionals have their payments postponed due to disallowances, which make them resort to loans from hospitals. Institutions, such as banks, charge interest on these transactions. With the experience in using AI in credit risk analysis, it is possible to identify the probability that the operator's repayment will be delayed, in how long, or not, which allows these loans to have interest appropriate to the risk of each procedure, the that favors both involved.
The presence of major international players is favorable for health managers to understand the gains that artificial intelligence can provide to the sector, which has been intensifying in recent years. There are, however, cultural factors, characteristics of each region, that need to be understood locally. This aspect, as well as the mastery of technology, is already made available by Brazilian companies.
Disclaimer: The opinion presented in this article is the responsibility of its author and not of ABES - Brazilian Association of Software Companies
EN 
PT 
