diagnosis e3

diagnosis.health.eco.br New non-relational storage structures are being developed, as previously mentioned, Neo4J's Cloud Aura DB data storage platform can place the entire table structure of a relational database into a Graph DB. In this context, it is important to develop technology to generate a knowledge base for a diagnostic support system, based on a set of information (e.g., anamnesis) associated with medical procedures, to enable the transformation of textual information related to the diagnosis into concepts that could be integrated into knowledge bases in the patient's medical record. The analysis of textual information from the anamnesis makes it possible to create an algorithm to detect joining terms (Ex.: de), allowing a certain algorithm to relate the word to a symptom (Ex.: Pain), and its respective subsequent context (Ex.: Head). The result would be the identification of the symptom "Headache". Other possibilities are the occurrence of a related number, a connecting element between the symptoms (Ex.: e), or a complement to the detected symptom (Ex.: High). In this way, the text "Headache and High Fever" could be related to two concepts: "Headache" and "High Fever". By dividing the symptoms in the text, it would be possible to identify the likely related diagnostic contexts. The algorithm would treat all words as related and ordered concepts within a map. PLN (Natural Language Processing) of the data would be carried out to acquire lists of terms that would not be considered important, classifying the text as a network of words represented through a concept map. In this area, the research focus is supported by the principle of least effort defined by George Kingsley Zipf. The Zipf law directs the focus on project language R&D. The structuring of the data allows the recognition of patterns, as well as the implementation of an assistance system, based on the diagnoses and symptoms highlighted by medical specialists, described in text fields of a conventional system. Recording all information about medical interactions on the health.eco.br platform can contribute to a technological initiative and bring significant benefits to the general population. If the patient allowed the tool to analyze their data, the system could inform possible risks associated with the occurrence of a certain disease. Furthermore, the system would suggest healthier behavior to the person in a certain aspect. With data access authorizations granted, the system could recognize more patterns, based on the epidemiological analysis of the population. If the software identified the actual pattern of the disease, the application would indicate a specialist doctor to analyze the person's data. With the operational process of acceptance of both parties existing, the person would now be a patient of the respective doctor who would have access to all the historical data necessary for the aforementioned diagnostic analysis. If the need for more data and/or exams arises, the system could provide, in an interactive process between the doctor and the patient, specifically personalized dynamic forms for viewing and recording related information, both synchronously and asynchronously. The information identified in the diagnostic process would be associated with Graph DBs used to represent knowledge through related symptoms, and integrated with those pre-defined diagnostic symptoms in the specialist doctor's cognitive structure, containing information related to the respective possibilities. This mapping aims to create protocols (guidelines) to facilitate data analysis and allow the formalization of information related to each diagnosis.