Oncoclinics creates web portal and mobile app to store clinical and medical procedures with Azure Cognitive Services

Oncoclínicas is the largest Oncology group in Latin America, with 129 care units in 13 Brazilian states and the Federal District. In its units, it brings together experts in the areas of Oncology, Radiotherapy, Hematology and Bone Marrow Transplantation, as well as multidisciplinary teams, complementary care, integrated care and individualized treatment.

The company has more than 7,000 employees and 1,800 medical specialists, conducts more than 180,000 consultations per year, as well as more than 300,000 chemotherapy and radiotherapy procedures per year.

Technology enhances care

With such a structure and such a large volume of diverse services, the demand for a digital data storage grew, as did the focus on structuring a series of information contained in physical or even digital documents, but without the insertion of the data contained therein into a system that would allow access.

"It's extremely valuable information, but it wasn't accessible. The solution proposed taking these data, centralizing them digitally, so we could use them for decision-making and making them available to physicians," explains Igor de Paula, a partner at Dataside, a Microsoft partner company that worked on implementing the project.

An app with unprecedented tools, mobility and intelligence

Marcio Guimarães, Data and Analytics executive manager at Oncoclínicas, explains that the choice for Microsoft was made after a POC (Proof of Concept), which reflected the most promising result for the company's needs and for the future.

The solution relied on Azure Storage to store scanned exams and host the web portal, Functions App as the application back end, Service Bus integrating microservices also in the back end, and Cosmos DB as the database.

This new technology also uses Azure Cognitive Services on two fronts. One is Computer Vision to extract text from documents via OCR (Optical Character Recognition).

The other is Text Analytics for Health, a tool that uses Machine Learning and AI to digitize medical reports, extracting crucial exam data, enabling large-scale collection – unprecedented use in Brazil.

Marcio explains that the implementation lasted approximately three months. "All the development was done with the Dataside team, but we have units throughout Brazil. With the solution in place, we went to the units to train the teams. Teams would go to the units and spend a few days training the team, monitoring the start of the work," he says.

An optimized operation helps you plan the next steps

The application optimizes the work of entering data into the system, relieving time-consuming manual work. "The application photographs, converts the image, processes, stores everything from the image to the extracted and structured data, classifies users in the application, the patient identification code, the type of exam – whether it is imaging, pathological or molecular," says Marcio.

The solution facilitated access to the data; in the last two months more than 20,000 exams were scanned at the units, which shows the dedication of the team hired to digitize this information in line with the new platform. The data is used for statistical purposes and patient confidentiality is preserved, and the patient signs a consent form that enables its use.

"The first benefit is being able to access important information in a structured way. The physician can even do so when caring for the patient, but now we can access it for other purposes, such as statistical, analysis and research, and to evaluate clinical outcomes. The second point is that it makes information structuring much easier," celebrates Marcio.

It is worth emphasizing that this information is specialized – it is not simple or trivial – it is medical records and medical exam reports, key clinical information that, with the new solution, much of the extraction of this data is now done automatically.

The next step is to expand the solution to cover more exams and patients with different demands, as well as improving the prospective part of the project, creating means for the patient's first interaction to already be entered into this system.

"Microsoft and Dataside have been great partners from the beginning. It's a super innovative project and we knew we'd face difficulties, but we've always had a lot of support, we've overcome each of the challenges we've faced in this period, and we've been very pleased with the outcome," concludes Marcio.