Thermodataset: sistema de acesso e base aberta de dados de termografia dinâmica em tumores de mama

Abstract

Breast cancer is the most common cause of cancer death in women in Brazil. Therefore, it is necessary to develop methodologies and tools that facilitate research in the area. This study aimed to make openly available data from research related to dynamic thermography technique in participants with suspected breast cancer (initially inaccessible and destandardized), including medical records, dynamic thermography files, and imaging and histopathological examinations related to breast neoplasia. Based on the requirements survey, a dataset and web access system were designed. The data were selected, organized, anonymized, and structured in an online dataset, containing general information for each participant, examinations, and family history. The web platform was developed using the React library for interface development and the Node runtime environment. Both the dataset and the platform were made available using AWS cloud services. The interface, available on http://thermodataset.utfpr.curitiba.br/, allows all users to register and view, filter, and download information from the dataset, as well as authorized users to manage and add data in a standardized way. After the development, the dataset contemplates 280 exam images (mammographies, ultrasounds, and thermographies) and data from 21 research participants, including their habits, family history and exams data, adding up to 60 attributes to each participant. It is expected that access to clean, validated, and integrated data related to dynamic thermography facilitates research on the use of this technique in early detection of breast cancer.