A carregar...
Publicação
MDCompass - Medical Diagnosis Software
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 4.4 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
MDCompass presents itself as a medical orientation software for diagnosis and an experimental
playground for merging the last generation of technologies with medicine and thus unravelling their gist
relations. The proposed approach model aims to analyze collected medical data in several formats. Such
as values related to blood, urine, stools, or other lab tests and combine them with patients' symptoms to
present multiple diagnoses ranked by their confidence score. Furthermore, the application has
implemented the most recent state-of-the-art speech-to-text model to provide the user with an almost
effortless experience.
To understand how the application was designed, this project will cover how the medical database was
structured and composed, the use of machine learning models for disease prediction, and how the latest
AI technology enabled seamless database querying, bringing contextual awareness to unseen data.
Additionally, it will also be pointed out how all these were complemented with speech-to-text assistance
to provide further dimensionality to the application by filtering audio information into the patient’s
symptoms.
The viability of these initiatives will be addresed by trying to understand the market space needs and
how they could be beneficial considering the worldwide diversity of healthcare models and accessibility,
providing a new option tool for more precarious situations. Thus, assessing the Potential for Maximizing
Value in Healthcare Economies through implementing a Medical Orientation Platform.
Moreover, this thesis aims to review the role of AI in medicine, understanding both the potential benefits
and challenges to medical development.
Descrição
Tese de mestrado, Engenharia Biomédica e Biofísica, 2024, Universidade de Lisboa, Faculdade de Ciências
Palavras-chave
Software de diagnóstico médico Serviço de saúde Aprendizagem profunda Modelos grandes de linguagem Automatização Teses de mestrado - 2024