Publicação
Utilização da câmara smartphone para monitorizar a aderência à terapia inalatória
| dc.contributor.advisor | Escudeiro, Nuno Filipe Fonseca Vasconcelos | |
| dc.contributor.advisor | Almeida, Rute | |
| dc.contributor.advisor | Marques, Pedro | |
| dc.contributor.author | Ferraz, Sofia Alexandra Gonçalves | |
| dc.date.accessioned | 2022-02-24T14:55:48Z | |
| dc.date.available | 2022-02-24T14:55:48Z | |
| dc.date.issued | 2021 | |
| dc.description.abstract | Self-management strategies can lead to improved health outcomes, fewer unscheduled treatments, and improved disease control. Compliance with inhaled control drugs is essential to achieve good clinical outcomes in patients with chronic respiratory diseases. However, compliance assessments suffer from the difficulty of achieving a high degree of trustworthiness, as patients often self-report high compliance rates and are considered unreliable. This thesis aims to enable reliable adhesion measurement by developing a mobile application module to objectively verify inhalation usage using image snapshots of the inhalation counter. To achieve this, a mobile application module featuring pre and post processing techniques and a default machine learning framework was built, for inhaler and dosage counter numbers detection. In addition, in an effort to improve the app’s capabilities of text recognition on a worst-performing inhaler, a machine learning model was trained on an inhaler image dataset. Some of the features worked on during this project were incorporated on the current version of the app InspirerMundi, a medication management mobile application, planned to be made available at the PlayStore by the end of 2021. The proposed approach was validated through a series of different inhaler image datasets. The carried-out tests with the default machine learning configuration showed correct detection of dosage counters for 70% of inhaler registration events and 93% for three commonly used inhalers in Portugal. On the other hand, the trained model had an average accuracy of 88 % in recognizing the digits on the dose counter of one of the worst-performing inhaler models. These results show the potential to explore mobile and embedded capabilities to gain additional evidence for inhaler compliance. These systems can help bridge the gap between patients and healthcare professionals. By empowering patients with disease selfmanagement and drug adherence tools and providing additional relevant data, these systems pave the way for informed disease management decisions. | pt_PT |
| dc.identifier.tid | 202936597 | pt_PT |
| dc.identifier.uri | http://hdl.handle.net/10400.22/20058 | |
| dc.language.iso | eng | pt_PT |
| dc.rights.uri | http://creativecommons.org/licenses/by-nc-nd/4.0/ | pt_PT |
| dc.subject | Optical character recognition | pt_PT |
| dc.subject | Medication adherence | pt_PT |
| dc.subject | mHealth | pt_PT |
| dc.subject | Remote monitoring | pt_PT |
| dc.title | Utilização da câmara smartphone para monitorizar a aderência à terapia inalatória | pt_PT |
| dc.title.alternative | Use of the Smartphone Camera to Monitor Adherence to Inhaled Therapy | pt_PT |
| dc.type | master thesis | |
| dspace.entity.type | Publication | |
| rcaap.rights | openAccess | pt_PT |
| rcaap.type | masterThesis | pt_PT |
| thesis.degree.name | Mestrado em Engenharia Biomédica | pt_PT |