Poor adherence in medical treatments is one of the biggest problems to tackle in the medical field. Poor adherence reduces clinical benefit provided by the treatment and increases the likelihood of hospitalization, especially in chronic illnesses, while good adherence increases the effectiveness of the treatment and promotes healthy lifestyles. This affects directly not only the quality of life of the patient but also the costs and quality of the healthcare system. Wearable technologies are a great tool to measure and promote medical adherence because they can monitor the at-home behavior of the patient and can also give reminders and feedback. This project addresses poor adherence by monitoring human activities through the movement of an arm, that could allow physicians to understand whether their patient is following the recommendations, by taking a pill. The approach we gave to this project is the design of a system that can acquire accelerometric signals from the forearm and send them to a smart-phone. In order to allow the data to be processed for gesture tracking, the device is connected to a smart-phone application that sends the data to the cloud. The designed system includes software and hardware. The wearable device thought for this application is a wrist band. The hardware design included a 6-axis IMU featuring an accelerometer and a gyroscope. This device performs an acquisition only when requested by the smartphone with which communicates via Bluetooth. Considering that the software for gesture tracking analysis (non developed in this thesis) is based on the knowledge of the gesture specific motion pattern that needs to be detected, customized to the specific patient, a training function is needed by which such gesture can be acquired in an univocal way in a dedicated session, the software design includes a systems interface that is adapted for two types of users: regular app user and expert user. Each type of user has specific interactions in the system. The expert user (physician) specifies the gestures to be tracked and the relevant training mode, meaning it can create and train gestures as well as managing the alarms for the regular app users. While the regular app user (patient) can train the already existing gestures and set alarms according to when the medication is needed to be taken. For storing all the data and information a cloud database was also designed. In the document, all the considerations are taken for the circuit design, the schematics and layouts of the circuit, the state machine that was programmed as firmware in the MCU, as well as a suggested method for testing the device are written. There is also presented the system's use case scenario, activity diagrams; the database class diagram, entity-relationship diagram, technical description of tables, and queries; and the app environments overview. And will explain overall how the system was built employing a database and a smartphone-app, and how this system works. At the end of the document, there is a discussion about the applicability in the market, the possible improvements, and future work for the designed device.
La scarsa aderenza ai trattamenti medici è uno dei maggiori problemi da affrontare in campo medico perché riduce il beneficio clinico fornito dal trattamento e aumenta la probabilità di ospedalizzazione, soprattutto nelle malattie croniche, mentre una buona aderenza aumenta l'efficacia del trattamento e promuove stili di vita sani. Ciò influisce direttamente non solo sulla qualità della vita del paziente, ma anche sui costi e sulla qualità del sistema sanitario. Le tecnologie wearable sono un ottimo strumento per misurare e promuovere l'aderenza medica perché possono monitorare il comportamento del paziente mentre è a casa e possono anche dargli indicazioni e feedback. Questo progetto affronta la scarsa aderenza monitorando le attività umane attraverso il movimento di un braccio, che potrebbe consentire ai medici di capire quando il loro paziente sta seguendo le raccomandazioni, relative al prendere una pillola. L'approccio che abbiamo avuto con questo progetto è la progettazione di un sistema in grado di acquisire segnali accelerometrici dall'avambraccio e inviarli ad uno smartphone. Per consentire l'elaborazione dei dati per fare gesture tracking, il dispositivo è collegato a un'applicazione per smartphone che invia i dati al cloud. Il sistema progettato include software e hardware. Il dispositivo wearable pensato per questa applicazione è un bracciale. La progettazione hardware include un sensore IMU a 6 assi con un accelerometro e un giroscopio. Questo dispositivo esegue un'acquisizione solo quando richiesto dallo smartphone con cui comunica tramite Bluetooth. Considerando che il software di analisi dei gesti (non incluso in questa tesi) presuppone la conoscenza del pattern motorio relativo al gesto che si vuole riconoscere, customizzato per il singolo paziente, è necessario predisporre una funzione di "allenamento" in cui tale gesto può essere registrato in modo univoco in una opportuna sessione, il design include un'interfaccia di sistema che è adattata per due tipi di utenti: utente normale dell'app e utente esperto. Ogni tipo di utente ha interazioni specifiche nel sistema. L'utente esperto (medico) specifica i gesti di cui tenere traccia e la modalità di allenamento pertinente, il che significa che può creare e addestrare gesti e gestire gli allarmi per gli utenti normali dell'app, mentre l'utente normale (paziente) dell'app può allenare i gesti già esistenti e impostare allarmi in base a quando è necessario prendere delle medicine. Per l'archiviazione di tutti i dati e le informazioni è stato progettato anche un database sul cloud. Nel documento, sono presenti tutte le considerazioni che sono state effettuate per la progettazione del circuito, gli schemi e i layout del circuito, la macchina a stati programmata come firmware nell'MCU e un metodo suggerito per testare il dispositivo. Viene inoltre presentato lo scenario del caso d'uso del sistema, i diagrammi di attività; il diagramma di classe del database, il diagramma entità-relazione, la descrizione tecnica delle tabelle e le query; e la panoramica degli ambienti delle app. Inoltre si spiegherà come è stato creato il sistema utilizzando un database e un'app per smartphone e come esso funzioni. Alla fine del documento, una discussione sull'applicabilità al mercato, i possibili miglioramenti e il lavoro futuro per il dispositivo progettato conclude il documento.
DESIGN OF A WRIST WEARABLE DEVICE FOR GESTURE RECOGNITION IN THE CONTEXT OF ADHERENCE MEASUREMENT
GONZÁLEZ RAMÍREZ, DIANA MARITZA
2018/2019
Abstract
Poor adherence in medical treatments is one of the biggest problems to tackle in the medical field. Poor adherence reduces clinical benefit provided by the treatment and increases the likelihood of hospitalization, especially in chronic illnesses, while good adherence increases the effectiveness of the treatment and promotes healthy lifestyles. This affects directly not only the quality of life of the patient but also the costs and quality of the healthcare system. Wearable technologies are a great tool to measure and promote medical adherence because they can monitor the at-home behavior of the patient and can also give reminders and feedback. This project addresses poor adherence by monitoring human activities through the movement of an arm, that could allow physicians to understand whether their patient is following the recommendations, by taking a pill. The approach we gave to this project is the design of a system that can acquire accelerometric signals from the forearm and send them to a smart-phone. In order to allow the data to be processed for gesture tracking, the device is connected to a smart-phone application that sends the data to the cloud. The designed system includes software and hardware. The wearable device thought for this application is a wrist band. The hardware design included a 6-axis IMU featuring an accelerometer and a gyroscope. This device performs an acquisition only when requested by the smartphone with which communicates via Bluetooth. Considering that the software for gesture tracking analysis (non developed in this thesis) is based on the knowledge of the gesture specific motion pattern that needs to be detected, customized to the specific patient, a training function is needed by which such gesture can be acquired in an univocal way in a dedicated session, the software design includes a systems interface that is adapted for two types of users: regular app user and expert user. Each type of user has specific interactions in the system. The expert user (physician) specifies the gestures to be tracked and the relevant training mode, meaning it can create and train gestures as well as managing the alarms for the regular app users. While the regular app user (patient) can train the already existing gestures and set alarms according to when the medication is needed to be taken. For storing all the data and information a cloud database was also designed. In the document, all the considerations are taken for the circuit design, the schematics and layouts of the circuit, the state machine that was programmed as firmware in the MCU, as well as a suggested method for testing the device are written. There is also presented the system's use case scenario, activity diagrams; the database class diagram, entity-relationship diagram, technical description of tables, and queries; and the app environments overview. And will explain overall how the system was built employing a database and a smartphone-app, and how this system works. At the end of the document, there is a discussion about the applicability in the market, the possible improvements, and future work for the designed device.| File | Dimensione | Formato | |
|---|---|---|---|
|
ThesisDianaMGR.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Testo della tesi
Dimensione
3.22 MB
Formato
Adobe PDF
|
3.22 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/164990