This research focuses on creating composite biomarkers that can classify diagnoses, estimate symptom severity, and detect treatment effects using data from wearable sensors and smartphone... Show moreThis research focuses on creating composite biomarkers that can classify diagnoses, estimate symptom severity, and detect treatment effects using data from wearable sensors and smartphone applications. The thesis consists of an introduction to machine learning techniques and their use in developing biomarkers for the central nervous system; a narrative review of the relevant literature; and detailed studies on the application of these techniques in various health conditions. Specifically, the research includes observational and cross-sectional studies on facioscapulohumeral muscular dystrophy (FSHD) and major depressive disorder (MDD), demonstrating how smartphone and wearable sensor data can be used to monitor disease severity and progression. Additionally, the research identified the use of a tablet-based finger tapping task to monitor the real-time effects of antiparkinson's drugs on Parkinson's symptom severity. Key findings highlight the potential of mHealth biomarkers to provide continuous, real-time monitoring of patients, which can enhance the accuracy of clinical assessments and potentially reduce the burden on patients and healthcare systems. The thesis also addresses the challenges of variability in mHealth device data and emphasizes the need for robust validation and standardization to ensure the reliability of these biomarkers in clinical settings. Show less
BACKGROUND:Pain evaluation remains largely subjective in neurosurgical practice, but machine learning provides the potential for objective pain assessment tools.OBJECTIVE:To predict daily pain... Show moreBACKGROUND:Pain evaluation remains largely subjective in neurosurgical practice, but machine learning provides the potential for objective pain assessment tools.OBJECTIVE:To predict daily pain levels using speech recordings from personal smartphones of a cohort of patients with diagnosed neurological spine disease.METHODS:Patients with spine disease were enrolled through a general neurosurgical clinic with approval from the institutional ethics committee. At-home pain surveys and speech recordings were administered at regular intervals through the Beiwe smartphone application. Praat audio features were extracted from the speech recordings to be used as input to a K-nearest neighbors (KNN) machine learning model. The pain scores were transformed from a 0 to 10 scale to low and high pain for better discriminative capacity.RESULTS:A total of 60 patients were enrolled, and 384 observations were used to train and test the prediction model. Using the KNN prediction model, an accuracy of 71% with a positive predictive value of 0.71 was achieved in classifying pain intensity into high and low. The model showed 0.71 precision for high pain and 0.70 precision for low pain. Recall of high pain was 0.74, and recall of low pain was 0.67. The overall F1 score was 0.73.CONCLUSION:Our study uses a KNN to model the relationship between speech features and pain levels collected from personal smartphones of patients with spine disease. The proposed model is a stepping stone for the development of objective pain assessment in neurosurgery clinical practice. Show less
Water is all around us and is vital for all aspects of life. Studying the various compounds and life forms that inhabit natural waters lets us better understand the world around us.Remote sensing... Show moreWater is all around us and is vital for all aspects of life. Studying the various compounds and life forms that inhabit natural waters lets us better understand the world around us.Remote sensing enables global measurements with rapid response and high consistency. Citizen science provides new knowledge and greatly increases the scientific and social impact of research.In this thesis, we investigate several aspects of citizen science and remote sensing of water, with a focus on uncertainty and accessibility. We improve existing techniques and develop new methods to use smartphone cameras for accessible remote sensing of water. Show less
Burggraaff, O.; Werther, M.; Boss E.S.; Simis, S.G.H.; Snik, F. 2022
Consumer cameras, especially on smartphones, are popular and effective instruments for above-water radiometry. The remote sensing reflectance Rrs is measured above the water surface and used to... Show moreConsumer cameras, especially on smartphones, are popular and effective instruments for above-water radiometry. The remote sensing reflectance Rrs is measured above the water surface and used to estimate inherent optical properties and constituent concentrations. Two smartphone apps, HydroColor and EyeOnWater, are used worldwide by professional and citizen scientists alike. However, consumer camera data have problems with accuracy and reproducibility between cameras, with systematic differences of up to 40% in intercomparisons. These problems stem from the need, until recently, to use JPEG data. Lossless data, in the RAW format, and calibrations of the spectral and radiometric response of consumer cameras can now be used to significantly improve the data quality. Here, we apply these methods to above-water radiometry. The resulting accuracy in Rrs is around 10% in the red, green, and blue (RGB) bands and 2% in the RGB band ratios, similar to professional instruments and up to 9 times better than existing smartphone-based methods. Data from different smartphones are reproducible to within measurement uncertainties, which are on the percent level. The primary sources of uncertainty are environmental factors and sensor noise. We conclude that using RAW data, smartphones and other consumer cameras are complementary to professional instruments in terms of data quality. We offer practical recommendations for using consumer cameras in professional and citizen science. Show less
Spectropolarimetry is a powerful technique for remote sensing of the environment. It enables the retrieval of particle shape and size distributions in air and water to an extent that traditional... Show moreSpectropolarimetry is a powerful technique for remote sensing of the environment. It enables the retrieval of particle shape and size distributions in air and water to an extent that traditional spectroscopy cannot. SPEX is an instrument concept for spectropolarimetry through spectral modulation, providing snapshot, and hence accurate, hyperspectral intensity and degree and angle of linear polarization. Successful SPEX instruments have included groundSPEX and SPEX airborne, which both measure aerosol optical thickness with high precision, and soon SPEXone, which will fly on PACE. Here, we present a low-cost variant for consumer cameras, iSPEX 2, with universal smartphone support. Smartphones enable citizen science measurements which are significantly more scaleable, in space and time, than professional instruments. Universal smartphone support is achieved through a modular hardware design and SPECTACLE data processing. iSPEX 2 will be manufactured through injection molding and 3D printing. A smartphone app for data acquisition and processing is in active development. Production, calibration, and validation will commence in the summer of 2020. Scientific applications will include citizen science measurements of aerosol optical thickness and surface water reflectance, as well as low-cost laboratory and portable spectroscopy. Show less
