Crowdsourcing digital health measures to predict Parkinson's disease severity: the Parkinson's Disease Digital Biomarker DREAM Challenge.
Ontology highlight
ABSTRACT:
SUBMITTER: Sieberts SK
PROVIDER: S-EPMC7979931 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Every year, 11% of infants are born preterm with significant health consequences, with the vaginal microbiome a risk factor for preterm birth. We crowdsource models to predict (1) preterm birth (PTB; <37 weeks) or (2) early preterm birth (ePTB; <32 weeks) from 9 vaginal microbiome studies representing 3,578 samples from 1,268 pregnant individuals, aggregated from public raw data via phylogenetic harmonization. The predictive models are validated on two independent unpublished datasets representing 331 samples from 148 pregnant individuals. The top-performing models (among 148 and 121 submissions from 318 teams) achieve area under the receiver operator characteristic (AUROC) curve scores of 0.69 and 0.87 predicting PTB and ePTB, respectively. Alpha diversity, VALENCIA community state types, and composition are important features in the top-performing models, most of which are tree-based methods. This work is a model for translation of microbiome data into clinically relevant predictive models and to better understand preterm birth.
| S-EPMC10829755 | biostudies-literature
Project description:Globally, every year about 11% of infants are born preterm, defined as a birth prior to 37 weeks of gestation, with significant and lingering health consequences. Multiple studies have related the vaginal microbiome to preterm birth. We present a crowdsourcing approach to predict: (a) preterm or (b) early preterm birth from 9 publicly available vaginal microbiome studies representing 3,578 samples from 1,268 pregnant individuals, aggregated from raw sequences via an open-source tool, MaLiAmPi. We validated the crowdsourced models on novel datasets representing 331 samples from 148 pregnant individuals. From 318 DREAM challenge participants we received 148 and 121 submissions for our two separate prediction sub-challenges with top-ranking submissions achieving bootstrapped AUROC scores of 0.69 and 0.87, respectively. Alpha diversity, VALENCIA community state types, and composition (via phylotype relative abundance) were important features in the top performing models, most of which were tree based methods. This work serves as the foundation for subsequent efforts to translate predictive tests into clinical practice, and to better understand and prevent preterm birth.
| S-EPMC10029035 | biostudies-literature
Project description:One of the promising opportunities of digital health is its potential to lead to more holistic understandings of diseases by interacting with the daily life of patients and through the collection of large amounts of real-world data. Validating and benchmarking indicators of disease severity in the home setting is difficult, however, given the large number of confounders present in the real world and the challenges in collecting ground truth data in the home. Here we leverage two datasets collected from patients with Parkinson's disease, which couples continuous wrist-worn accelerometer data with frequent symptom reports in the home setting, to develop digital biomarkers of symptom severity. Using these data, we performed a public benchmarking challenge in which participants were asked to build measures of severity across 3 symptoms (on/off medication, dyskinesia, and tremor). 42 teams participated and performance was improved over baseline models for each subchallenge. Additional ensemble modeling across submissions further improved performance, and the top models validated in a subset of patients whose symptoms were observed and rated by trained clinicians.
| S-EPMC10047543 | biostudies-literature
Project description:BackgroundImpaired gait plays an important role for quality of life in patients with Huntington's disease (HD). Measuring objective gait parameters in HD might provide an unbiased assessment of motor deficits in order to determine potential beneficial effects of future treatments.ObjectiveTo objectively identify characteristic features of gait in HD patients using sensor-based gait analysis. Particularly, gait parameters were correlated to the Unified Huntington's Disease Rating Scale, total motor score (TMS), and total functional capacity (TFC).MethodsPatients with manifest HD at two German sites (n = 43) were included and clinically assessed during their annual ENROLL-HD visit. In addition, patients with HD and a cohort of age- and gender-matched controls performed a defined gait test (4 × 10 m walk). Gait patterns were recorded by inertial sensors attached to both shoes. Machine learning algorithms were applied to calculate spatio-temporal gait parameters and gait variability expressed as coefficient of variance (CV).ResultsStride length (- 15%) and gait velocity (- 19%) were reduced, while stride (+ 7%) and stance time (+ 2%) were increased in patients with HD. However, parameters reflecting gait variability were substantially altered in HD patients (+ 17% stride length CV up to + 41% stride time CV with largest effect size) and showed strong correlations to TMS and TFC (0.416 ≤ rSp ≤ 0.690). Objective gait variability parameters correlated with disease stage based upon TFC.ConclusionsSensor-based gait variability parameters were identified as clinically most relevant digital biomarker for gait impairment in HD. Altered gait variability represents characteristic irregularity of gait in HD and reflects disease severity.
| S-EPMC7293689 | biostudies-literature
Project description:Smartphones and wearables are widely recognised as the foundation for novel Digital Health Technologies (DHTs) for the clinical assessment of Parkinson's disease. Yet, only limited progress has been made towards their regulatory acceptability as effective drug development tools. A key barrier in achieving this goal relates to the influence of a wide range of sources of variability (SoVs) introduced by measurement processes incorporating DHTs, on their ability to detect relevant changes to PD. This paper introduces a conceptual framework to assist clinical research teams investigating a specific Concept of Interest within a particular Context of Use, to identify, characterise, and when possible, mitigate the influence of SoVs. We illustrate how this conceptual framework can be applied in practice through specific examples, including two data-driven case studies.
| S-EPMC9284971 | biostudies-literature
Project description:Therapeutic trials of disease-modifying agents in neurodegenerative disease typically require several hundred participants and long durations for clinical endpoints. Trials of this size are not feasible for prion diseases, rare dementia disorders associated with misfolding of prion protein. In this situation, biomarkers are particularly helpful. On diagnostic imaging, prion diseases demonstrate characteristic brain signal abnormalities on diffusion-weighted MRI. The aim of this study was to determine whether cerebral water diffusivity could be a quantitative imaging biomarker of disease severity. We hypothesized that the basal ganglia were most likely to demonstrate functionally relevant changes in diffusivity. Seventy-one subjects (37 patients and 34 controls) of whom 47 underwent serial scanning (23 patients and 24 controls) were recruited as part of the UK National Prion Monitoring Cohort. All patients underwent neurological assessment with the Medical Research Council Scale, a functionally orientated measure of prion disease severity, and diffusion tensor imaging. Voxel-based morphometry, voxel-based analysis of diffusion tensor imaging and regions of interest analyses were performed. A significant voxel-wise correlation of decreased Medical Research Council Scale score and decreased mean, radial and axial diffusivities in the putamen bilaterally was observed (P < 0.01). Significant decrease in putamen mean, radial and axial diffusivities over time was observed for patients compared with controls (P = 0.01), and there was a significant correlation between monthly decrease in putamen mean, radial and axial diffusivities and monthly decrease in Medical Research Council Scale (P < 0.001). Step-wise linear regression analysis, with dependent variable decline in Medical Research Council Scale, and covariates age and disease duration, showed the rate of decrease in putamen radial diffusivity to be the strongest predictor of rate of decrease in Medical Research Council Scale (P < 0.001). Sample size calculations estimated that, for an intervention study, 83 randomized patients would be required to provide 80% power to detect a 75% amelioration of decline in putamen radial diffusivity. Putamen radial diffusivity has potential as a secondary outcome measure biomarker in future therapeutic trials in human prion diseases.
| S-EPMC7425333 | biostudies-literature
Project description:Parkinson's disease (PD) is one of the first diseases where digital biomarkers demonstrated excellent performance in differentiating disease from healthy individuals. However, no study has systematically compared and leveraged multiple types of digital biomarkers to predict PD. Particularly, machine learning works on the fine-motor skills of PD are limited. Here, we developed deep learning methods that achieved an AUC (Area Under the receiver operator characteristic Curve) of 0.933 in identifying PD patients on 6418 individuals using 75048 tapping accelerometer and position records. Performance of tapping is superior to gait/rest and voice-based models obtained from the same benchmark population. Assembling the three models achieved a higher AUC of 0.944. Notably, the models not only correlated strongly to, but also performed better than patient self-reported symptom scores in diagnosing PD. This study demonstrates the complementary predictive power of tapping, gait/rest and voice data and establishes integrative deep learning-based models for identifying PD.
| S-EPMC8763910 | biostudies-literature
Project description:BackgroundAdoption of new digital measures for clinical trials and practice has been hindered by lack of actionable qualitative data demonstrating relevance of these metrics to people with Parkinson's disease.ObjectiveThis study evaluated of relevance of WATCH-PD digital measures to monitoring meaningful symptoms and impacts of early Parkinson's disease from the patient perspective.MethodsParticipants with early Parkinson's disease (N = 40) completed surveys and 1:1 online-interviews. Interviews combined: 1) symptom mapping to delineate meaningful symptoms/impacts of disease, 2) cognitive interviewing to assess content validity of digital measures, and 3) mapping of digital measures back to personal symptoms to assess relevance from the patient perspective. Content analysis and descriptive techniques were used to analyze data.ResultsParticipants perceived mapping as deeply engaging, with 39/40 reporting improved ability to communicate important symptoms and relevance of measures. Most measures (9/10) were rated relevant by both cognitive interviewing (70-92.5%) and mapping (80-100%). Two measures related to actively bothersome symptoms for more than 80% of participants (Tremor, Shape rotation). Tasks were generally deemed relevant if they met three participant context criteria: 1) understanding what the task measured, 2) believing it targeted an important symptom of PD (past, present, or future), and 3) believing the task was a good test of that important symptom. Participants did not require that a task relate to active symptoms or "real" life to be relevant.ConclusionDigital measures of tremor and hand dexterity were rated most relevant in early PD. Use of mapping enabled precise quantification of qualitative data for more rigorous evaluation of new measures.
| S-EPMC10357170 | biostudies-literature
Project description:Parkinson's disease (PD) results in progressively worsening gait and balance dysfunction that can be measured using computerized devices. We utilized the longitudinal database of the Parkinson's Disease Biomarker Program to determine if baseline gait and balance measures predict future rates of symptom progression. We included 230, 222, 164, and 177 PD subjects with 6, 12, 18, and 24 months of follow-up, respectively, and we defined progression as worsening of the following clinical parameters: MDS-UPDRS total score, Montreal Cognitive Assessment, PDQ-39 mobility subscale, levodopa equivalent daily dose, Schwab and England score, and global composite outcome. We developed ridge regression models to independently estimate how each gait or balance measure, or combination of measures, predicted progression. The accuracy of each ridge regression model was calculated by cross-validation in which 90% of the data were used to estimate the ridge regression model which was then tested on the 10% of data left out. While the models modestly predicted change in outcomes at the 6-month follow-up visit (accuracy in the range of 66-71%) there was no change in the outcome variables during this short follow-up (median change in MDS-UPDRS total score = 0 and change in LEDD = 0). At follow-up periods of 12, 18, and 24 months, the models failed to predict change (accuracy in the held-out sets ranged from 42 to 60%). We conclude that this set of computerized gait and balance measures performed at baseline is unlikely to help predict future disease progression in PD. Research scientists must continue to search for progression predictors to enhance the performance of disease modifying clinical trials.
| S-EPMC9681812 | biostudies-literature
Project description:Abnormalities of oculometric measures (OM) are widely described in people with Parkinson's disease (PD). However, knowledge of correlations between abnormal OM, disease severity and clinical assessment in PD patients is still lacking. To evaluate these correlations, PD patients (215 patients, mean age 69 ± 9.1 years, 79 females) with severe (H&Y > 3) and mild to moderate (H&Y ≤ 2) disease, and 215 age-matched healthy subjects were enrolled. All patients were evaluated using MDS-UPDRS and an oculometric test using computer vision and deep learning algorithms. Comparisons of OM between groups and correlations between OM and MDS-UPDRS scores were calculated. Saccadic latency (ms) was prolonged in patients with severe compared with mild to moderate disease (pro-saccades: 267 ± 69 vs. 238 ± 53, p = 0.0011; anti-saccades: 386 ± 119 vs. 352 ± 106, p = 0.0393) and in patients with mild to moderate disease versus healthy subjects (pro-saccades: 238 ± 53 vs. 220 ± 45, p = 0.0003; anti-saccades: 352 ± 106 vs. 289 ± 71, p < 0.0001). Error rate (%) was higher among patients with severe (64.06 ± 23.08) versus mild to moderate disease (49.84 ± 24.81, p = 0.0001), and versus healthy subjects (49.84 ± 24.81 vs. 28.31 ± 21.72, p = 0.00001). Response accuracy (%) was lower for patients with severe (75.66 ± 13.11) versus mild to moderate disease (79.66 ± 13.56, p = 0.0462), and versus healthy subjects (79.66 ± 13.56 vs. 90.27 ± 8.79, p < 0.0001). Pro- and anti-saccadic latency, error rate and accuracy were correlated with MDS-UPDRS scores (r = 0.32, 0.28, 0.36 and -0.30, respectively, p < 0.0001) and similar correlations were found with its axial subscore (R = 0.38, 0.29, 0.44, and -0.30, respectively, p < 0.0001). Several OM were different in patients under levodopa treatment. OM worsened as PD severity increases, and were correlated with MDS-UPDRS scores. Using OM can be implemented for PD patients' assessment as a tool to follow disease progression.
| S-EPMC10480268 | biostudies-literature