Digital Biomarkers最新文献

Introduction: We investigated the possibilities and opportunities for using wearable devices that measure physical activity and physiometric signals in conjunction with ecological momentary assessment (EMA) data to improve the assessment and treatment of pain.

Methods: We considered studies with cross-sectional and longitudinal designs as well as interventional or observational studies correlating pain scores with measures derived from wearable devices. A search was also performed on studies that investigated physical activity and physiometric signals among patients with pain.

Results: Few studies have assessed the possibility of incorporating wearable devices as objective tools for contextualizing pain and physical function in free-living environments. Of the studies that have been conducted, most focus solely on physical activity and functional outcomes as measured by a wearable accelerometer. Several studies report promising correlations between pain scores and signals derived from wearable devices, objectively measured physical activity, and physical function. In addition, there is a known association between physiologic signals that can be measured by wearable devices and pain, though studies using wearable devices to measure these signals and associate them with pain in free-living environments are limited.

Conclusion: There exists a great opportunity to study the complex interplay between physiometric signals, physical function, and pain in a real-time fashion in free-living environments. The literature supports the hypothesis that wearable devices can be used to develop reproducible biosignals that correlate with pain. The combination of wearable devices and EMA will likely lead to the development of clinically meaningful endpoints that will transform how we understand and treat pain patients.

At the end of 2020, Karger's Digital Biomarkers, together with Evidation Health, produced a special issue entitled "The Future of Digital Health." This brief meeting report provides an overview of the expert panel and workshop that were held in early 2021 to explore key topics raised in the special issue.

Diseases can affect organs such as the heart, lungs, brain, muscles, or vocal folds, which can then alter an individual's voice. Therefore, voice analysis using artificial intelligence opens new opportunities for healthcare. From using vocal biomarkers for diagnosis, risk prediction, and remote monitoring of various clinical outcomes and symptoms, we offer in this review an overview of the various applications of voice for health-related purposes. We discuss the potential of this rapidly evolving environment from a research, patient, and clinical perspective. We also discuss the key challenges to overcome in the near future for a substantial and efficient use of voice in healthcare.

Introduction: The effect of rhythmic acoustic stimuli on body sway is of increasing interest due to their positive contribution when training or restoring the control of movement. Inertial sensors show promise as a portable, easier, and more affordable method compared to the force plate "gold standard" concerning the evaluation of postural sway. This study examined the concurrent validity of inertially sensed measures of voluntary body sway against those obtained with a force plate, in silence and while exposed to a rhythmic acoustic stimulus.

Methods: Temporal (sway duration and variability) and spatial (trajectory length, variability, range, velocity, and area) body sway variables were extracted using an inertial sensor (at L5) in synchronization with a force plate, during anteroposterior body sway in silence and while exposed to a rhythmic acoustic stimulus (n = 18 young women; two 70-s trials in each condition). Statistics included bivariate correlations between the inertially sensed and the force plate measures, separately, in silence and with a rhythmic acoustic stimulus, as well as for the effect of the rhythmic acoustic stimulus (percentage difference from silence) (p ≤ 0.05, SPSS v25.0).

Results: The inertially sensed measures demonstrated good-to-excellent concurrent validity for all temporal and almost all spatial variables, both in silence and with rhythmic acoustic stimulus (r > 0.75, p = 0.000), as well as for the rhythmic acoustic-stimulus effect (r > 0.75, p ≤ 0.05).

Conclusion: The inertially sensed measures of the voluntary anteroposterior body sway demonstrated an overall good-to-excellent concurrent validity against those obtained with the force plate "gold standard," both in the silence and the rhythmic acoustic stimulus conditions, as well as for the rhythmic acoustic-stimulus effect. The findings of this pilot study allow the recommendation of inertial sensing for the evaluation of postural control alterations when exposed to rhythmic acoustic stimuli, a condition of increasing interest due to the positive contribution of such stimuli when training or restoring the control of movement.

To support the successful adoption of digital measures into internal decision making and evidence generation for medical product development, we present a unified lexicon to aid communication throughout this process, and highlight key concepts including the critical role of participant engagement in development of digital measures. We detail the steps of bringing a successful proof of concept to scale, focusing on key decisions in the development of a new digital measure: asking the right question, optimized approaches to evaluating new measures, and whether and how to pursue qualification or acceptance. Building on the V3 framework for establishing verification and analytical and clinical validation, we discuss strategic and practical considerations for collecting this evidence, illustrated with concrete examples of trailblazing digital measures in the field.

Background: Mild cognitive impairment (MCI) is a condition that entails a slight yet noticeable decline in cognition that exceeds normal age-related changes. Older adults living with MCI have a higher chance of progressing to dementia, which warrants regular cognitive follow-up at memory clinics. However, due to time and resource constraints, this follow-up is conducted at separate moments in time with large intervals in between. Casual games, embedded into the daily life of older adults, may prove to be a less resource-intensive medium that yields continuous and rich data on a patient's cognition.

Objective: To explore whether digital biomarkers of cognitive performance, found in the casual card game Klondike Solitaire, can be used to train machine-learning models to discern games played by older adults living with MCI from their healthy counterparts.

Methods: Digital biomarkers of cognitive performance were captured from 23 healthy older adults and 23 older adults living with MCI, each playing 3 games of Solitaire with 3 different deck shuffles. These 3 deck shuffles were identical for each participant. Using a supervised stratified, 5-fold, cross-validated, machine-learning procedure, 19 different models were trained and optimized for F1 score.

Results: The 3 best performing models, an Extra Trees model, a Gradient Boosting model, and a Nu-Support Vector Model, had a cross-validated F1 training score on the validation set of ≥0.792. The F1 score and AUC of the test set were, respectively, >0.811 and >0.877 for each of these models. These results indicate psychometric properties comparative to common cognitive screening tests.

Conclusion: The results suggest that commercial card games, not developed to address specific mental processes, may be used for measuring cognition. The digital biomarkers derived from Klondike Solitaire show promise and may prove useful to fill the current blind spot between consultations.

Introduction: Real-time digital heart rate (HR) monitoring in sports can provide unique physiological insights into athletic performance. However, most HR monitoring of elite athletes is limited to non-real-time, non-competition settings while utilizing sensors that are cumbersome. The present study was undertaken to test the feasibility of using small, wearable medical-grade sensors, paired with a novel technology system, to capture and process real-time HR data from elite athletes during professional competition.

Methods: We examined the performance of the BioStamp nPoint® sensor compared to the Polar chest strap HR sensor in 15 Professional Squash Association (PSA) tournament matches in 2019-2020. Fourteen male professional squash players volunteered for the study (age = 23.8 ± 4.9 years; height = 177.9 ± 7.1 cm; weight = 71 ± 7.0 kg), which was approved by the PSA in accordance with their Code of General Conduct and Ethics. Algorithms developed by Sports Data Labs (SDL; Detroit, MI, USA) used proprietary data collection, transmission, and signal processing protocols to produce HR values in real-time during matches. We calculated the mean and maximum HR from both sensors and used widely accepted measures of agreement to compare their performance.

Results: The system captured 99.8% of HR data across all matches (range 98.3-100%). The BioStamp's mean HR was 170.4 ± 20.3 bpm, while the Polar's mean HR was 169.4 ± 21.7 bpm. Maximum HR ranged from 182 to 202 bpm (Polar) and 185 to 203 bpm (BioStamp). Spearman's correlation coefficient (r _s) was 0.986 (p < 0.001), indicating a strong correlation between the 2 devices. The mean difference (d) in HR was 1.0 bpm, the mean absolute error was 2.2 bpm, and the percent difference was 0.72%, demonstrating high agreement between device measurements.

Conclusions: It is feasible to accurately measure and monitor real-time HR in elite athletes during competition using BioStamp's and SDL's proprietary system. This system facilitates development and understanding of physiological digital biomarkers of athletic performance and physical and psychosocial demands in elite athletic competition.

Introduction: Motor abnormalities have been shown to be a distinct component of schizophrenia symptomatology. However, objective and scalable methods for assessment of motor functioning in schizophrenia are lacking. Advancements in machine learning-based digital tools have allowed for automated and remote "digital phenotyping" of disease symptomatology. Here, we assess the performance of a computer vision-based assessment of motor functioning as a characteristic of schizophrenia using video data collected remotely through smartphones.

Methods: Eighteen patients with schizophrenia and 9 healthy controls were asked to remotely participate in smartphone-based assessments daily for 14 days. Video recorded from the smartphone front-facing camera during these assessments was used to quantify the Euclidean distance of head movement between frames through a pretrained computer vision model. The ability of head movement measurements to distinguish between patients and healthy controls as well as their relationship to schizophrenia symptom severity as measured through traditional clinical scores was assessed.

Results: The rate of head movement in participants with schizophrenia (1.48 mm/frame) and those without differed significantly (2.50 mm/frame; p = 0.01), and a logistic regression demonstrated that head movement was a significant predictor of schizophrenia diagnosis (p = 0.02). Linear regression between head movement and clinical scores of schizophrenia showed that head movement has a negative relationship with schizophrenia symptom severity (p = 0.04), primarily with negative symptoms of schizophrenia.

Conclusions: Remote, smartphone-based assessments were able to capture meaningful visual behavior for computer vision-based objective measurement of head movement. The measurements of head movement acquired were able to accurately classify schizophrenia diagnosis and quantify symptom severity in patients with schizophrenia.

Background: Assuring adequate antibiotic tissue concentrations at the point of infection, especially in skin and soft tissue infections, is pivotal for an effective treatment and cure. Despite the global issue, a reliable AB monitoring test is missing. Inadequate antibiotic treatment leads to the development of antimicrobial resistances and toxic side effects. β-lactam antibiotics were already detected in sweat of patients treated with the respective antibiotics intravenously before. With the emergence of smartphone-based biosensors to analyse sweat on the spot of need, next-generation molecular digital biomarkers will be increasingly available for a non-invasive pharmacotherapy monitoring.

Objective: Here, we investigated if the glycopeptide antibiotic vancomycin is detectable in sweat samples of in-patients treated with intravenous vancomycin.

Methods: Eccrine sweat samples were collected using the Macroduct Sweat Collector®. Along every sweat sample, a blood sample was taken. Bio-fluid analysis was performed by Ultra-high Pressure Liquid Chromatograph-Tandem Quadrupole Mass Spectrometry coupled with tandem mass spectrometry.

Results: A total of 5 patients were included. Results demonstrate that vancomycin was detected in 5 out of 5 sweat samples. Specifically, vancomycin concentrations ranged from 0.011 to 0.118 mg/L in sweat and from 4.7 to 8.5 mg/L in blood.

Conclusion: Our results serve as proof-of-concept that vancomycin is detectable in eccrine sweat and may serve as a surrogate marker for antibiotic tissue penetration. A targeted vancomycin treatment is crucial in patients with repetitive need for antibiotics and a variable antibiotic distribution such as in peripheral artery disease to optimize treatment effectiveness. If combined with on-skin smartphone-based biosensors and smartphone applications, the detection of antibiotic concentrations in sweat might enable a first digital, on-spot, lab-independent and non-invasive therapeutic drug monitoring in skin and soft tissue infections.

Background: Alterations in multiple domains of cognition have been observed in individuals who have experienced a traumatic stressor. These domains may provide important insights in identifying underlying neurobiological dysfunction driving an individual's clinical response to trauma. However, such assessments are burdensome, costly, and time-consuming. To overcome barriers, efforts have emerged to measure multiple domains of cognitive functioning through the application of machine learning (ML) models to passive data sources.

Methods: We utilized automated computer vision and voice analysis methods to extract facial, movement, and speech characteristics from semi-structured clinical interviews in 81 trauma survivors who additionally completed a cognitive assessment battery. A ML-based regression framework was used to identify variance in visual and auditory measures that relate to multiple cognitive domains.

Results: Models derived from visual and auditory measures collectively accounted for a large variance in multiple domains of cognitive functioning, including motor coordination (R² = 0.52), processing speed (R² = 0.42), emotional bias (R² = 0.52), sustained attention (R² = 0.51), controlled attention (R² = 0.44), cognitive flexibility (R² = 0.43), cognitive inhibition (R² = 0.64), and executive functioning (R² = 0.63), consistent with the high test-retest reliability of traditional cognitive assessments. Face, voice, speech content, and movement have all significantly contributed to explaining the variance in predicting functioning in all cognitive domains.

Conclusions: The results demonstrate the feasibility of automated measurement of reliable proxies of cognitive functioning through low-burden passive patient evaluations. This makes it easier to monitor cognitive functions and to intervene earlier and at a lower threshold without requiring a time-consuming neurocognitive assessment by, for instance, a licensed psychologist with specialized training in neuropsychology.