Convolutional neural network based detection of early stage Parkinson’s disease using the six minute walk test

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Sept. 30, 2024

Language: Английский

---

Multimodal Digital Phenotyping of Behavior in a Neurology Clinic: Development of the Neurobooth Platform and the First Two Years of Data Collection

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 31, 2024

Abstract Quantitative analysis of human behavior is critical for objective characterization neurological phenotypes, early detection neurodegenerative diseases, and development more sensitive measures disease progression to support clinical trials translation new therapies into practice. Sophisticated computational modeling can these objectives, but requires large, information-rich data sets. This work introduces Neurobooth, a customizable platform time-synchronized multimodal capture behavior. Over two year period, Neurobooth implementation integrated setting facilitated collection across multiple behavioral domains from cohort 470 individuals (82 controls 388 with neurologic diseases) who participated in collective 782 sessions. Visualization the time series demonstrates presence rich phenotypic signs range diseases. These open-source offer potential advancing our understanding diseases facilitating therapy development, may be valuable resource related fields that study

Language: Английский

---

Comparative Assessment of Gait and Balance in Patients with Parkinson’s Disease and Normal Pressure Hydrocephalus

SiSli Etfal Hastanesi Tip Bulteni / The Medical Bulletin of Sisli Hospital, Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

Objectives:We aim to compare balance and gait parameters in patients diagnosed with Parkinson's disease (PD) normal pressure hydrocephalus (NPH).Methods: A total of 13 NPH, 20 PD, healthy controls (HC) recruited the study.Three IMU sensors (Ambulatory PD Monitoring Inc., OR, USA) were placed on lumbar area feet participants.The evaluations comprised eight successive standing tasks; modified clinical test sensory interaction test.These tasks involved apart eyes open as well closed a firm foam surface, together closed, tandem stance right foot front left front.Functional conducted using 10-M Walk Test (10 MWT), 2 min-Walk (2 timed-up go (TUG).Parameters analyzed then compared.Results: NPH displayed notable decrease both stride length speed compared tests revealed that group demonstrated significantly poorer performance, specifically feet-apart eyes-closed foam-surface test, test.During while surfaces, groups showed an increase root mean square sway, range, velocity (p<0.05) sway anteroposterior plane.In addition, during TUG exhibited significant prolongation time needed complete task decline turning but no difference was seen comparison HC group.Conclusion: Our study indicated notably worse measurements groups.These findings emphasize significance monitoring managing impairments patients.Sensor-based technologies may offer objective for more precise efficient follow-up these terms balance.

Language: Английский

---

Quantitative assessment of head movement dynamics in dystonia using visual perceptive deep learning: a multi-centre retrospective longitudinal cohort study

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Sept. 11, 2023

Abstract Background Dystonia is a neurological movement disorder characterised by abnormal involuntary movements and postures, particularly affecting the head neck. However, current clinical assessment methods for dystonia rely on simplified rating scales which lack ability to capture intricate spatiotemporal features of dystonic phenomena, hindering management limiting understanding underlying neurobiology. To address this, we developed visual perceptive deep learning framework that utilizes standard videos comprehensively evaluate quantify disease states impact therapeutic interventions, specifically brain stimulation. This overcomes limitations traditional offers an efficient accurate method rater-independent evaluating monitoring patients. Methods framework, leveraged semi-standardized video data collected in three retrospective, longitudinal cohort studies across seven academic centres Germany. We extracted static angle excursions validation derived kinematic variables reflecting naturalistic dynamics predict severity, subtype, neuromodulation effects. The was validated fully independent generalised Findings Computer vision-derived measurements showed strong correlation with clinically assigned scores, outperforming previous approaches employing specialised camera equipment. Across comparisons, discovered consistent set from full assessments, encoded information relevant effects neural circuit intervention more strongly independently deviations predominantly used scoring. Interpretation proposed machine reveals pathosignatures may be utilized augment management, facilitate scientific translation inform personalised precision Neurology. Research context Evidence before this study Clinical dystonia, disorder, has traditionally relied aim simplify complex phenomenology into lowerdimensional items. these score-based assessments have significant clinimetric do not rich are crucial judgment pathophysiological understanding. In contrast, recent investigations animal models already demonstrated utility relevance quantitative phenotyping, gradually supersedes observer-dependent behavioural analyses. Taken together, led need objective detailed evaluation dystonia. performed PubMed search up July 2023 combining terms “dystonia” AND (”deep learning” OR “machine or “computer vision” “vision-based” “video-based”) (”angle” “kinematic” “rating” “scoring” “movement analysis”) including abstracts English German. yielded vision-based frameworks automating cervical severity compared clinician-annotated ratings. Two focused deriving setups, while third utilised computer vision retrospective dataset recorded using conventional These reported fair moderately correlations between scores. Additionally, two investigated assessing tremor dystonia: one single case report validity metrics, cross-sectional agreement oscillation metrics different subgroups. additional visionbased kinematics dystonia-like phenomena rodent monogenetic demonstrating both phenotype genotype predictions. most were limited task conditions, where patients attempted hold neutral position head, thus providing account Moreover, beyond angular no explored broader feature space reflects true complexity movements. assessed at time points without considering therapy stimulation, highly effective targeting circuits. Nor did they compare sub-types, such as systonia. Added value study, present comprehensive addresses gaps assessments. use retrospectively analyse unique multi-centric, encompassing examinations along continuum, stimulation states. Our goes automation suboptimal symptom reverse engineering inspired features. resulting high dimensional, yet intuitively interpretable enabled us explore therapies level detail comparable experimental neuroscientific investigations. Through data-driven approach, identified only four dynamic parameters encode efficacy interventions. Notably, deviations, play central role pointing involvement partially distinct neurobiological processes captured findings align emerging concepts symptom-specific circuits thereby exemplifying framework’s potential bridge translational disorders research. By precise our valuable insights improved treatment strategies further dystonia’s Implications all available evidence collectively underscores capturing informative movements, emphasizing granular methods. line quantification, shown metrics. their designs inadvertently reinforce associated scoring process. introduce serves powerful platform judgement generate extracting inspired, implications showcasing enhancing fostering integration advanced neuroimaging neurotechnological opens doors future research application techniques derive signatures species promising can significantly improve patient outcomes.

Language: Английский

---

Convolutional neural network-based detection of early-stage Parkinson’s disease using the six-minute walk test

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: June 12, 2024

The heterogeneity of Parkinson’s disease (PD) generates significant challenges for accurate diagnosis, especially in early-stage disease, when symptoms may be very subtle. This study aimed to determine the accuracy a convolutional neural network (CNN) technique based on 6-min walk test (6MWT) using wearable sensors distinguishing patients with PD (n = 78) from healthy controls 50). Wearing six sensors, participants performed 6MWT, and time-series data were converted into new images. main results showed that gyroscopic vertical component lumbar spine had highest classification 83.5%, followed by thoracic (83.1%) right thigh (79.5%) segment. These suggest 6MWT CNN models pave way clinicians diagnose track earlier thus provide timely treatment during golden transition geriatric pathologic gait patterns.

Language: Английский

---

Kinematic IMU-Based Assessment of Postural Transitions: A Preliminary Application in Clinical Context

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(16), P. 7011 - 7011

Published: Aug. 9, 2024

This study aims to develop a new methodology for assessing postural transitions, such as sit-to-stand movements, and preliminarily apply it in clinical setting. These movements provide valuable information about the state of movement effector system components, whether musculoskeletal, nervous, or cognitive, their evaluation is key point functional assessment setting patients with complex rehabilitative needs. The objective this was developed by pursuing three goals: verifying ability discriminate between healthy pathological subjects, defining set parameters assessment, thus designing preliminary paradigm future applications. We investigated signals from single IMU sensor applied subjects (20 13 patients) performing five different transitions. A six kinematic variables that allowed quantitative motion identified, namely total time, smoothness, fluency, velocity, jerk root mean square, maximum variation. At end study, adopted were shown be able quantitatively assess transitions context distinguish subjects. This, together studies, will researchers clinicians resource evaluating results rehabilitation program, well keeping track patients’ status follow-up evaluations.

Language: Английский

---

A Statistical Approach for Functional Reach-to-Grasp Segmentation Using a Single Inertial Measurement Unit

Sensors, Journal Year: 2024, Volume and Issue: 24(18), P. 6119 - 6119

Published: Sept. 22, 2024

The aim of this contribution is to present a segmentation method for the identification voluntary movements from inertial data acquired through single measurement unit placed on subject’s wrist. Inertial were recorded 25 healthy subjects while performing 75 consecutive reach-to-grasp movements. approach herein presented, called DynAMoS, based an adaptive thresholding step angular velocity norm, followed by statistics-based post-processing movement duration distribution. Post-processing aims at reducing number erroneous transitions in segmentation. We assessed quality using stereophotogrammetric system as gold standard. Two popular methods already presented literature compared DynAMoS terms identified, onset and offset mean absolute errors, duration. Moreover, we analyzed sub-phase durations drinking further characterize task. results show that proposed performs significantly better than two state-of-the-art approaches (i.e., percentage = 3%; error < 0.08 s), suggesting could make more effective home monitoring applications assessing motion improvements patients following domicile rehabilitation protocols.

Language: Английский

---

Hand Tremor Detection and Feedback System Based on Inertial Measurement Unit (IMU) for Parkinson's Disease: A Pilot Study

Journal of Mechanics in Medicine and Biology, Journal Year: 2024, Volume and Issue: 24(09)

Published: Sept. 27, 2024

The hand tremor symptoms of Parkinson’s disease (PD) patients are expressed in a wide variety forms. These tremors not limited to specific areas, but the most experienced by PD hand-related. To detect such symptoms, device based on an inertial measurement unit (IMU) was used measure changes vibration speed around wrist. sensor data collected from gyroscope through dedicated algorithm can be reflected convert it into result value number user, and characteristics extracted. performance system validated using simulator. experiments confirmed that operated normally under conditions set (5–18[Formula: see text]Hz) for detection, operation results matched those theoretically detectable range. wearable user’s bio-kinetic signals daily life gyroscope, adjust algorithm’s sensitivity severity continuously monitor transmitting recognition via Bluetooth communication.

Language: Английский

---

Biodegradable, Self‐Adhesive, Stretchable, Transparent, and Versatile Electronic Skins Based on Intrinsically Hydrophilic Poly(Caproactone‐Urethane) Elastomer

Advanced Engineering Materials, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 25, 2024

In biomedical sciences, there is a demand for electronic skins with highly sensitive tactile sensors that have applications in patient monitoring, human–machine interfaces, and on‐body sensors. Sensor fabrication requires high‐performance conductive surfaces are transparent, breathable, flexible, easy to fabricate. It also preferable if the electrodes easily processable as wastes, example, degradable. this work, design of hydrophilic silanol/amine‐terminated poly(caprolactone‐urethane) (SA‐PCLU) elastomer‐based stretchable, biodegradable reported. Ag nanowires dispersed water sprayed onto intrinsically electrospun SA‐PCLU became embedded into scaffold formed conformal polyurethane‐based networks (HPCN). The used fabricate capacitive, curvature, strain sensors, all having monomaterial composition. addition displaying particularly good transparencies at low sheet resistances, stretchability, hydrophilicity, tight bonding target surface, allow evaporation perspiration, making them suitable epidermal long‐time use. application HPCN flexible electronics bionic skin demonstrated through gesture monitoring experiments swelling

Language: Английский

---

Feasibility of Shoulder Kinematics Assessment Using Magnetic Inertial Measurement Units in Hemiplegic Patients after Stroke: A Pilot Study

Applied Sciences, Journal Year: 2023, Volume and Issue: 13(21), P. 11900 - 11900

Published: Oct. 31, 2023

Scapulothoracic movements are altered after stroke, with resulting shoulder dysfunction. The scapulohumeral rhythm (SHR) is complex and poorly studied. Magnetic inertial measurement units (MIMUs) allow a rapid accurate analysis of kinematics. MIMUs were used to assess the SHR during active flexion abduction over 60°. values obtained from hemiplegic shoulders stroke patients (n = 7) compared those healthy controls 25) correlated clinical–functional measurements. impairment paretic arms was assessed using Fugl-Meyer Assessment (FMA). We found that in shoulders, scapular tilt significantly lower at maximal arm 60° 90° abduction. On side, also consistently for all measured movements. FMA anterior–posterior (Rho 0.847, p 0.016, Rho 0.757, 0.049, respectively). This pilot study demonstrates feasibility assessing confirms previous findings on dysfunction patients.

Language: Английский

---