Search

Accurate tremor classification is crucial for effective patient management and treatment. However, clinical diagnoses are often hindered by misdiagnoses, necessitating the development of robust technical methods. Here, we present a two-stage convolutional neural network (CNN)-based system for classifying physiological tremor, essential tremor (ET), and Parkinson’s disease (PD) tremor. Employing acceleration signals from the hands of 408 patients, our system utilizes both medically motivated signal features and (nearly) raw data (by means of spectrograms) as system inputs. Our model employs a hybrid approach of data-based and feature-based methods to leverage the strengths of both while mitigating their weaknesses. By incorporating various data augmentation techniques for model training, we achieved an overall accuracy of 88.12%. This promising approach demonstrates improved accuracy in discriminating between the three tremor types, paving the way for more precise tremor diagnosis and enhanced patient care.

Background Veterans transitioning to civilian life often have chronic pain from service-related musculoskeletal disorders (MSD) with higher risk for substance misuse. Many seek VA (Department of Veterans Affairs) compensation for MSD. Use of Screening, Brief Intervention, and Referral to Treatment for Pain Management (SBIRT-PM) by VA Post-9/11 Military2VA (M2VA) case managers presents an opportunity to engage these veterans in VA pain care and address substance misuse. Implementation facilitation might help case managers use SBIRT-PM and engage veterans in services to improve outcomes. Design This study is a 2-cohort multisite cluster-randomized hybrid type 2 effectiveness-implementation trial. Within 2 separate cohorts of 14 VA sites each, sites will be allocated to receive an implementation strategy through the use of a constrained randomization procedure: virtual implementation facilitation or training-as-usual. Sites and M2VA case managers will receive the assigned implementation strategy to support use of SBIRT-PM. Recently discharged veterans (n  = 1848) claiming service-connected MSD will be recruited, with case managers blind to veterans' study enrollment. The proportion of participants who receive any SBIRT-PM will be the primary implementation outcome. Veteran participants will complete baseline, 12-week, and 36-week assessments, irrespective of whether case managers conduct SBIRT-PM with them (intent-to-treat). Pain intensity and interference will be the primary clinical outcomes. The study emphasizes pragmatic over explanatory methodological features. Summary This pragmatic trial will examine implementation facilitation versus training-as-usual in implementing SBIRT-PM to promote veteran engagement in nonpharmacological pain services. Using innovative methods to train and support VA case managers in SBIRT-PM, study outcomes could have broad implications for case management systems of care across the VA. [ABSTRACT FROM AUTHOR]

Accurate tremor classification is crucial for effective patient management and treatment. However, clinical diagnoses are often hindered by misdiagnoses, necessitating the development of robust technical methods. Here, we present a two-stage convolutional neural network (CNN)-based system for classifying physiological tremor, essential tremor (ET), and Parkinson's disease (PD) tremor. Employing acceleration signals from the hands of 408 patients, our system utilizes both medically motivated signal features and (nearly) raw data (by means of spectrograms) as system inputs. Our model employs a hybrid approach of data-based and feature-based methods to leverage the strengths of both while mitigating their weaknesses. By incorporating various data augmentation techniques for model training, we achieved an overall accuracy of 88.12%. This promising approach demonstrates improved accuracy in discriminating between the three tremor types, paving the way for more precise tremor diagnosis and enhanced patient care. [ABSTRACT FROM AUTHOR]