-
JMIR Rehabilitation and Assistive... Jun 2024Impaired cognitive function is observed in many pathologies, including neurodegenerative diseases such as Alzheimer disease. At present, the pharmaceutical treatments...
BACKGROUND
Impaired cognitive function is observed in many pathologies, including neurodegenerative diseases such as Alzheimer disease. At present, the pharmaceutical treatments available to counter cognitive decline have only modest effects, with significant side effects. A nonpharmacological treatment that has received considerable attention is computerized cognitive training (CCT), which aims to maintain or improve cognitive functioning through repeated practice in standardized exercises. CCT allows for more regular and thorough training of cognitive functions directly at home, which represents a significant opportunity to prevent and fight cognitive decline. However, the presence of assistance during training seems to be an important parameter to improve patients' motivation and adherence to treatment. To compensate for the absence of a therapist during at-home CCT, a relevant option could be to include a virtual assistant to accompany patients throughout their training.
OBJECTIVE
The objective of this exploratory study was to evaluate the interest of including a virtual assistant to accompany patients during CCT. We investigated the relationship between various individual factors (eg, age, psycho-affective functioning, personality, personal motivations, and cognitive skills) and the appreciation and usefulness of a virtual assistant during CCT. This study is part of the THERADIA (Thérapies Digitales Augmentées par l'Intelligence Artificielle) project, which aims to develop an empathetic virtual assistant.
METHODS
A total of 104 participants were recruited, including 52 (50%) young adults (mean age 21.2, range 18 to 27, SD 2.9 years) and 52 (50%) older adults (mean age 67.9, range 60 to 79, SD 5.1 years). All participants were invited to the laboratory to answer several questionnaires and perform 1 CCT session, which consisted of 4 cognitive exercises supervised by a virtual assistant animated by a human pilot via the Wizard of Oz method. The participants evaluated the virtual assistant and CCT at the end of the session.
RESULTS
Analyses were performed using the Bayesian framework. The results suggest that the virtual assistant was appreciated and perceived as useful during CCT in both age groups. However, older adults rated the assistant and CCT more positively overall than young adults. Certain characteristics of users, especially their current affective state (ie, arousal, intrinsic relevance, goal conduciveness, and anxiety state), appeared to be related to their evaluation of the session.
CONCLUSIONS
This study provides, for the first time, insight into how young and older adults perceive a virtual assistant during CCT. The results suggest that such an assistant could have a beneficial influence on users' motivation, provided that it can handle different situations, particularly their emotional state. The next step of our project will be to evaluate our device with patients experiencing mild cognitive impairment and to test its effectiveness in long-term cognitive training.
PubMed: 38901017
DOI: 10.2196/48129 -
Proceedings of the National Academy of... Jun 2024Proteomics has been revolutionized by large protein language models (PLMs), which learn unsupervised representations from large corpora of sequences. These models are...
Proteomics has been revolutionized by large protein language models (PLMs), which learn unsupervised representations from large corpora of sequences. These models are typically fine-tuned in a supervised setting to adapt the model to specific downstream tasks. However, the computational and memory footprint of fine-tuning (FT) large PLMs presents a barrier for many research groups with limited computational resources. Natural language processing has seen a similar explosion in the size of models, where these challenges have been addressed by methods for parameter-efficient fine-tuning (PEFT). In this work, we introduce this paradigm to proteomics through leveraging the parameter-efficient method LoRA and training new models for two important tasks: predicting protein-protein interactions (PPIs) and predicting the symmetry of homooligomer quaternary structures. We show that these approaches are competitive with traditional FT while requiring reduced memory and substantially fewer parameters. We additionally show that for the PPI prediction task, training only the classification head also remains competitive with full FT, using five orders of magnitude fewer parameters, and that each of these methods outperform state-of-the-art PPI prediction methods with substantially reduced compute. We further perform a comprehensive evaluation of the hyperparameter space, demonstrate that PEFT of PLMs is robust to variations in these hyperparameters, and elucidate where best practices for PEFT in proteomics differ from those in natural language processing. All our model adaptation and evaluation code is available open-source at https://github.com/microsoft/peft_proteomics. Thus, we provide a blueprint to democratize the power of PLM adaptation to groups with limited computational resources.
Topics: Proteomics; Proteins; Natural Language Processing; Protein Interaction Mapping; Computational Biology; Humans; Algorithms
PubMed: 38900798
DOI: 10.1073/pnas.2405840121 -
Journal of Rehabilitation Medicine Jun 2024To explore how people with stroke, discharged to skilled nursing facilities before returning home, experience the chain of care and rehabilitation.
OBJECTIVE
To explore how people with stroke, discharged to skilled nursing facilities before returning home, experience the chain of care and rehabilitation.
DESIGN
Qualitative, semi-structured interview design.
METHODS
Thirteen stroke survivors discharged from a stroke unit to a skilled nursing facility before returning to independent living participated. Semi-structured telephone interviews were conducted 2-5 months after stroke and analysed with content analysis.
RESULTS
The analysis resulted in three categories, Organizational processes, critical and complex, Rehabilitation, the right support at the right time and Adaptation to the changed situation, with a total of 9 subcategories. The informants perceived low participation in planning and goalsetting and limited information. Support from the healthcare services was important to proceed with improvements although the amount of supported training varied. Factors hindering and facilitating managing everyday life were described, as well as lingering uncertainty of what the future would be like.
CONCLUSION
Support and rehabilitation as well as individuals' needs varied, throughout the chain of care. To enable participation in the rehabilitation, assistance in setting goals and repeated information is warranted. Tailored care and rehabilitation throughout the chain of care should be provided, followed up at home, and coordinated for smooth transitions between organizations.
Topics: Humans; Stroke Rehabilitation; Skilled Nursing Facilities; Female; Male; Patient Discharge; Aged; Middle Aged; Qualitative Research; Aged, 80 and over; Stroke; Continuity of Patient Care
PubMed: 38899476
DOI: 10.2340/jrm.v56.35240 -
Frontiers in Psychology 2024Although Cognitive Behavioral Therapy (CBT) is the most often used intervention in forensic treatment, its effectivity is not consistently supported. Interventions...
INTRODUCTION
Although Cognitive Behavioral Therapy (CBT) is the most often used intervention in forensic treatment, its effectivity is not consistently supported. Interventions incorporating knowledge from neuroscience could provide for more successful intervention methods.
METHODS
The current pilot study set out to assess the feasibility and usability of the study protocol of a 4-week neuromeditation training in adult forensic outpatients with impulse control problems. The neuromeditation training, which prompts awareness and control over brain states of restlessness with EEG neurofeedback, was offered in addition to treatment as usual (predominantly CBT).
RESULTS
Eight patients completed the neuromeditation training under guidance of their therapists. Despite some emerging obstacles, overall, the training was rated sufficiently usable and feasible by patients and their therapists.
DISCUSSION
The provided suggestions for improvement can be used to implement the intervention in treatment and set up future trials to study the effectiveness of neuromeditation in offender treatment.
PubMed: 38899124
DOI: 10.3389/fpsyg.2024.1354997 -
PCN Reports : Psychiatry and Clinical... Jun 2024
PubMed: 38899052
DOI: 10.1002/pcn5.219 -
BMC Medical Genomics Jun 2024Immunoregulatory drugs regulate the ubiquitin-proteasome system, which is the main treatment for multiple myeloma (MM) at present. In this study, bioinformatics analysis...
BACKGROUND
Immunoregulatory drugs regulate the ubiquitin-proteasome system, which is the main treatment for multiple myeloma (MM) at present. In this study, bioinformatics analysis was used to construct the risk model and evaluate the prognostic value of ubiquitination-related genes in MM.
METHODS AND RESULTS
The data on ubiquitination-related genes and MM samples were downloaded from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. The consistent cluster analysis and ESTIMATE algorithm were used to create distinct clusters. The MM prognostic risk model was constructed through single-factor and multiple-factor analysis. The ROC curve was plotted to compare the survival difference between high- and low-risk groups. The nomogram was used to validate the predictive capability of the risk model. A total of 87 ubiquitination-related genes were obtained, with 47 genes showing high expression in the MM group. According to the consistent cluster analysis, 4 clusters were determined. The immune infiltration, survival, and prognosis differed significantly among the 4 clusters. The tumor purity was higher in clusters 1 and 3 than in clusters 2 and 4, while the immune score and stromal score were lower in clusters 1 and 3. The proportion of B cells memory, plasma cells, and T cells CD4 naïve was the lowest in cluster 4. The model genes KLHL24, HERC6, USP3, TNIP1, and CISH were highly expressed in the high-risk group. AICAr and BMS.754,807 exhibited higher drug sensitivity in the low-risk group, whereas Bleomycin showed higher drug sensitivity in the high-risk group. The nomogram of the risk model demonstrated good efficacy in predicting the survival of MM patients using TCGA and GEO datasets.
CONCLUSIONS
The risk model constructed by ubiquitination-related genes can be effectively used to predict the prognosis of MM patients. KLHL24, HERC6, USP3, TNIP1, and CISH genes in MM warrant further investigation as therapeutic targets and to combat drug resistance.
Topics: Humans; Multiple Myeloma; Computational Biology; Prognosis; Ubiquitination; Gene Expression Regulation, Neoplastic; Biomarkers, Tumor; Nomograms; Cluster Analysis
PubMed: 38898455
DOI: 10.1186/s12920-024-01937-0 -
Archives of Osteoporosis Jun 2024Interviews and focus groups with patients, FLS clinicians, and GPs identified challenges relating to clinical and shared decision-making about bone health and...
UNLABELLED
Interviews and focus groups with patients, FLS clinicians, and GPs identified challenges relating to clinical and shared decision-making about bone health and osteoporosis medicines. Findings will inform the development of the multicomponent iFraP intervention to address identified training needs and barriers to implementation to facilitate SDM about osteoporosis medicines.
PURPOSE
The iFraP (improving uptake of Fracture Prevention treatments) study aimed to develop a multicomponent intervention, including an osteoporosis decision support tool (DST), to support shared decision-making (SDM) about osteoporosis medicines. To inform iFraP intervention development, this qualitative study explored current practice in relation to communication about bone health and osteoporosis medicines, anticipated barriers to, and facilitators of, an osteoporosis DST, and perceived training needs.
METHODS
Patients attending an FLS consultation (n = 8), FLS clinicians (n = 9), and general practitioners (GPs; n = 7) were purposively sampled to participate in a focus group and/or telephone interview. Data were transcribed, inductively coded, and then mapped to the Theoretical Domains Framework (TDF) as a deductive framework to systematically identify possible barriers to, and facilitators of, implementing a DST.
RESULTS
Inductive codes were deductively mapped to 12 TDF domains. FLS clinicians were perceived to have specialist expertise (knowledge). However, clinicians described aspects of clinical decision-making and risk communication as difficult (cognitive skills). Patients reflected on decisional uncertainty about medicines (decision processes). Discussions about current practice and the proposed DST indicated opportunities to facilitate SDM, if identified training needs are met. Potential individual and system-level barriers to implementation were identified, such as differences in FLS configuration and a move to remote consulting (environmental context and resources).
CONCLUSIONS
Understanding of current practice revealed unmet training needs, indicating that using a DST in isolation would be unlikely to produce a sustained shift to SDM. Findings will shape iFraP intervention development to address unmet needs.
Topics: Humans; Qualitative Research; Osteoporosis; Female; Male; Focus Groups; Bone Density Conservation Agents; Decision Making, Shared; Middle Aged; Aged; Osteoporotic Fractures
PubMed: 38898212
DOI: 10.1007/s11657-024-01410-6 -
Scientific Reports Jun 2024According to the literature, seizure prediction models should be developed following a patient-specific approach. However, seizures are usually very rare events, meaning...
According to the literature, seizure prediction models should be developed following a patient-specific approach. However, seizures are usually very rare events, meaning the number of events that may be used to optimise seizure prediction approaches is limited. To overcome such constraint, we analysed the possibility of using data from patients from an external database to improve patient-specific seizure prediction models. We present seizure prediction models trained using a transfer learning procedure. We trained a deep convolutional autoencoder using electroencephalogram data from 41 patients collected from the EPILEPSIAE database. Then, a bidirectional long short-term memory and a classifier layers were added on the top of the encoder part and were optimised for 24 patients from the Universitätsklinikum Freiburg individually. The encoder was used as a feature extraction module. Therefore, its weights were not changed during the patient-specific training. Experimental results showed that seizure prediction models optimised using pretrained weights present about four times fewer false alarms while maintaining the same ability to predict seizures and achieved more 13% validated patients. Therefore, results evidenced that the optimisation using transfer learning was more stable and faster, saving computational resources. In summary, adopting transfer learning for seizure prediction models represents a significant advancement. It addresses the data limitation seen in the seizure prediction field and offers more efficient and stable training, conserving computational resources. Additionally, despite the compact size, transfer learning allows to easily share data knowledge due to fewer ethical restrictions and lower storage requirements. The convolutional autoencoder developed in this study will be shared with the scientific community, promoting further research.
Topics: Humans; Seizures; Electroencephalography; Databases, Factual; Machine Learning; Female; Male; Neural Networks, Computer; Adult
PubMed: 38898066
DOI: 10.1038/s41598-024-64802-1 -
Scientific Reports Jun 2024One of the focal points in the field of intelligent transportation is the intelligent control of traffic signals (TS), aimed at enhancing the efficiency of urban road...
One of the focal points in the field of intelligent transportation is the intelligent control of traffic signals (TS), aimed at enhancing the efficiency of urban road networks through specific algorithms. Deep Reinforcement Learning (DRL) algorithms have become mainstream, yet they suffer from inefficient training sample selection, leading to slow convergence. Additionally, enhancing model robustness is crucial for adapting to diverse traffic conditions. Hence, this paper proposes an enhanced method for traffic signal control (TSC) based on DRL. This approach utilizes dueling network and double q-learning to alleviate the overestimation issue of DRL. Additionally, it introduces a priority sampling mechanism to enhance the utilization efficiency of samples in memory. Moreover, noise parameters are integrated into the neural network model during training to bolster its robustness. By representing high-dimensional real-time traffic information as matrices, and employing a phase-cycled action space to guide the decision-making of intelligent agents. Additionally, utilizing a reward function that closely mirrors real-world scenarios to guide model training. Experimental results demonstrate faster convergence and optimal performance in metrics such as queue length and waiting time. Testing experiments further validate the method's robustness across different traffic flow scenarios.
PubMed: 38898047
DOI: 10.1038/s41598-024-64885-w -
Scientific Reports Jun 2024Maintaining driving independence is important for older adults. However, cognitive decline, a common issue in older populations, can impair older adults' driving...
Maintaining driving independence is important for older adults. However, cognitive decline, a common issue in older populations, can impair older adults' driving abilities and overall safety on the roads. This study explores how cognitive impairment influences driving patterns and driving choices among older adults. We analyzed real-world driving patterns of 246 older adults using GPS dataloggers. Our sample included 230 cognitively normal older adults (CN; Clinical Dementia Rating [CDR] = 0) and 16 older adults with incident cognitive impairment (ICI; CDR = 0.5). The CN group had an average age of 68.2 years, with 46% females and an average of 16.5 years of education, while the ICI group's average age was 69.2 years, with 36% females and an average of 16.0 years of education. We employed spatial clustering and hashing algorithms to evaluate driving behaviours. Significant differences emerged: The ICI group used fewer distinct routes to their most common destination. These differences can be leveraged to develop driving as a digital biomarker for the early detection and continuous monitoring of cognitive impairment.
Topics: Humans; Automobile Driving; Female; Aged; Male; Cognitive Dysfunction; Aged, 80 and over; Middle Aged; Choice Behavior
PubMed: 38898026
DOI: 10.1038/s41598-024-63663-y