XGBoost based machine learning approach to predict the risk of fall in older adults using gait outcomes

Noh, Byungjoo; Youm, Changhong; Goh, Eun-Kyoung; Lee, Myeounggon; Park, Hwayoung; Jeon, Hyo Jeong; Kim, Oh Yoen

doi:10.1038/s41598-021-91797-w

Cited by 41 publications

(30 citation statements)

References 37 publications

(50 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They managed to discriminate between fallers and nonfallers using shallow learning (SVM [82,83], Partial least square discriminant analysis (PLS-DA) [81]) and improve our understanding of how falls-related gait impairments in neurological patients are manifested. Similarly, [84] used Spatio-temporal variables from shoes' IMU sensors during 20 m walking and shallow learning (extreme gradient boosting (XGBoost), a decision tree-based ensemble machine learning technique). They found that stride length and walking speed are the most important variables regarding future risk of fall.…”

Section: Gait Analysis Sensors Analytics and Future Challengesmentioning

confidence: 99%

Preventing falls: the use of machine learning for the prediction of future falls in individuals without history of fall

et al. 2022

View full text Add to dashboard Cite

Nowadays, it becomes of paramount societal importance to support many frail-prone groups in our society (elderly, patients with neurodegenerative diseases, etc.) to remain socially and physically active, maintain their quality of life, and avoid their loss of autonomy. Once older people enter the prefrail stage, they are already likely to experience falls whose consequences may accelerate the deterioration of their quality of life (injuries, fear of falling, reduction of physical activity). In that context, detecting frailty and high risk of fall at an early stage is the first line of defense against the detrimental consequences of fall. The second line of defense would be to develop original protocols to detect future fallers before any fall occur. This paper briefly summarizes the current advancements and perspectives that may arise from the combination of affordable and easy-to-use non-wearable systems (force platforms, 3D tracking motion systems), wearable systems (accelerometers, gyroscopes, inertial measurement units-IMUs) with appropriate machine learning analytics, as well as the efforts to address these challenges.

show abstract

Section: Gait Analysis Sensors Analytics and Future Challengesmentioning

confidence: 99%

Preventing falls: the use of machine learning for the prediction of future falls in individuals without history of fall

et al. 2022

View full text Add to dashboard Cite

show abstract

“…PAFA (Zhou and Zhao, 2018) utilizes sparse logistic regression, which is trained by noncoding pathogenic variants in ClinVar and GWAS SNPs in GWASdb (Li et al, 2012), to predict noncoding risk variants; ncER (Wells et al, 2019) adopts XGBoost (Noh et al, 2021), which is trained by noncoding regulatory variants in HGMD and noncoding pathogenetic variants in ClinVar, to predict noncoding pathogenic variants. Moreover, unsupervised methods have also been developed for their advantage of not relying on labelled functional NCVs especially when functional NCVs are few.…”

Section: Introductionmentioning

confidence: 99%

Exploiting deep transfer learning for the prediction of functional non-coding variants using genomic sequence

Chen

Wang

2022

Bioinformatics

View full text Add to dashboard Cite

Motivation Though genome-wide association studies have identified tens of thousands of variants associated with complex traits and most of them fall within the noncoding regions, they may not the causal ones. The development of high-throughput functional assays leads to the discovery of experimental validated noncoding functional variants. However, these validated variants are rare due to technical difficulty and financial cost. The small sample size of validated variants makes it less reliable to develop a supervised machine learning model for achieving a whole genome-wide prediction of noncoding causal variants. Results We will exploit a deep transfer learning model, which is based on convolutional neural network, to improve the prediction for functional noncoding variants. To address the challenge of small sample size, the transfer learning model leverages both large-scale generic functional noncoding variants to improve the learning of low-level features and context-specific functional noncoding variants to learn high-level features toward the context-specific prediction task. By evaluating the deep transfer learning model on three MPRA datasets and 16 GWAS datasets, we demonstrate that the proposed model outperforms deep learning models without pretraining or retraining. In addition, the deep transfer learning model outperforms 18 existing computational methods in both MPRA and GWAS datasets. Availability https://github.com/lichen-lab/TLVar Supplementary Information Supplementary data are available at Bioinformatics online.

show abstract

“…Sarcopenia and frailty, characterized by decreased MM, MS, and MF, are associated with decreased gait ability [ 4 , 5 , 6 ]. Gait analysis effectively identifies early pathology, assesses disease progression, and predicts fall risk [ 7 , 8 ]. Therefore, gait analysis may help predict and identify decreased MM, MS, and MF [ 4 , 5 , 6 ].…”

Section: Introductionmentioning

confidence: 99%

Association of Muscle Mass, Muscle Strength, and Muscle Function with Gait Ability Assessed Using Inertial Measurement Unit Sensors in Older Women

Kim

Youm

Park

et al. 2022

IJERPH

Self Cite

View full text Add to dashboard Cite

Aging-related muscle atrophy is associated with decreased muscle mass (MM), muscle strength (MS), and muscle function (MF) and may cause motor control, balance, and gait pattern impairments. This study determined associations of three speed-based gait variables with loss of MM, MS, and MF in older women. Overall, 432 older women aged ≥65 performed appendicular skeletal muscle, handgrip strength, and five times sit-to-stand test to evaluate MM, MS, and MF. A gait test was performed at three speeds by modifying the preferred walking speed (PWS; slower walking speed (SWS); faster-walking speed (FWS)) on a straight 19 m walkway. Stride length (SL) at PWS was significantly associated with MM. FWS and coefficient of variance (CV) of double support phase (DSP) and DSP at PWS showed significant associations with MS. CV of step time and stride time at SWS, FWS, and single support phase (SSP) at PWS showed significant associations with MF. SL at PWS, DSP at FWS, CV of DSP at PWS, stride time at SWS, and CV of SSP at PWS showed significant associations with composite MM, MS, and MF variables. Our study indicated that gait tasks under continuous and various speed conditions are useful for evaluating MM, MS, and MF.

show abstract

XGBoost based machine learning approach to predict the risk of fall in older adults using gait outcomes

Cited by 41 publications

References 37 publications

Preventing falls: the use of machine learning for the prediction of future falls in individuals without history of fall

Preventing falls: the use of machine learning for the prediction of future falls in individuals without history of fall

Exploiting deep transfer learning for the prediction of functional non-coding variants using genomic sequence

Association of Muscle Mass, Muscle Strength, and Muscle Function with Gait Ability Assessed Using Inertial Measurement Unit Sensors in Older Women

Contact Info

Product

Resources

About