AI-Powered Fall Risk Prediction in Nursing Care

NCT07000981 | Completed Results

• 1 other identifier: INONU-FN-AC-01
• Study Type: observational
• Target: 177
• Locations: 1 country
• Sites: 1

Brief Summary

The goal of this study is to develop a nursing clinical decision support system for fall risk prediction using machine learning and computer vision techniques. The system is intended to offer advantages over traditional scales, including real-time analysis, contactless monitoring, objective evaluation, and personalized risk prediction-ultimately aiming to improve patient safety and reduce complications related to falls in clinical settings. This study aims to answer the following questions: Can machine learning models serve as valid tools for fall risk prediction? Is the proposed system feasible for use in clinical environments? Inclusion criteria for participants:

• Aged 18 years or older
• Able to read and write in Turkish
• Able to walk with or without assistance
• Willing to voluntarily participate in the study Exclusion criteria:
• Inability to speak or understand Turkish adequately
• Being intubated
• Being physically restrained
• Being immobile
• Having a diagnosed cognitive impairment Participants' basic information-including age, height, and weight-will be collected through a demographic data form. Fall risk will be initially assessed using the Morse Fall Scale. Then, a walking assessment will be conducted using a digital camera-based computer vision system as participants walk at a comfortable pace in a clinical corridor. Additionally, an accelerometer placed in the participants' pockets will record three-axis acceleration (X, Y, Z) during walking. The data obtained will be analyzed using machine learning algorithms to estimate lower and upper limb biomechanics in real time. Features such as step length, cadence, gait cycle, and range of motion (ROM) will be extracted. These features, combined with Morse Fall Scale scores, will be used to train and validate an artificial neural network (ANN). The study aims to contribute to the development of a reliable, objective, and real-time system capable of predicting fall risk in clinical environments through gait analysis.

Conditions

Accidental Falls; Risk Assessment; Nursing Assessment; Machine Learning; Computer Vision

Outcome Measures

Primary Outcomes (3)

• Fall Risk Categories Based on the Morse Fall Scale

  Assessment of the patient's fall risk with the Morse Fall Scale It is an effective and simple measurement tool that is frequently used in hospitals in Turkey and used to diagnose potential patient fall risks for the nursing profession. The scale consists of six criteria (secondary diagnosis, presence of a history of falls, mobilization support, presence of intravenous access or heparin use, gait/transfer, and mental status) that diagnose fall risk. According to this assessment tool, if the patient has a score below 25 points, he/she is in the low risk group for falls. If the score is between 25 and 50, the patient is in the medium risk group, and if the score is 51 and above, the patient is in the high risk group. A minimum score of 0 and a maximum score of 125 can be obtained from the scale. This scale allows a systematic determination of the fall risk of patients in clinical settings.

  Day 1

• Fall Risk Classification Accuracy of the Decision Support System

  Classification accuracy of the decision support system was evaluated based on the percentage of test units correctly classified. The scale ranges from 0% to 100%, where higher values indicate better performance. This metric reflects the proportion of correctly identified cases by the system during model evaluation.

  Day 1

• Classification Performance Metrics of the Decision Support System (F1 Score, Precision, Recall)

  This outcome measure evaluates the classification performance of a clinical decision support system using standard machine learning metrics: precision, recall, and F1-score. These metrics are based on a scale ranging from 0 to 1. Higher values indicate better classification performance. Precision is defined as the proportion of true positive predictions among all positive predictions. Recall is defined as the proportion of true positive predictions among all actual positives. The F1-score is the harmonic mean of precision and recall.

  Day 1

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Adult (18-64), Older Adult (65+)
• Sampling Method: Non-Probability Sample

Study Population

The study is planned to be conducted on patients who meet the inclusion criteria and are hospitalized in the Physical Medicine and Rehabilitation department of a research hospital located in the eastern region of Turkey.

You may qualify if:

• years of age or older
• Accepting voluntary participation in the study
• Be able to read and write Turkish
• Being able to walk with or without support

You may not qualify if:

• Not being able to speak or understand Turkish adequately
• Being intubated.
• To be identified.
• Being immobile
• Having a mental disability.

Sponsors & Collaborators

• Inonu University: lead

Study Sites (1)

Turgut Özal Medical Center

Malatya, Battalgazi, 44280, Turkey (Türkiye)

Location

Related Publications (7)

• Akhtaruzzaman M, Shafie AA, Khan MR. Gait Analysis: Systems, Technologies and Significance. J Mech Med Biol. 2016;16(7).

  BACKGROUND

• Demir NY, Intepeler ŞS. Adaptation Of Morse Fall Scale To Turkish And Determination Of Sensitivity And Specificity. J Ege Univ Nurs Fac. 2012;28(1):57-71.

  BACKGROUND

• Zhao M, Chang CH, Xie W, Xie Z, Hu J. Cloud Shape Classification System Based on Multichannel CNN and Enhanced FDM. IEEE Access. 2020;8:44111-24.

  BACKGROUND

• Krizhevsky A, Sutskever I, Hinton GE. ImageNet classification with deep convolutional neural networks. Commun ACM. 2017;60(6):84-90.

  BACKGROUND

• Santos GL, Endo PT, Monteiro KHC, Rocha EDS, Silva I, Lynn T. Accelerometer-Based Human Fall Detection Using Convolutional Neural Networks. Sensors (Basel). 2019 Apr 6;19(7):1644. doi: 10.3390/s19071644.

  PMID: 30959877 BACKGROUND

• Yunas SU, Ozanyan KB. Gait Activity Classification Using Multi-Modality Sensor Fusion: A Deep Learning Approach. IEEE Sens J. 2021;21(15):16870-9.

  BACKGROUND

• Manssor SAF, Sun S, Elhassan MAM. Real-Time Human Recognition at Night via Integrated Face and Gait Recognition Technologies. Sensors (Basel). 2021 Jun 24;21(13):4323. doi: 10.3390/s21134323.

  PMID: 34202659 BACKGROUND

Results Point of Contact

• Title: Asst. Prof. Gürkan ÖZDEN
• Organization: Inonu University

gurkan.ozden@inonu.edu.tr

+90 505 715 62 65

Study Officials

• Gürkan Özden, Assistant Professor

  Inonu University

  STUDY DIRECTOR

Publication Agreements

• PI is Sponsor Employee: No
• Restrictive Agreement: No

Study Design

• Study Type: observational
• Observational Model: OTHER
• Time Perspective: CROSS SECTIONAL
• Sponsor Type: OTHER
• Responsible Party: PRINCIPAL INVESTIGATOR
• PI Title: ACeviz [Ahmet CEVİZ] Principal Investigator

Study Record Dates

• First Submitted: May 13, 2025
• First Posted: June 3, 2025
• Study Start: June 1, 2025
• Primary Completion: July 1, 2025
• Study Completion: July 24, 2025
• Last Updated: December 2, 2025
• Results First Posted: December 2, 2025

Record last verified: 2025-11

Data Sharing

• IPD Sharing: Will not share

Locations

TU (1)