Smartphone Data Research Articles

Perceived Mobility Challenges in Rare Bone Diseases: How Real-world Gait Data Aligns

Rare bone diseases (RBD) are highly diverse and complex, and there is a lack of understanding of the correlation between perceived mobility challenges during everyday life and the physical abilities of people with RBD. This work aimed to elucidate perceived mobility challenges among people with RBDs and examine relationships with quantitative gait metrics captured in free-living environments to draw a link between physical abilities and everyday life challenges, which can be tracked with a smartphone. This study used a comprehensive approach involving a questionnaire including 70 daily mobility challenges across different areas of daily living and semi-supervised walks in an open-world setting using a smartphone on the thigh carried in the left jeans pocket. Gait analysis is conducted using standard (cadence, variance, covariance) and nonlinear variability calculations (Lyaopnov exponent and entropy) to assess stride characteristics and walking patterns, aligning with established methodologies for gait assessment in similar contexts. In summary, nine people with RBD and nine healthy controls completed the questionnaire and provided the smartphone data. People with RBD rated balance and walking-related mobility challenges significantly higher than healthy controls. Statistical analysis indicates a strong association between the sum score of relevant challenges and the dynamic gait stability (largest Lyapunov exponent), suggesting that gait irregularity correlates with perceived mobility challenges in people with RBD. Overall, the study draws a link between physical abilities and perceived mobility challenges using a smartphone and thus presents a way to track the progress of a disease remotely using a smartphone.

A novel maximum likelihood based probabilistic behavioral data fusion algorithm for modeling residential energy consumption.

The current research effort is focused on improving the effective use of the multiple disparate sources of data available by proposing a novel maximum likelihood based probabilistic data fusion approach for modeling residential energy consumption. To demonstrate our data fusion algorithm, we consider energy usage by fuel type variables (for electricity and natural gas) in residential dwellings as our dependent variable of interest, drawn from residential energy consumption survey (RECS) data. The national household travel survey (NHTS) dataset was considered to incorporate additional variables that are not available in the RECS data. With a focus on improving the model for the residential energy use by fuel type, our proposed research provides a probabilistic mechanism for appropriately fusing records from the NHTS data with the RECS data. Specifically, instead of strictly matching records with only common attributes, we propose a flexible differential weighting method (probabilistic) based on attribute similarity (or dissimilarity) across the common attributes for the two datasets. The fused dataset is employed to develop an updated model of residential energy use with additional independent variables contributed from the NHTS dataset. The newly estimated energy use model is compared with models estimated RECS data exclusively to see if there is any improvement offered by the newly fused variables. In our analysis, the model fit measures provide strong evidence for model improvement via fusion as well as weighted contribution estimation, thus highlighting the applicability of our proposed fusion algorithm. The analysis is further augmented through a validation exercise that provides evidence that the proposed algorithm offers enhanced explanatory power and predictive capability for the modeling energy use. Our proposed data fusion approach can be widely applied in various sectors including the use of location-based smartphone data to analyze mobility and ridehailing patterns that are likely to influence energy consumption with increasing electric vehicle (EV) adoption.

A Systematic Review of Features Forecasting Patient Arrival Numbers.

Adequate nurse staffing is crucial for quality healthcare, necessitating accurate predictions of patient arrival rates. These forecasts can be determined using supervised machine learning methods. Optimization of machine learning methods is largely about minimizing the prediction error. Existing models primarily utilize data such as historical patient visits, seasonal trends, holidays, and calendars. However, it is unclear what other features reduce the prediction error. Our systematic literature review identifies studies that use supervised machine learning to predict patient arrival numbers using nontemporal features, which are features not based on time or dates. We scrutinized 26 284 studies, eventually focusing on 27 relevant ones. These studies highlight three main feature groups: weather data, internet search and usage data, and data on (social) interaction of groups. Internet data and social interaction data appear particularly promising, with some studies reporting reduced errors by up to 33%. Although weather data are frequently used, its utility is less clear. Other potential data sources, including smartphone and social media data, remain largely unexplored. One reason for this might be potential data privacy challenges. In summary, although patient arrival prediction has become more important in recent years, there are still many questions and opportunities for future research on the features used in this area.

Generalized and Improved Human Activity Recognition for Real-Time Wellness Monitoring

Human activity categorization using smartphone data can be useful for physicians in real-time data monitoring in sports or lifestyle monitoring. The goal of this research is to develop a methodology that can identify strong machine-learning classifiers applied to various human activity datasets. The first step is pre-processing the data, followed by feature extraction, selection, and classification. Relying on a single dataset does not yield high confidence in the findings. Instead, examining multiple datasets is crucial for a comprehensive understanding, as it avoids the pitfalls of basing conclusions on one dataset alone. Multiple datasets and classifiers are applied in different experiments to achieve improved and generalized human activity recognition performance. Experimental results of the support vector machine (SVM) with its generalized performance of 99% encourage us to use the trained SVM-based model to monitor normal human activities inside the home, in the park, in the gym, etc. enhancing wellness monitoring.

A Sequence-Based Hybrid Ensemble Approach for Estimating Trail Pavement Roughness Using Smartphone and Bicycle Data

Trail pavement roughness significantly impacts user experience and safety. Measuring roughness over large areas using traditional equipment is challenging and expensive. The utilization of smartphones and bicycles offers a more feasible approach to measuring trail roughness, but the current methods to capture data using these have accuracy limitations. While machine learning has the potential to improve accuracy, there have been few applications of real-time roughness evaluation. This study proposes a hybrid ensemble machine learning model that combines sequence-based modeling with support vector regression (SVR) to estimate trail roughness using smartphone sensor data mounted on bicycles. The hybrid model outperformed traditional methods like double integration and whole-body vibration in roughness estimation. For the 0.031 mi (50 m) segments, it reduced RMSE by 54–74% for asphalt concrete (AC) trails and 50–59% for Portland cement concrete (PCC) trails. For the 0.31 mi (499 m) segments, RMSE reductions of 37–60% and 49–56% for AC and PCC trails were achieved, respectively. Additionally, the hybrid model outperformed the base random forest model by 17%, highlighting the effectiveness of combining ensemble learning with sequence modeling and SVR. These results demonstrate that the hybrid model provides a cost-effective, scalable, and highly accurate alternative for large-scale trail roughness monitoring and assessment.

Improved User Identification through Calibrated Monte-Carlo Dropout

This paper presents an enhanced approach to user identification using smartphone and wearable sensor data. Our methodology involves segmenting input data and independently analyzing subsequences with CNNs. During testing, we apply calibrated Monte-Carlo Dropout to measure prediction uncertainty. By leveraging the weights obtained from uncertainty quantification, we integrate the results through weighted averaging, thereby improving overall identification accuracy. The main motivation behind this paper is the need to calibrate the CNN for improved weighted averaging. It has been observed that incorrect predictions often receive high confidence, while correct predictions are assigned lower confidence. To tackle this issue, we have implemented the Ensemble of Near Isotonic Regression (ENIR) as an advanced calibration technique. This ensures that certainty scores more accurately reflect the true likelihood of correctness. Furthermore, our experiment shows that calibrating CNN reduces the need for Monte Carlo samples in uncertainty quantification, thereby reducing computational costs. Our thorough evaluation and comparison of different calibration methods have shown improved accuracy in user identification across multiple datasets. Our results showed notable performance improvements when compared to the latest models available. In particular, our approach achieved better results than DB2 by 1.12% and HAR by 0.3% in accuracy.

Identifying older adults at risk for dementia based on smartphone data obtained during a wayfinding task in the real world.

Alzheimer's disease (AD), as the most common form of dementia and leading cause for disability and death in old age, represents a major burden to healthcare systems worldwide. For the development of disease-modifying interventions and treatments, the detection of cognitive changes at the earliest disease stages is crucial. Recent advancements in mobile consumer technologies provide new opportunities to collect multi-dimensional data in real-life settings to identify and monitor at-risk individuals. Based on evidence showing that deficits in spatial navigation are a common hallmark of dementia, we assessed whether a memory clinic sample of patients with subjective cognitive decline (SCD) who still scored normally on neuropsychological assessments show differences in smartphone-assisted wayfinding behavior compared with cognitively healthy older and younger adults. Guided by a mobile application, participants had to find locations along a short route on the medical campus of the Magdeburg university. We show that performance measures that were extracted from GPS and user input data distinguish between the groups. In particular, the number of orientation stops was predictive of the SCD status in older participants. Our data suggest that subtle cognitive changes in patients with SCD, whose risk to develop dementia in the future is elevated, can be inferred from smartphone data, collected during a brief wayfinding task in the real world.

Neutrosophic Topological Framework for Uncertainty Management in Smartphone Data Extraction Applications

Neutrosophic Topological Framework for Uncertainty Management in Smartphone Data Extraction Applications

An Artificial Neural Network approach to assess road roughness using smartphone-based crowdsourcing data

Monitoring road surface conditions is a crucial task for road authorities to develop effective infrastructure maintenance programs. Despite smartphones have been introduced as cost-effective and real-time solution for this purpose, several challenges must be addressed before their real-world application. This study investigates the utilization of smartphone-based crowdsourcing data and Artificial Neural Networks (ANN) to enhance the precision of road surface condition estimation. Initially, data are collected from four different smartphone models mounted in various vehicles, including vertical acceleration, geographic location, and speed. The root mean square of the vertical acceleration data, along with vehicle speed, is then employed as input features for the ANN, while the true International Roughness Index (IRI) values serve as the corresponding output features. Comparative analysis between ANN and regression models based on statistical metrics such as Mean Squared Error (MSE) and Pearson correlation revealed that ANN outperforms regression models. The obtained MSE and Pearson correlation values for ANN (0.56 and 0.91) surpass those of regression models (0.72 and 0.88). Moreover, results indicated that utilizing crowdsourcing smartphone data yielded superior outcomes compared to using a single smartphone for this purpose.

A comparison of smartphone and infrasound microphone data from a fuel air explosive and a high explosive.

For prompt detection of large (>1 kt) above-ground explosions, infrasound microphone networks and arrays are deployed at surveyed locations across the world. Denser regional and local networks are deployed for smaller explosions, however, they are limited in number and are often deployed temporarily for experiments. With the expanded interest in smaller yield explosions targeted at vulnerable areas such as population centers and key infrastructures, the need for more dense microphone networks has increased. An "attritable" (affordable, reusable, and replaceable) and flexible alternative can be provided by smartphone networks. Explosion signals from a fuel air explosive (thermobaric bomb) and a high explosive with trinitrotoluene equivalent yields of 6.35 and 3.63 kg, respectively, were captured on both an infrasound microphone and a network of smartphones. The resulting waveforms were compared in time, frequency, and time-frequency domains. The acoustic waveforms collected on smartphones produced a filtered explosion pulse due to the smartphone's diminishing frequency response at infrasound frequencies (<20 Hz) and was found difficult to be used with explosion characterization methods utilizing waveform features (peak overpressure, impulse, etc.). However, the similarities in time frequency representations and additional sensor inputs are promising for other explosion signal identification and analysis. As an example, a method utilizing the relative acoustic amplitudes for source localization using the smartphone sensor network is presented.

A Review of Vision-Based Pothole Detection Methods Using Computer Vision and Machine Learning.

Potholes and other road surface damages pose significant risks to vehicles and traffic safety. The current methods of in situ visual inspection for potholes or cracks are inefficient, costly, and hazardous. Therefore, there is a pressing need to develop automated systems for assessing road surface conditions, aiming to efficiently and accurately reconstruct, recognize, and locate potholes. In recent years, various methods utilizing (a) computer vision, (b) three-dimensional (3D) point clouds, or (c) smartphone data have been employed to map road surface quality conditions. Machine learning and deep learning techniques have increasingly enhanced the performance of these methods. This review aims to provide a comprehensive overview of cutting-edge computer vision and machine learning algorithms for pothole detection. It covers topics such as sensing systems for acquiring two-dimensional (2D) and 3D road data, classical algorithms based on 2D image processing, segmentation-based algorithms using 3D point cloud modeling, machine learning, deep learning algorithms, and hybrid approaches. The review highlights that hybrid methods combining traditional image processing and advanced machine learning techniques offer the highest accuracy in pothole detection. Machine learning approaches, particularly deep learning, demonstrate superior adaptability and detection rates, while traditional 2D and 3D methods provide valuable baseline techniques. By reviewing and evaluating existing vision-based methods, this paper clarifies the current landscape of pothole detection technologies and identifies opportunities for future research and development. Additionally, insights provided by this review can inform the design and implementation of more robust and effective systems for automated road surface condition assessment, thereby contributing to enhanced roadway safety and infrastructure management.

Estimating traffic demand of different transportation modes using floating smartphone data

Traffic demand is crucial for traffic planning, helping to understand traffic volume, identify congestion points, and develop strategies for efficient and sustainable transportation. This paper introduces a method to estimate origin-destination (OD) matrices for drivers, cyclists, and pedestrians using floating smartphone data (FSD) and the information minimisation (IM) model. The FSD is used in the gravity model (GM) to provide more accurate inputs for the IM model by estimating an a priori matrix and traffic demand at intersections. Additionally, FSD allowed a direct route choice proportions deriviation without needing a traffic assignment model. The proposed estimation model was tested with real data and analyzed in a simulation environment, examining the impact of increasing the penetration rate (PR) of FSD. Results indicate that traffic demand in the study area was accurately estimated with an average Rxy value of 92%, aligning well with simulation results.

A translationally informed approach to vital signs for psychiatry: a preliminary proof of concept

The nature of data obtainable from the commercial smartphone – bolstered by a translational model emphasizing the impact of social and physical zeitgebers on circadian rhythms and mood – offers the possibility of scalable and objective vital signs for major depression. Our objective was to explore associations between passively sensed behavioral smartphone data and repeatedly measured depressive symptoms to suggest which features could eventually lead towards vital signs for depression. We collected continuous behavioral data and bi-weekly depressive symptoms (PHQ-8) from 131 psychiatric outpatients with a lifetime DSM-5 diagnosis of depression and/or anxiety over a 16-week period. Using linear mixed-effects models, we related depressive symptoms to concurrent passively sensed behavioral summary features (mean and variability of sleep, activity, and social engagement metrics), considering both between- and within-person associations. Individuals with more variable wake-up times across the study reported higher depressive symptoms relative to individuals with less variable wake-up times (B [95% CI] = 1.53 [0.13, 2.93]). On a given week, having a lower step count (−0.16 [−0.32, −0.01]), slower walking rate (−1.46 [−2.60, −0.32]), lower normalized location entropy (−3.01 [−5.51, −0.52]), more time at home (0.05 [0.00, 0.10]), and lower distances traveled (−0.97 [−1.72, −0.22]), relative to one’s own typical levels, were each associated with higher depressive symptoms. With replication in larger samples and a clear understanding of how these components are best combined, a behavioral composite measure of depression could potentially offer the kinds of vital signs for psychiatric medicine that have proven invaluable to assessment and decision-making in physical medicine. Clinical Trials Registration: The data that form the basis of this report were collected as part of clinical trial number NCT03152864.

Real-world smartphone data can trace the behavioural impact of epilepsy: A case study.

Neurobehavioural comorbidities have a detrimental effect on the quality of life of people with epilepsy, yet tracking their impact is challenging as behaviour may vary with seizures and anti-seizure medication (ASM) side effects. Smartphones have the potential to monitor day-to-day neurobehavioural patterns objectively. We present the case of a man in his late twenties with drug-resistant focal epilepsy in whom we ascertained the effects of ASM withdrawal and a convulsive seizure on his touchscreen interactions. Using a dedicated app, we recorded over 185 days the timestamps of 718,357 interactions. We divided the various smartphone behaviours according to the next-interval dynamics of the interactions by using a joint interval distribution (JID). During two ASM load transitions, namely before versus during tapering and tapering versus restarting medication, we used cluster-based permutation tests to compare the JIDs. We also compared the JID of the seizure day to the average of the previous 3 days. The cluster-based permutation tests revealed significant differences, with accelerated next-interval dynamics during tapering and a reversal upon medication restart. The day of the convulsion exhibited a marked slowing of next-interval dynamics compared to the preceding 3 days. Our findings suggest that the temporal dynamics of smartphone touchscreen interactions may help monitor neurobehavioural comorbidities in neurological care.

Identifying Person-Specific Drivers of Depression in Adolescents: Protocol for a Smartphone-Based Ecological Momentary Assessment and Passive Sensing Study.

Adolescence is marked by an increasing risk of depression and is an optimal window for prevention and early intervention. Personalizing interventions may be one way to maximize therapeutic benefit, especially given the marked heterogeneity in depressive presentations. However, empirical evidence that can guide personalized intervention for youth is lacking. Identifying person-specific symptom drivers during adolescence could improve outcomes by accounting for both developmental and individual differences. This study leverages adolescents' everyday smartphone use to investigate person-specific drivers of depression and validate smartphone-based mobile sensing data against established ambulatory methods. We describe the methods of this study and provide an update on its status. After data collection is completed, we will address three specific aims: (1) identify idiographic drivers of dynamic variability in depressive symptoms, (2) test the validity of mobile sensing against ecological momentary assessment (EMA) and actigraphy for identifying these drivers, and (3) explore adolescent baseline characteristics as predictors of these drivers. A total of 50 adolescents with elevated symptoms of depression will participate in 28 days of (1) smartphone-based EMA assessing depressive symptoms, processes, affect, and sleep; (2) mobile sensing of mobility, physical activity, sleep, natural language use in typed interpersonal communication, screen-on time, and call frequency and duration using the Effortless Assessment of Risk States smartphone app; and (3) wrist actigraphy of physical activity and sleep. Adolescents and caregivers will complete developmental and clinical measures at baseline, as well as user feedback interviews at follow-up. Idiographic, within-subject networks of EMA symptoms will be modeled to identify each adolescent's person-specific drivers of depression. Correlations among EMA, mobile sensor, and actigraph measures of sleep, physical, and social activity will be used to assess the validity of mobile sensing for identifying person-specific drivers. Data-driven analyses of mobile sensor variables predicting core depressive symptoms (self-reported mood and anhedonia) will also be used to assess the validity of mobile sensing for identifying drivers. Finally, between-subject baseline characteristics will be explored as predictors of person-specific drivers. As of October 2023, 84 families were screened as eligible, of whom 70% (n=59) provided informed consent and 46% (n=39) met all inclusion criteria after completing baseline assessment. Of the 39 included families, 85% (n=33) completed the 28-day smartphone and actigraph data collection period and follow-up study visit. This study leverages depressed adolescents' everyday smartphone use to identify person-specific drivers of adolescent depression and to assess the validity of mobile sensing for identifying these drivers. The findings are expected to offer novel insights into the structure and dynamics of depressive symptomatology during a sensitive period of development and to inform future development of a scalable, low-burden smartphone-based tool that can guide personalized treatment decisions for depressed adolescents. DERR1-10.2196/43931.

Smartphone Voice Calls Provide Early Biomarkers of Parkinsonism in Rapid Eye Movement Sleep Behavior Disorder.

Speech dysfunction represents one of the initial motor manifestations to develop in Parkinson's disease (PD) and is measurable through smartphone. The aim was to develop a fully automated and noise-resistant smartphone-based system that can unobtrusively screen for prodromal parkinsonian speech disorder in subjects with isolated rapid eye movement sleep behavior disorder (iRBD) in a real-world scenario. This cross-sectional study assessed regular, everyday voice call data from individuals with iRBD compared to early PD patients and healthy controls via a developed smartphone application. The participants also performed an active, regular reading of a short passage on their smartphone. Smartphone data were continuously collected for up to 3 months after the standard in-person assessments at the clinic. A total of 3525 calls that led to 5990 minutes of preprocessed speech were extracted from 72 participants, comprising 21 iRBD patients, 26 PD patients, and 25 controls. With a high area under the curve of 0.85 between iRBD patients and controls, the combination of passive and active smartphone data provided a comparable or even more sensitive evaluation than laboratory examination using a high-quality microphone. The most sensitive features to induce prodromal neurodegeneration in iRBD included imprecise vowel articulation during phone calls (P = 0.03) and monopitch in reading (P = 0.05). Eighteen minutes of speech corresponding to approximately nine calls was sufficient to obtain the best sensitivity for the screening. We consider the developed tool widely applicable to deep longitudinal digital phenotyping data with future applications in neuroprotective trials, deep brain stimulation optimization, neuropsychiatry, speech therapy, population screening, and beyond. © 2024 The Author(s). Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.

Untapping the potential of mobile location data: The opportunities and challenges for retail analytics

Smartphone technology has created a burgeoning mobile location data (MLD) marketplace. MLD has been adopted by many retailers to track consumer behaviours at a granular level across space and time. This paper provides a critical empirical case study approach to generate spatiotemporal MLD-based insights using raw smartphone data at five super-regional shopping centres. The analysis and visualization of MLD-based metrics provide the basis to rethink how we approach retail analytics and location decision-making. Many challenges need to be addressed to leverage the full potential of MLD, including data bias, spatial and temporal accuracy and significant consumer privacy concerns.

Bridging the Gap: commodifying infrastructure spatial dynamics with crowdsourced smartphone data

Structural information deficits about aging bridges have led to several avoidable catastrophes in recent years. Data-driven methods for bridge vibration monitoring enable frequent, accurate structural assessments; however, the high costs of widespread deployments of these systems make important condition information a luxury for bridge owners. Smartphone-based monitoring is inexpensive and has produced structural information, i.e., modal frequencies, in crowdsensing applications. Even so, current methods cannot extract spatial vibration characteristics with uncontrolled datasets that are needed for damage identification. Here we present an extensive real-world study with crowdsourced smartphone-vehicle trips within motor vehicles in which we estimate absolute value mode shapes and simulate damage detection capabilities. Our method analyzes over 800 trips across four road bridges with main spans ranging from 30 to 1300 m in length, representing about one-quarter of bridges in the United States. We demonstrate a bridge health monitoring platform compatible with ride-sourcing data streams that check conditions daily. The result has the potential to commodify data-driven structural assessments globally.

1103 Gait Analysis Using Wearables and AI in Predicting Peripheral Arterial Disease: A Systematic Review

Abstract Introduction Peripheral arterial disease (PAD) affects over 230 million people globally. The trajectory of this disease process can culminate in major amputation. These patients demonstrate statistically significant changes in gait including changes in stride length, plantarflexion and other spatiotemporal parameters. Aim To determine if alterations in gait parameters can be detected and measured accurately by wearables and AI in PAD patients and if these measurements are predictive of PAD progression. Method A systematic review was performed in line with PRISMA guidelines. MEDLINE, Embase, Web of Science and Scopus were used for database searches. A combination of medical subject headings (MeSH) regarding ‘PAD’, ‘gait’, ‘kinetics’, ‘biomechanics’, ‘ambulation’, ‘artificial intelligence’, ‘sensors’ and ‘smartphone data’ were employed in the primary search string. Screening, full text reviews and data extraction was conducted by 2 reviewers using Covidence. Bias was assessed using the Newcastle-Ottawa scale. Results 4719 studies were identified. 3873 were screened after de-duplication. 5 studies were included in the final review. The main methods of gait measurements included smartphones (iPhone and Samsung Galaxy), wearable sensors and motion sensor technology. Gait initiation parameters were sensitive in detecting gait impairment. iPhone pedometer demonstrated accuracy measuring steps comparable to reference Actigraph but was inaccurate measuring distance. Machine learning models including Random Forest algorithms and extreme gradient boosting, detected PAD patients with 89% and 92% accuracy respectively. Conclusions Wearables and AI algorithms can detect gait changes and differentiate PAD from non-PAD gait. All studies showed detective capability, however none demonstrated predictability of PAD. Limitations include accuracy and validation of measurement methods.

Mortality and its associated factors among mechanically ventilated adult patients in the intensive care units of referral hospitals in Northwest Amhara, Ethiopia, 2023.

Worldwide, nearly half of the patients admitted to intensive care units require ventilatory support. Despite advances in intensive care unit patient management and mechanical ventilator utilization, the odds of mortality among mechanically ventilated patients are higher in resource-limited settings. Little is known about the mortality of patients on mechanical ventilation outside the capital of Ethiopia. This study aimed to assess mortality and its associated factors among mechanically ventilated adult patients in intensive care units. An institutional-based cross-sectional study was conducted on mechanically ventilated patients in intensive care units from 1 February 2020 to 1 March 2023. A simple random sampling technique was used to select 434 patients' charts. A data extraction tool designed on the Kobo toolbox, a smartphone data collection platform, was used to collect the data. The data were exported into Microsoft Excel 2019 and then into Stata 17 for data management and analysis. Descriptive statistics were used to summarize the characteristics of the study participants. A bivariable logistic regression was conducted, and variables with p ≤ 0.20 were recruited for multivariable analysis. Statistical significance was declared at p < 0.05, and the strength of associations was summarized using an adjusted odds ratio with 95% confidence intervals. A total of 404 charts of mechanically ventilated patients were included, with a completeness rate of 93.1%. The overall proportion of mortality was 62.87%, with a 95% CI of (58.16-67.58). In the multivariable logistic regression, age 41-70 years (AOR: 4.28, 95% CI: 1.89-9.62), sepsis (AOR: 2.43, 95% CI: 1.08-5.46), reintubation (AOR: 2.76, 95% CI: 1.06-7.21), and sedation use (AOR: 0.41, 95% CI: 0.18-0.98) were found to be significant factors associated with the mortality of mechanically ventilated patients in the intensive care unit. The magnitude of mortality among mechanically ventilated patients was high. Factors associated with increased odds of death were advanced age, sepsis, and reintubation. However, sedation use was a factor associated with decreased mortality. Healthcare professionals in intensive care units should pay special attention to patients with sepsis, those requiring reintubation, those undergoing sedation, and those who are of advanced age.