Development and Technical Validation of a Smartphone-Based Cry Detection Algorithm.
Ontology highlight
ABSTRACT:
SUBMITTER: ZhuParris A
PROVIDER: S-EPMC8076575 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:IntroductionCoughing is a common symptom in pediatric lung disease and cough frequency has been shown to be correlated to disease activity in several conditions. Automated cough detection could provide a noninvasive digital biomarker for pediatric clinical trials or care. The aim of this study was to develop a smartphone-based algorithm that objectively and automatically counts cough sounds of children.MethodsThe training set was composed of 3228 pediatric cough sounds and 480,780 noncough sounds from various publicly available sources and continuous sound recordings of 7 patients admitted due to respiratory disease. A Gradient Boost Classifier was fitted on the training data, which was subsequently validated on recordings from 14 additional patients aged 0-14 admitted to the pediatric ward due to respiratory disease. The robustness of the algorithm was investigated by repeatedly classifying a recording with the smartphone-based algorithm during various conditions.ResultsThe final algorithm obtained an accuracy of 99.7%, sensitivity of 47.6%, specificity of 99.96%, positive predictive value of 82.2% and negative predictive value 99.8% in the validation dataset. The correlation coefficient between manual- and automated cough counts in the validation dataset was 0.97 (p < .001). The intra- and interdevice reliability of the algorithm was adequate, and the algorithm performed best at an unobstructed distance of 0.5-1 m from the audio source.ConclusionThis novel smartphone-based pediatric cough detection application can be used for longitudinal follow-up in clinical care or as digital endpoint in clinical trials.
| S-EPMC9306830 | biostudies-literature
Project description:Background/aimsTo compare intraocular pressure (IOP) measurements using a prototype smartphone tonometer with other tonometers used in clinical practice.MethodsPatients from an academic glaucoma practice were recruited. The smartphone tonometer uses fixed force applanation and in conjunction with a machine-learning computer algorithm is able to calculate the IOP. IOP was also measured using Goldmann applanation tonometry (GAT) in all subjects. A subset of patients were also measured using ICare, pneumotonometry (upright and supine positions) and Tono-Pen (upright and supine positions) and the results were compared.Results92 eyes of 81 subjects were successfully measured. The mean difference (in mm Hg) for IOP measurements of the smartphone tonometer versus other devices was +0.24 mm Hg for GAT, -1.39 mm Hg for ICare, -3.71 mm Hg for pneumotonometry and -1.30 mm Hg for Tono-Pen. The 95% limits of agreement for the smartphone tonometer versus other devices was -4.35 to 4.83 mm Hg for GAT, -6.48 to 3.70 mm Hg for ICare, -7.66 to -0.15 mm Hg for pneumotonometry and -5.72 to 3.12 mm Hg for Tono-Pen. Overall, the smartphone tonometer results correlated best with GAT (R2=0.67, p<0.001). Of the 92 videos, 90 (97.8%) were within ±5 mm Hg of GAT and 58 (63.0%) were within ±2 mm Hg of GAT.ConclusionsPreliminary IOP measurements using a prototype smartphone-based tonometer was grossly equivalent to the reference standard.
| S-EPMC7397723 | biostudies-literature
Project description:PurposeContrast sensitivity (CS) testing is an important measure of visual function reflecting variations in everyday visual experience in different conditions and helps to identify more subtle vision loss. However, it is only infrequently used. To make this more accessible, we have developed and validated a smartphone-based CS test.MethodsA new tumbling-E smartphone-based CS test was developed, Peek Contrast Sensitivity (PeekCS). This was field tested and refined through several iterations. Reference standard was a tumbling-E Pelli-Robson CS test (PRCS). The validation study was conducted in community clinics in Ethiopia. Test-retest variability was measured for both PRCS and PeekCS. PRCS and PeekCS were then compared. Correlation coefficients and 95% confidence intervals (CIs) were calculated; 95% limits of agreement were calculated and displayed on Bland-Altman plots.ResultsPeekCS showed strong repeatability (correlation coefficient: 0.93; 95% CI: 0.91-0.95), which was comparable with PRCS (correlation coefficient: 0.96; 95% CI: 0.95-0.97). The 95% limit of agreement for test-retest variability of PRCS and PeekCS were -0.20 to 0.21 and -0.31 to 0.29, respectively. PRCS and PeekCS were highly correlated: 0.94 (95% CI: 0.93-0.95); 95% limits of agreement -0.27 to 0.29; and mean difference 0.010 (95% CI: -0.001 to 0.022). PeekCS had a faster testing time (44.6 seconds) than PRCS (48.6 seconds): mean difference -3.98 (95% CI: -5.38 to -2.58); P < 0.001.ConclusionsThe smartphone-based PeekCS is a repeatable and rapid test, providing results that are highly comparable with the commonly used PRCS test.Translational relevancePeekCS provides an accessible and easy to perform alternative for CS testing, particularly in the community setting.
| S-EPMC6743644 | biostudies-literature
Project description:This paper proposes a smartphone-based colorimetric pH detection method using a color adaptation algorithm for point-of-care applications. Although a smartphone camera can be utilized to measure the color information of colorimetric sensors, ambient light changes and unknown built-in automatic image correction operations make it difficult to obtain stable color information. This paper utilizes a 3D printed mini light box and performs a calibration procedure with a paper-printed comparison chart and a reference image which overcomes the drawbacks of smartphone cameras and the difficulty in preparing for the calibration procedure. The color adaptation is performed in the CIE 1976 u'v' color space by using the reference paper in order to stabilize the color variations. Non-rigid u'v' curve interpolation is used to produce the high-resolution pH estimate. The final pH value is estimated by using the best-matching method to handle the nonlinear curve properties of multiple color patches. The experimental results obtained using a pH indicator paper show that the proposed algorithm provides reasonably good estimation of pH detection. With paper-printed accurate color comparison charts and smart color adaptation techniques, superior estimation is achieved in the smartphone-based colorimetric pH detection system for point-of-care application.
| S-EPMC5539506 | biostudies-other
Project description:There exist significant clinical and basic research needs for accurate, automated seizure detection algorithms. These algorithms have translational potential in responsive neurostimulation devices and in automatic parsing of continuous intracranial electroencephalography data. An important barrier to developing accurate, validated algorithms for seizure detection is limited access to high-quality, expertly annotated seizure data from prolonged recordings. To overcome this, we hosted a kaggle.com competition to crowdsource the development of seizure detection algorithms using intracranial electroencephalography from canines and humans with epilepsy. The top three performing algorithms from the contest were then validated on out-of-sample patient data including standard clinical data and continuous ambulatory human data obtained over several years using the implantable NeuroVista seizure advisory system. Two hundred teams of data scientists from all over the world participated in the kaggle.com competition. The top performing teams submitted highly accurate algorithms with consistent performance in the out-of-sample validation study. The performance of these seizure detection algorithms, achieved using freely available code and data, sets a new reproducible benchmark for personalized seizure detection. We have also shared a 'plug and play' pipeline to allow other researchers to easily use these algorithms on their own datasets. The success of this competition demonstrates how sharing code and high quality data results in the creation of powerful translational tools with significant potential to impact patient care.
| S-EPMC6075622 | biostudies-literature
Project description:BackgroundDetection of active pulmonary tuberculosis on chest radiographs (CRs) is critical for the diagnosis and screening of tuberculosis. An automated system may help streamline the tuberculosis screening process and improve diagnostic performance.MethodsWe developed a deep learning-based automatic detection (DLAD) algorithm using 54c221 normal CRs and 6768 CRs with active pulmonary tuberculosis that were labeled and annotated by 13 board-certified radiologists. The performance of DLAD was validated using 6 external multicenter, multinational datasets. To compare the performances of DLAD with physicians, an observer performance test was conducted by 15 physicians including nonradiology physicians, board-certified radiologists, and thoracic radiologists. Image-wise classification and lesion-wise localization performances were measured using area under the receiver operating characteristic (ROC) curves and area under the alternative free-response ROC curves, respectively. Sensitivities and specificities of DLAD were calculated using 2 cutoffs (high sensitivity [98%] and high specificity [98%]) obtained through in-house validation.ResultsDLAD demonstrated classification performance of 0.977-1.000 and localization performance of 0.973-1.000. Sensitivities and specificities for classification were 94.3%-100% and 91.1%-100% using the high-sensitivity cutoff and 84.1%-99.0% and 99.1%-100% using the high-specificity cutoff. DLAD showed significantly higher performance in both classification (0.993 vs 0.746-0.971) and localization (0.993 vs 0.664-0.925) compared to all groups of physicians.ConclusionsOur DLAD demonstrated excellent and consistent performance in the detection of active pulmonary tuberculosis on CR, outperforming physicians, including thoracic radiologists.
| S-EPMC6695514 | biostudies-literature
Project description:ImportanceInterpretation of chest radiographs is a challenging task prone to errors, requiring expert readers. An automated system that can accurately classify chest radiographs may help streamline the clinical workflow.ObjectivesTo develop a deep learning-based algorithm that can classify normal and abnormal results from chest radiographs with major thoracic diseases including pulmonary malignant neoplasm, active tuberculosis, pneumonia, and pneumothorax and to validate the algorithm's performance using independent data sets.Design, setting, and participantsThis diagnostic study developed a deep learning-based algorithm using single-center data collected between November 1, 2016, and January 31, 2017. The algorithm was externally validated with multicenter data collected between May 1 and July 31, 2018. A total of 54 221 chest radiographs with normal findings from 47 917 individuals (21 556 men and 26 361 women; mean [SD] age, 51 [16] years) and 35 613 chest radiographs with abnormal findings from 14 102 individuals (8373 men and 5729 women; mean [SD] age, 62 [15] years) were used to develop the algorithm. A total of 486 chest radiographs with normal results and 529 with abnormal results (1 from each participant; 628 men and 387 women; mean [SD] age, 53 [18] years) from 5 institutions were used for external validation. Fifteen physicians, including nonradiology physicians, board-certified radiologists, and thoracic radiologists, participated in observer performance testing. Data were analyzed in August 2018.ExposuresDeep learning-based algorithm.Main outcomes and measuresImage-wise classification performances measured by area under the receiver operating characteristic curve; lesion-wise localization performances measured by area under the alternative free-response receiver operating characteristic curve.ResultsThe algorithm demonstrated a median (range) area under the curve of 0.979 (0.973-1.000) for image-wise classification and 0.972 (0.923-0.985) for lesion-wise localization; the algorithm demonstrated significantly higher performance than all 3 physician groups in both image-wise classification (0.983 vs 0.814-0.932; all P < .005) and lesion-wise localization (0.985 vs 0.781-0.907; all P < .001). Significant improvements in both image-wise classification (0.814-0.932 to 0.904-0.958; all P < .005) and lesion-wise localization (0.781-0.907 to 0.873-0.938; all P < .001) were observed in all 3 physician groups with assistance of the algorithm.Conclusions and relevanceThe algorithm consistently outperformed physicians, including thoracic radiologists, in the discrimination of chest radiographs with major thoracic diseases, demonstrating its potential to improve the quality and efficiency of clinical practice.
| S-EPMC6583308 | biostudies-literature
Project description:PurposeTo describe the development and validation of a smartphone-based visual acuity (VA) test called Vision at home (V@home).MethodsThree study populations (elderly Chinese, adolescent Chinese, and Australian groups) underwent distance and near VA testing using standard Early Treatment Diabetic Retinopathy Study (ETDRS) charts and the V@home device; all VA tests used tumbling E optotypes. VA tests were repeated with one eye, selected randomly. Distance VA was measured monocularly at 2 m, and near VA was measured binocularly at 40 cm. Participants also completed a questionnaire about their satisfaction with the device. V@home VA (logMAR) was compared to VA for ETDRS charts at distance and near and test-retest reliability.ResultsThe mean difference between V@home and ETDRS distance VA across all groups ranged from -0.010 to -0.100 logMAR. Tolerant weighted kappa (TWK) agreement ranged from substantial (0.742) in the Australian group to almost perfect (0.950) in the adolescent Chinese group. There was high agreement of V@home with near ETDRS VA across all groups, with a mean difference of -0.092 to -0.042 logMAR and a TWK of 0.736 to 0.837. Test-retest reliability was also high (difference: -0.018 to 0.026) for both distance and near VA tests (95% limits of agreement: -0.289 to 0.258 for distance and -0.235 to 0.199 for near). The majority of participants were satisfied with V@home.ConclusionsV@home could accurately and reliably measure both distance and near VA and is well accepted by participants.Translational relevanceThe V@home system could potentially serve as a useful tool to improve eye care accessibility, especially in underdeveloped areas with limited eye care personnel and resources.
| S-EPMC6701871 | biostudies-literature
Project description:Background:This study developed, validated, and disseminated a generalizable informatics algorithm for detecting breast cancer recurrence and timing using a gold standard measure of recurrence coupled with data derived from a readily available common data model that pools health insurance claims and electronic health records data. Methods:The algorithm has two parts: to detect the presence of recurrence and to estimate the timing of recurrence. The primary data source was the Cancer Research Network Virtual Data Warehouse (VDW). Sixteen potential indicators of recurrence were considered for model development. The final recurrence detection and timing models were determined, respectively, by maximizing the area under the ROC curve (AUROC) and minimizing average absolute error. Detection and timing algorithms were validated using VDW data in comparison with a gold standard recurrence capture from a third site in which recurrences were validated through chart review. Performance of this algorithm, stratified by stage at diagnosis, was compared with other published algorithms. All statistical tests were two-sided. Results:Detection model AUROCs were 0.939 (95% confidence interval [CI] = 0.917 to 0.955) in the training data set (n?=?3370) and 0.956 (95% CI?=?0.944 to 0.971) and 0.900 (95% CI?=?0.872 to 0.928), respectively, in the two validation data sets (n?=?3370 and 3961, respectively). Timing models yielded average absolute prediction errors of 12.6% (95% CI?=?10.5% to 14.5%) in the training data and 11.7% (95% CI?=?9.9% to 13.5%) and 10.8% (95% CI?=?9.6% to 12.2%) in the validation data sets, respectively, and were statistically significantly lower by 12.6% (95% CI?=?8.8% to 16.5%, P < .001) than those estimated using previously reported timing algorithms. Similar covariates were included in both detection and timing algorithms but differed substantially from previous studies. Conclusions:Valid and reliable detection of recurrence using data derived from electronic medical records and insurance claims is feasible. These tools will enable extensive, novel research on quality, effectiveness, and outcomes for breast cancer patients and those who develop recurrence.
| S-EPMC5972574 | biostudies-literature
Project description:BackgroundOutside of a clinical setting, oscillometric devices make remote monitoring of blood pressure and virtual care more convenient and feasible. HeartBeat Technologies Ltd developed a novel approach to measuring blood pressure remotely after an initial blood pressure reading by a nurse using the conventional measurement method. Using a finger pulse oximeter, a photoplethysmogram wave is transmitted by Bluetooth to a smartphone or tablet. A smartphone app (MediBeat) transmits the photoplethysmogram to a server for analysis by a proprietary algorithm-the person's current blood pressure is sent back to the smartphone and to the individual's health care provider.ObjectiveThis study sought to determine whether the HeartBeat algorithm calculates blood pressure as accurately as required by the European Society of Hypertension International Protocol revision 2010 (ESH-IP2) for validation of blood pressure measuring devices.MethodsESH-IP2 requirements, modified to conform to a more recent international consensus statement, were followed. The ESH-IP2 establishes strict guidelines for the conduct and reporting of any validation of any device to measure blood pressure, including using the standard manual blood pressure instrument as a comparator and specific required accuracy levels for low, medium, and high ranges of blood pressure readings. The consensus statement requires a greater number of study participants for each of the blood pressure ranges. The validation of the accuracy of the algorithm was conducted with a Contec CMS50EW pulse oximeter and a Samsung Galaxy XCover 4 smartphone.ResultsThe differences between the HeartBeat-calculated and the manually measured blood pressures of 62 study participants did not meet the ESH-IP2 standards for accuracy for either systolic or diastolic blood pressure measurements. There was no discernible pattern in the inaccuracies of the HeartBeat-calculated measurements.ConclusionsThe October 4, 2019 version of the HeartBeat algorithm, implemented in combination with the MediBeat app, a pulse oximeter, and an Android smartphone, was not sufficiently accurate for use in a general adult population.Trial registrationClinicalTrials.gov NCT04082819; http://clinicaltrials.gov/ct2/show/NCT04082819.
| S-EPMC8105754 | biostudies-literature