New Linguistic Description Approach for Time Series and its Application to Bed Restlessness Monitoring for Eldercare.

Time series analysis has been an active area of research for years, with important applications in forecasting or discovery of hidden information such as patterns or anomalies in observed data. In recent years, the use of time series analysis techniques for the generation of descriptions and summaries in natural language of any variable, such as temperature, heart rate or CO2 emission has received increasing attention. Natural language has been recognized as more effective than traditional graphical representations of numerical data in many cases, in particular in situations where a large amount of data needs to be inspected or when the user lacks the necessary background and skills to interpret it. In this work, we describe a novel mechanism to generate linguistic descriptions of time series using natural language and fuzzy logic techniques. The proposed method generates quality summaries capturing the time series features that are relevant for a user in a particular application, and can be easily customized for different domains. This approach has been successfully applied to the generation of linguistic descriptions of bed restlessness data from residents at TigerPlace (Columbia, Missouri), which is used as a case study to illustrate the modeling process and show the quality of the descriptions obtained.

Corrections to "Cardiovascular Function and Ballistocardiogram: A Relationship Interpreted via Mathematical Modeling" [Oct 19 2906-2917].

Presents a missing funding acknowledgment for Dr. Mihail Popescu in the above named paper.

Cardiovascular Function and Ballistocardiogram: A Relationship Interpreted via Mathematical Modeling.

OBJECTIVE: To develop quantitative methods for the clinical interpretation of the ballistocardiogram (BCG). METHODS: A closed-loop mathematical model of the cardiovascular system is proposed to theoretically simulate the mechanisms generating the BCG signal, which is then compared with the signal acquired via accelerometry on a suspended bed. RESULTS: Simulated arterial pressure waveforms and ventricular functions are in good qualitative and quantitative agreement with those reported in the clinical literature. Simulated BCG signals exhibit the typical I, J, K, L, M, and N peaks and show good qualitative and quantitative agreement with experimental measurements. Simulated BCG signals associated with reduced contractility and increased stiffness of the left ventricle exhibit different changes that are characteristic of the specific pathological condition. CONCLUSION: The proposed closed-loop model captures the predominant features of BCG signals and can predict pathological changes on the basis of fundamental mechanisms in cardiovascular physiology. SIGNIFICANCE: This paper provides a quantitative framework for the clinical interpretation of BCG signals and the optimization of BCG sensing devices. The present paper considers an average human body and can potentially be extended to include variability among individuals.

Sleep Posture Classification Using Bed Sensor Data and Neural Networks.

Sleep posture has been shown to be important in monitoring health conditions such as congestive heart failure (CHF), sleep apnea, pressure ulcers, and even blood pressure abnormalities. In this paper, we investigate the use of four hydraulic bed transducers placed underneath the mattress to classify different sleep postures. For classification, we employed a simple neural network. Different combinations of parameters were studied to determine the best configuration. Data were collected on four major postures from 58 subjects. We report the results of classification for different combinations of these four postures. Both 10-Fold and Leave-One-Subject-Out (LOSO) Cross-validations (CV) were used to evaluate the accuracy of our predictions. Our results show that there are multiple configuration settings that make classification accuracy as high as 100% using k-Fold CV for all postures. Maximum classification accuracy after applying LOSO is 93% for a two-class classification of separating Left vs. Right lateral positions. The second-best classification accuracy with LOSO is 92% for the classification of lateral versus non-lateral.

Shear wave elastography and Afirma™ gene expression classifier in thyroid nodules with indeterminate cytology: a comparison study.

PURPOSE: To compare shear wave elastography (SWE) and Afirma™ gene expression classifier (GEC) for diagnosis of malignancy in thyroid nodules (TNs) with Bethesda Classification (BC) III or IV indeterminate cytology. METHODS: This preliminary single-center prospective study was approved by the Institutional Review Board. We evaluated 151 consented patients with 151 indeterminate TNs (123 BC III, 28 BC IV) on fine-needle aspiration biopsy (FNAB). B-mode ultrasound, vascularity, and SWE were performed prior to FNAB. TN stiffness was measured as shear wave velocity (SWV) in meters per second (m/s). The stiffest area of the TN was selected for SWV measurement. GEC testing was performed with a second FNAB. Surgery was recommended for GEC-suspicious TNs, or GEC-benign TNs with two or more worrisome B-mode US features. RESULTS: Surgical pathology confirmed 31 malignant TNs. Among the GEC-suspicious group, 28 of 59 TNs were malignant. The SWV value of ≥3.59 m/s was the best cut-off for malignancy risk based on the receiver operating curve (ROC). Twenty-six malignant TNs had SWV ≥ 3.59 m/s. The sensitivity and specificity for SWV ≥ 3.59 m/s were 83.9 and 79.2%, respectively. Positive predictive value (PPV) was 51.0% and negative predictive value (NPV) was 95.0%. For the GEC-suspicious group, sensitivity, specificity, PPV, and NPV were 90.3, 74.2, 47.5, and 96.7%, respectively. In multivariate analysis, SWV and GEC-suspicious were significant predictors of malignancy, but B-mode features and vascularity were not. CONCLUSION: This preliminary study indicates that SWE and GEC are independent predictors of malignancy in TNs with BC III or IV.

Shear wave elastography and parathyroid adenoma: A new tool for diagnosing parathyroid adenomas.

OBJECTIVES: This study prospectively determines the shear wave elastography characteristics of parathyroid adenomas using virtual touch imaging quantification, a non-invasive ultrasound based shear wave elastography method. METHODS: This prospective study examined 57 consecutive patients with biochemically proven primary hyperparathyroidism and solitary parathyroid adenoma identified by ultrasound and confirmed by at least one of the following: surgical resection, positive Technetium-99m Sestamibi Scintigraphy (MIBI) scan, or fine needle aspiration biopsy with positive PTH washout (performed only in MIBI negative patients). Vascularity and shear wave elastography were performed for all patients. Parathyroid adenoma stiffness was measured as shear wave velocity in meters per second. RESULTS: The median (range) pre-surgical value for PTH and calcium were 58pg/mL (19, 427) and 10.8mg/dL (9.5, 12.1), respectively. 37 patients had positive MIBI scan. 20 patients had negative MIBI scan but diagnosis was confirmed with positive PTH washout. 42 patients underwent parathyroidectomy, and an adenoma was confirmed in all. The median (range) shear wave velocity for all parathyroid adenomas enrolled in this study was 2.02m/s (1.53, 2.50). The median (range) shear wave velocity for thyroid tissue was 2.77m/s (1.89, 3.70). The shear wave velocity of the adenomas was independent of adenoma size, serum parathyroid hormone concentration, or plasma parathyroid hormone concentration. CONCLUSIONS: Tissue elasticity of parathyroid adenoma is significantly lower than thyroid tissue. B-mode features and distinct vascularity pattern are helpful tools in diagnosing parathyroid adenoma with ultrasound. Shear wave elastography may provide valuable information in diagnosing parathyroid adenoma.

Shear Wave Elastography and Cervical Lymph Nodes: Predicting Malignancy.

This prospective study evaluates the accuracy of virtual touch imaging quantification (VTIQ), a non-invasive shear wave elastography method for measuring cervical lymph nodes (LN) stiffness in differentiating benign from malignant LN. The study evaluated 270 LN in 236 patients with both conventional B-mode ultrasound and VTIQ shear wave elastography before fine-needle aspiration biopsy (FNAB). LN stiffness was measured as shear wave velocity (SWV) in m/s. Surgical resection was advised for FNAB results that were not clearly benign. Surgical pathology confirmed 54 malignant LN. The receiver operating curve (ROC) identified a single cut-off value of 2.93 m/s as the maximum SWV for predicting a malignant cervical LN. The sensitivity and specificity were 92.59% and 75.46%, respectively. Positive predictive value (PPV) was 48.54% and negative predictive value (NPV) was 97.60%. LN stiffness measured by VTIQ-generated shear wave elastography is an independent predictor of malignancy.

Sleep stage recognition using respiration signal.

This paper presents a sleep stage recognition system for Awake, rapid eye movement (REM) and non-REM (NREM) sleep detection. Two respiratory variability (RV) features are extracted from oro-nasal airflow signals provided in the sleep-EDF (Expanded) database. A two layer system with threshold comparison classifier is implemented. This system achieved state-of-the-art performance with simple features and classifiers. The average accuracy of 74.00%±5.30% and Cohen's kappa coefficient of 0.49±0.08 were achieved with 21 recordings. In the end, the measure of sleep efficiency was calculated and the average absolute error was 3.61%±3.66%.

Thyroid Nodules and Shear Wave Elastography: A New Tool in Thyroid Cancer Detection.

This study determines the performance of virtual touch imaging quantification (VTIQ), a non-invasive shear wave elastography method for measuring thyroid nodule (TN) stiffness, in distinguishing benign from malignant TNs. This prospective study evaluates 707 TNs in 676 patients with fine-needle aspiration biopsy (FNAB). Before FNAB, both conventional B-mode ultrasound and shear wave elastography were performed. Surgical resection was recommended for FNAB results that were not clearly benign. Surgical pathology confirmed 82 malignant TNs. The receiver operating curve identified a single cut-off of 3.54 m/s as the maximum shear wave velocity (SWV) for predicting thyroid cancer (TC). The sensitivity and specificity were 79.27% and 71.52%, respectively. Positive predictive value (PPV) was 26.75% and negative predictive value (NPV) was 96.34%. Compared with B-mode US features for predicting malignancy, SWV ≥3.54 m/s has a higher sensitivity, specificity, PPV and NPV. TN stiffness measured by VTIQ-generated shear wave elastography is an independent predictor of TC.

Textual summarization of events leading to health alerts.

Extracting information from the sensors installed in the homes of elderly pose a unique set of challenges. Add to it the short amount of time the clinicians and nurses have to analyze this data, and the problem becomes more complicated. A system already in place at an "Aging in Place" facility monitors the activities of residents through multiple non-intrusive sensors and sends alerts on detecting an unusual event. We present an approach to generate textual summaries of events leading to the alerts. We analyze our system using four case studies and also list the comments provided by collaborators in healthcare domain. The system was then iterated to take some of those suggestions into account to give a glimpse of what an ideal system should look like.

Sit-to-stand measurement for in-home monitoring using voxel analysis.

We present algorithms to segment the activities of sitting and standing, and identify the regions of sit-to-stand (STS) transitions in a given image sequence. As a means of fall risk assessment, we propose methods to measure STS time using the 3-D modeling of a human body in voxel space as well as ellipse fitting algorithms and image features to capture orientation of the body. The proposed algorithms were tested on ten older adults with ages ranging from 83 to 97. Two techniques in combination yielded the best results, namely the voxel height in conjunction with the ellipse fit. Accurate STS time was computed on various STSs and verified using a marker-based motion capture system. This application can be used as part of a continuous video monitoring system in the homes of older adults and can provide valuable information to help detect fall risk and enable early interventions.

Monitoring patients in hospital beds using unobtrusive depth sensors.

We present an approach for patient activity recognition in hospital rooms using depth data collected using a Kinect sensor. Depth sensors such as the Kinect ensure that activity segmentation is possible during day time as well as night while addressing the privacy concerns of patients. It also provides a technique to remotely monitor patients in a non-intrusive manner. An existing fall detection algorithm is currently generating fall alerts in several rooms in the University of Missouri Hospital (MUH). In this paper we describe a technique to reduce false alerts such as pillows falling off the bed or equipment movement. We do so by detecting the presence of the patient in the bed for the times when the fall alert is generated. We test our algorithm on 96 hours obtained in two hospital rooms from MUH.

Toward a passive low-cost in-home gait assessment system for older adults.

In this paper, we propose a webcam-based system for in-home gait assessment of older adults. A methodology has been developed to extract gait parameters including walking speed, step time, and step length from a 3-D voxel reconstruction, which is built from two calibrated webcam views. The gait parameters are validated with a GAITRite mat and a Vicon motion capture system in the laboratory with 13 participants and 44 tests, and again with GAITRite for 8 older adults in senior housing. Excellent agreement with intraclass correlation coefficients of 0.99 and repeatability coefficients between 0.7% and 6.6% was found for walking speed, step time, and step length given the limitation of frame rate and voxel resolution. The system was further tested with ten seniors in a scripted scenario representing everyday activities in an unstructured environment. The system results demonstrate the capability of being used as a daily gait assessment tool for fall risk assessment and other medical applications. Furthermore, we found that residents displayed different gait patterns during their clinical GAITRite tests compared to the realistic scenario, namely a mean increase of 21% in walking speed, a mean decrease of 12% in step time, and a mean increase of 6% in step length. These findings provide support for continuous gait assessment in the home for capturing habitual gait.

A Cluster Validity Framework Based on Induced Partition Dissimilarity.

We describe a new cluster validity framework (CVF) that compares structure in the data (in dissimilarity form) to the structure of dissimilarity matrices induced by a matrix transformation of the partition being tested. As part of this framework, we show two possible cluster validation measures: one, visual cluster validity, that that uses visual comparison and another one, correlation cluster validity, based on correlation. Unlike many existing measures, the measures we propose can be applied to crisp or soft partitions obtained by any relational or object data clustering algorithm. We illustrate the new measures and compare them to several well-known existing measures using real and artificial data sets.

Genome-wide linkage analyses identify Hfhl1 and Hfhl3 with frequency-specific effects on the hearing spectrum of NIH Swiss mice.

BACKGROUND: The mammalian cochlea receives and analyzes sound at specific places along the cochlea coil, commonly referred to as the tonotopic map. Although much is known about the cell-level molecular defects responsible for severe hearing loss, the genetics responsible for less severe and frequency-specific hearing loss remains unclear. We recently identified quantitative trait loci (QTLs) Hfhl1 and Hfhl2 that affect high-frequency hearing loss in NIH Swiss mice. Here we used 2f1-f2 distortion product otoacoustic emissions (DPOAE) measurements to refine the hearing loss phenotype. We crossed the high frequency hearing loss (HFHL) line of NIH Swiss mice to three different inbred strains and performed linkage analysis on the DPOAE data obtained from the second-generation populations. RESULTS: We identified a QTL of moderate effect on chromosome 7 that affected 2f1-f2 emissions intensities (Hfhl1), confirming the results of our previous study that used auditory brainstem response (ABR) thresholds to identify QTLs affecting HFHL. We also identified a novel significant QTL on chromosome 9 (Hfhl3) with moderate effects on 2f1-f2 emissions intensities. By partitioning the DPOAE data into frequency subsets, we determined that Hfhl1 and Hfhl3 affect hearing primarily at frequencies above 24 kHz and 35 kHz, respectively. Furthermore, we uncovered additional QTLs with small effects on isolated portions of the DPOAE spectrum. CONCLUSIONS: This study identifies QTLs with effects that are isolated to limited portions of the frequency map. Our results support the hypothesis that frequency-specific hearing loss results from variation in gene activity along the cochlear partition and suggest a strategy for creating a map of cochlear genes that influence differences in hearing sensitivity and/or vulnerability in restricted portions of the cochlea.

Resident identification using kinect depth image data and fuzzy clustering techniques.

As a part of our passive fall risk assessment research in home environments, we present a method to identify older residents using features extracted from their gait information from a single depth camera. Depth images have been collected continuously for about eight months from several apartments at a senior housing facility. Shape descriptors such as bounding box information and image moments were extracted from silhouettes of the depth images. The features were then clustered using Possibilistic C Means for resident identification. This technology will allow researchers and health professionals to gather more information on the individual residents by filtering out data belonging to non-residents. Gait related information belonging exclusively to the older residents can then be gathered. The data can potentially help detect changes in gait patterns which can be used to analyze fall risk for elderly residents by passively observing them in their home environments.

High-frequency sensorineural hearing loss and its underlying genetics (Hfhl1 and Hfhl2) in NIH Swiss mice.

Studies using inbred strains of mice have been invaluable for identifying alleles that adversely affect hearing. However, the efficacy of those studies is limited by the phenotypes that these strains express and the alleles that they segregate. Here, by selectively breeding phenotypically and genetically heterogeneous NIH Swiss mice, we generated two lines-the all-frequency hearing loss (AFHL) line and the high-frequency hearing loss (HFHL) line-with differential hearing loss. The AFHL line exhibited characteristics typical of severe, early-onset, sensorineural hearing impairment. In contrast, the HFHL line expressed a novel early-onset, mildly progressive, and frequency-specific sensorineural hearing loss. By quantitative trait loci (QTLs) analyses in these two lines, we identified QTLs on chromosomes 7, 8, and 10 that significantly affected hearing function. The loci on chromosomes 7 and 8 (Hfhl1 and Hfhl2, respectively) are novel and appear to adversely affect only high frequencies (≥30 kHz). Mice homozygous for NIH Swiss alleles at either Hfhl1 or Hfhl2 have 32-kHz auditory-evoked brain stem response thresholds that are 8-14 dB SPL higher than the corresponding heterozygotes. DNA sequence analyses suggest that both the Cdh23(ahl) and Gipc3(ahl5) variants contribute to the chromosome 10 QTL detected in the AFHL line. The frequency-specific hearing loss indicates that the Hfhl1 and Hfhl2 alleles may affect tonotopic development. In addition, dissecting the underlying complex genetics of high-frequency hearing loss may prove relevant in identifying less severe and common forms of hearing impairment in the human population.

Quantitative analysis of 180 degree turns for fall risk assessment using video sensors.

In this paper, we present a method for quantitatively and objectively assessing 180 degree turns using low cost video sensors. A three-dimensional voxel reconstruction, which is built using silhouettes captured from two calibrated web camera views, is used to represent the human body. Experiments were conducted in which participants performed the standard Timed Up and Go tests where 180 degree turns are evaluated. Our two calibrated cameras captured the images during the test. Two key parameters including turn time and turn steps are extracted using the voxel data. Good agreement for the turn time was found for our system compared to the expert rating. The extracted numbers of turn steps are one step less than the expert rating in many test runs. The difference comes mainly from the nature of the pivot turns, and the turn time difference between the expert rating and the algorithm, namely the determination of the time duration from the beginning to the end of the turn. The development of this technology provides potential for assessing 180 degree turns in the home setting as part of a balance, stability and fall risk assessment tool.

Extracting footfalls from voxel data.

In this paper, we present a method for extracting footfall locations from three dimensional voxel data created from a pair of silhouettes. With the growth of the elderly population, there is a need for passive monitoring of physical activity to allow older adults to continue living in independent settings. Prior research using anonymized video data has shown good results in passively acquiring information useful for assessing physical function; and, additionally, research has shown that video data anonymized through the use of silhouettes alleviates privacy concerns of older adults towards the technology. Previous work in acquiring gait information from voxel data has not included a technique for identifying individual footfall locations, from which additional information useful for assessing asymmetric gait patterns and other physical parameters may be obtained. Furthermore, visualization of the footfall locations during a walking sequence may provide additional insight to care providers for assessing physical function. To evaluate our approach, participants were asked to walk across a GAITRite electronic mat, used to validate our results, while also being monitored by our camera system. Results show good agreement between the footfalls extracted by our system and those from the GAITRite.

Body sway measurement for fall risk assessment using inexpensive webcams.

In this paper, we present a method for extracting body sway parameters from a three-dimensional voxel reconstruction, which is built using silhouettes captured from two calibrated web camera views. The results were validated with a Vicon motion capture system. Experiments were conducted in which subjects stand and sway in the anterior-posterior direction and then in the lateral directions with two different frequencies. In addition, experiments were also conducted where subjects walked in a straight path at different speeds. Through the experiment, the Vicon cameras recorded the motion of reflective markers attached to subjects, and our two calibrated cameras captured the images. Good agreement was found with our system compared to the Vicon results, given the limitation of voxel space resolution and frame rate. The development of this technology provides potential capability of measuring body sway in daily living environment for elderly people, and can be used as part of a balance, stability and fall risk assessment tool.