Search

A dominant sigmoid sinus with either diverticulum or dehiscence (SS-Div/SS-Deh) is a common cause of pulsatile tinnitus (PT). For PT originating from SS-Div/SS-Deh, an etiology-specific and secure reconstruction using firm materials is vital for optimal outcomes. As a follow-up to our previous reports on transmastoid SS resurfacing or reshaping for SS-Div/SS-Deh, this study aimed to evaluate the long-term results of transmastoid resurfacing/reshaping. We retrospectively reviewed 20 PT patients who were diagnosed with SS-Div/SS-Deh, underwent transmastoid resurfacing/reshaping, and were followed up for more than 1 year postoperatively. For PT, immediate and long-term changes (> 1 year) in loudness and annoyance were analyzed using the visual analog scale (VAS). Additionally, pre and postoperative objective measurements of PT using transcanal sound recording and spectro-temporal analysis (TSR-STA), imaging results, and audiological findings were comprehensively analyzed. Significant improvements in PT were sustained or enhanced for > 1 year (median follow-up period: 37 months, range: 12–54 months). On TSR-STA, both peak and root mean square amplitudes decreased after surgery. Also, the average pure-tone threshold at 250 Hz improved after surgery. Thus, our long-term follow-up data confirmed that the surgical management of PT originating from SS-Div/SS-Deh is successful with regard to both objective and subjective measures.

To predict the apnea-hypopnea index (AHI) in patients with obstructive sleep apnea (OSA) using data from breathing sounds recorded using a noncontact device during sleep.Prospective cohort study.Tertiary referral hospital.Audio recordings during sleep were performed using an air-conduction microphone during polysomnography. Breathing sounds recorded from all sleep stages were analyzed. After noise reduction preprocessing, the audio data were segmented into 5-second windows and sound features were extracted. Estimation of AHI by regression analysis was performed using a Gaussian process, support vector machine, random forest, and simple linear regression, along with 10-fold cross-validation.In total, 116 patients who underwent attended, in-laboratory, full-night polysomnography were included. Overall, random forest resulted in the highest performance with the highest correlation coefficient (0.83) and least mean absolute error (9.64 events/h) and root mean squared error (13.72 events/h). Other models resulted in somewhat lower but similar performances, with correlation coefficients ranging from 0.74 to 0.79. The estimated AHI tended to be underestimated as the severity of OSA increased. Regarding bias and precision, estimation performances in the severe OSA subgroup were the lowest, regardless of the model used. Among sound features, derivative of the area methods of moments of overall standard deviation demonstrated the highest correlation with AHI.AHI was fairly predictable by using data from breathing sounds generated during sleep. The prediction model may be useful not only for prescreening but also for follow-up after treatment in patients with OSA.

Objectives The current study sought to evaluate whether nursing narratives can be used to predict postoperative length of hospital stay (LOS) following curative surgery for ovarian cancer. Methods A total of 33 patients, aged over 65 years, underwent curative surgery for newly diagnosed ovarian cancer between 2008 and 2012. Based on the median postoperative LOS, patients were divided into two groups: long-stay (>12 days; n = 13) and short-stay (≤12 days; n = 20). Patterns in nursing narratives were examined and compared through a quantitative analysis. Specifically, the total number (TN) of narratives pertaining to care and the standardized number (SN), which was calculated by dividing the TN by the LOS, were compared. Experts evaluated the relevance of the phrases extracted. LOS was then predicted using machine learning techniques. Results The median postoperative LOS was 18 days (interquartile range [IQR]: 16–24 days) in the long-stay group and 9.5 days (IQR: 8–11.25 days) in the short-stay group. In the long-stay group, surgery duration was longer. Overall, patients in the long-stay group showed a higher volume of nursing narratives compared with patients in the short-stay group (SN: 68 vs. 46, p = 0.021). Thirty-two of the most frequently used nursing narratives were selected from 998 uniquely defined nursing narratives. Multiple t-tests were used to compare the TN and real standardized number (RSN; minimum p Conclusions Statistical tests showed that nursing narratives that utilized the words “urination,” “food supply,” “bowel mobility,” or “pain” were related to hospital stay in elderly females with ovarian cancer. Additionally, machine learning effectively predicted LOS. Summary The current study sought to determine whether elements of nursing narratives could be used to predict postoperative LOS among elderly ovarian cancer patients. Results indicated that nursing narratives that used the words “urination,” “food supply,” “bowel mobility,” and “pain” significantly predicted postoperative LOS in the study population. Additionally, it was found that machine learning could effectively predict LOS based on quantitative characteristics of nursing narratives.

OBJECTIVE To investigate the effects of cavitating lesions involving the internal auditory canal (IAC) in subjects with cochlear otosclerosis with regard to poststapedotomy hearing outcome. STUDY DESIGN Retrospective study. SETTING Tertiary referral center. PATIENTS A retrospective chart review of 134 subjects with otosclerosis treated from January 2011 to June 2017 at Seoul National University Bundang Hospital was conducted. Sixteen subjects (23 ears) with temporal bone computed tomography (TBCT)-confirmed cochlear otosclerosis who underwent stapedotomy were included in the study. MAIN OUTCOME MEASURES Pure tone audiometry (PTA) (i.e., air and bone conduction; AC and BC, respectively) thresholds and air-bone gap (ABG), measured at 6 months postoperatively were compared between cochlear otosclerosis with and without IAC involvement (IAC group and non-IAC group, respectively). RESULTS A total of 14 of 23 ears showed involvement of the IAC. There were no significant differences in age, side of otosclerosis, or preoperative hearing threshold between the two groups. The mean postoperative AC and BC thresholds and ABG of the IAC group were significantly poorer (45.7 dB, 33.8 dB, and 11.8 dB, respectively) than those of the nonIAC group (24.1 dB, 20.0 dB, and 4.1 dB, respectively). CONCLUSIONS Cochlear otosclerosis with cavitating lesions involving the IAC showed significantly poorer postoperative audiological outcomes than those without any cavitating lesion. Cavitation extending to the IAC may act as a third window providing a route for sound energy shunting, and thus precluding successful hearing outcome in some subjects with cavitating otosclerosis after stapedotomy.

Modern deep learning-based models have seen outstanding performance improvement with speech enhancement tasks. The number of parameters of state-of-the-art models, however, is often too large to be deployed on devices for real-world applications. To this end, we propose Tiny Recurrent U-Net (TRU-Net), a lightweight online inference model that matches the performance of current state-of- the-art models. The size of the quantized version of TRU-Net is 362 kilobytes, which is small enough to be deployed on edge devices. In addition, we combine the small-sized model with a new masking method called phase-aware s-sigmoid mask, which enables simultaneous denoising and dereverberation. Results of both objective and subjective evaluations have shown that our model can achieve competitive performance with the current state-of-the-art models on benchmark datasets using fewer parameters by orders of magnitude.

Collecting and accessing a large amount of medical data is very time-consuming and laborious, not only because it is difficult to find specific patients but also because it is required to resolve the confidentiality of a patient's medical records. On the other hand, there are deep learning models, trained on easily collectible, large scale datasets such as Youtube or Wikipedia, offering useful representations. It could therefore be very advantageous to utilize the features from these pre-trained networks for handling a small amount of data at hand. In this work, we exploit various multi-modal features extracted from pre-trained networks to recognize Alzheimer's Dementia using a neural network, with a small dataset provided by the ADReSS Challenge at INTERSPEECH 2020. The challenge regards to discern patients suspicious of Alzheimer's Dementia by providing acoustic and textual data. With the multi-modal features, we modify a Convolutional Recurrent Neural Network based structure to perform classification and regression tasks simultaneously and is capable of computing conversations with variable lengths. Our test results surpass baseline's accuracy by 18.75%, and our validation result for the regression task shows the possibility of classifying 4 classes of cognitive impairment with an accuracy of 78.70%.

Quantitative evaluation of piano performance is of interests in many fields, including music education and computational performance rendering. Previous studies utilized features extracted from audio or musical instrument digital interface (MIDI) files but did not address the difference between hands (DBH), which might be an important aspect of high-quality performance. Therefore, we investigated DBH as an important factor determining performance proficiency. To this end, 34 experts and 34 amateurs were recruited to play two excerpts on a Yamaha Disklavier. Each performance was recorded in MIDI, and handcrafted features were extracted separately for the right hand (RH) and left hand (LH). These were conventional MIDI features representing temporal and dynamic attributes of each note and computed as absolute values (e. g., MIDI velocity) or ratios between performance and corresponding scores (e. g., ratio of duration or inter-onset interval (IOI)). These note-based features were rearranged into additional features representing DBH by simple subtraction between features of both hands. Statistical analyses showed that DBH was more significant in experts than in amateurs across features. Regarding temporal features, experts pressed keys longer and faster with the RH than did amateurs. Regarding dynamic features, RH exhibited both greater values and a smoother change along melodic intonations in experts that in amateurs. Further experiments using principal component analysis (PCA) and support vector machine (SVM) verified that hand-difference features can successfully differentiate experts from amateurs according to performance proficiency. Moreover, existing note-based raw feature values (Basic features) and DBH features were tested repeatedly via 10-fold cross-validation, suggesting that adding DBH features to Basic features improved F1 scores to 93.6% (by 3.5%) over Basic features. Our results suggest that differently controlling both hands simultaneously is an important skill for pianists; therefore, DBH features should be considered in the quantitative evaluation of piano performance.

Researchers and service providers have focused on leveraging social information acquired from interactions between users to improve the accuracy of system recommendations. However, few have explained the characteristics of music recommendations through interpersonal relationships. To investigate how interpersonal relationships affect users’ evaluation of music recommendation, we conducted a survey-based study that compared two types of recommendation channels—interpersonal (i.e., from friends) and non-interpersonal (i.e., from systems). We found that relevance was evaluated higher in music recommended from non-interpersonal channels on average, while diversity, novelty, and serendipity were higher in interpersonal channels. Non-interpersonal channels surpassed interpersonal channels in terms of convenience, frequency, and adoption rate. These results illustrate that interpersonal and non-interpersonal channels have different strengths and that digital streaming platforms, which have mainly provided system recommendations thus far, need to better support interpersonal channels for richer user experience.

Background Vocal acoustic features are potential biomarkers of elderly depression. Previous automated diagnostic tests for depression have employed unstandardized speech samples, and few studies have considered differences in voice reactivity. We aimed to develop a voice-based screening test for depression measuring vocal acoustic features of elderly Koreans while they read a series of mood-inducing sentences (MIS). Methods In this case-control study, we recruited 61 individuals with major depressive disorder and 143 healthy controls (mean age [SD]: 72 [6]; female, 70%) from the community-dwelling elderly population. Participants were asked to read MIS and their variation pattern of acoustic features represented by the correlation distance between two MIS were analyzed as input features using the univariate feature selection technique and subsequently classified by AdaBoost. Results Acoustic features showing significant discriminatory performances were spectral and energy-related features for males (sensitivity 0.95, specificity 0.88, and accuracy 0.86) and prosody-related features for females (sensitivity 0.73, specificity 0.86, and accuracy 0.77). The correlation distance between negative and positive MIS was significantly shorter in the depressed group than in the healthy control (F = 18.574, P Limitations Small sample size and relatively homogenous clinical profile of depression could limit the generalizability. Conclusions While reading MIS, spectral and energy-related acoustic features for males and prosody-related features for females are good discriminators for major depressive disorder. These features may be used as biomarkers of depression in the elderly.

In this study, we investigate the possibility of protecting audio classification datasets used in deep learning by embedding a pattern in the magnitude of the time-frequency representation of a subset of the dataset. Previous studies on audio watermarking technologies require the actual sound of the watermarked audio to extract the information embedded in it. In our study, we propose an audio watermarking framework aimed to identify whether a deep learning based audio classification model is trained with the watermarked audio classification dataset or not by using only the classification results. The experimental results show that our proposed method can identify the usage of an audio classification dataset while having minimal effect on the overall classification performance. The results are consistent with three different audio classification datasets. The proposed method is robust to different types and parameters of time-frequency representations and classification models.

Objectives To develop a simple algorithm for prescreening of obstructive sleep apnea (OSA) on the basis of respiratory sounds recorded during polysomnography during all sleep stages between sleep onset and offset. Methods Patients who underwent attended, in-laboratory, full-night polysomnography were included. For all patients, audio recordings were performed with an air-conduction microphone during polysomnography. Analyses included all sleep stages (i.e., N1, N2, N3, rapid eye movement, and waking). After noise reduction preprocessing, data were segmented into 5-s windows and sound features were extracted. Prediction models were established and validated with 10-fold cross-validation by using simple logistic regression. Binary classifications were separately conducted for three different threshold criteria at apnea hypopnea index (AHI) of 5, 15, or 30. Prediction model characteristics, including accuracy, sensitivity, specificity, positive predictive value (precision), negative predictive value, and area under the curve (AUC) of the receiver operating characteristic were computed. Results A total of 116 subjects were included; their mean age, body mass index, and AHI were 50.4 years, 25.5 kg/m2 , and 23.0/hr, respectively. A total of 508 sound features were extracted from respiratory sounds recorded throughout sleep. Accuracies of binary classifiers at AHIs of 5, 15, and 30 were 82.7%, 84.4%, and 85.3%, respectively. Prediction performances for the classifiers at AHIs of 5, 15, and 30 were AUC, 0.83, 0.901, and 0.91; sensitivity, 87.5%, 81.6%, and 60%; and specificity, 67.8%, 87.5%, and 94.1%. Respective precision values of the classifiers were 89.5%, 87.5%, and 78.2% for AHIs of 5, 15, and 30. Conclusion This study showed that our binary classifier predicted patients with AHI of ≥15 with sensitivity and specificity of >80% by using respiratory sounds during sleep. Since our prediction model included all sleep stage data, algorithms based on respiratory sounds may have a high value for prescreening OSA with mobile devices.

Purpose Breathing sounds during sleep are altered and characterized by various acoustic specificities in patients with sleep disordered breathing (SDB). This study aimed to identify acoustic biomarkers indicative of the severity of SDB by analyzing the breathing sounds collected from a large number of subjects during entire overnight sleep. Methods The participants were patients who presented at a sleep center with snoring or cessation of breathing during sleep. They were subjected to full-night polysomnography (PSG) during which the breathing sound was recorded using a microphone. Then, audio features were extracted and a group of features differing significantly between different SDB severity groups was selected as a potential acoustic biomarker. To assess the validity of the acoustic biomarker, classification tasks were performed using several machine learning techniques. Based on the apnea–hypopnea index of the subjects, four-group classification and binary classification were performed. Results Using tenfold cross validation, we achieved an accuracy of 88.3% in the four-group classification and an accuracy of 92.5% in the binary classification. Experimental evaluation demonstrated that the models trained on the proposed acoustic biomarkers can be used to estimate the severity of SDB. Conclusions Acoustic biomarkers may be useful to accurately predict the severity of SDB based on the patient’s breathing sounds during sleep, without conducting attended full-night PSG. This study implies that any device with a microphone, such as a smartphone, could be potentially utilized outside specialized facilities as a screening tool for detecting SDB.

In this paper, we present a novel method for harmonic-percussive sound separation (HPSS) by exploiting continuity/discontinuity properties in a matrix decomposition framework. It is widely accepted in the HPSS research that the harmonic and percussive components have anisotropic characteristics: The spectrum of the harmonic components and the time activation of the percussive components are sparse, whereas the spectrum of the percussive components and the time activation of the harmonic components are smooth. However, conventional methods fail to fully utilize the characteristics leading to suboptimal performance. Based on the observations that not the degree of sparseness but the degree of fluctuation is an accurate measure for distinguishing the harmonic and percussive components, we propose a novel HPSS algorithm by incorporating the continuity control in the iterative update formula of the matrix decomposition algorithm. In doing so, we first utilize probabilistic latent component analysis with Dirichlet prior, and later reformulate the algorithm in the nonnegative matrix factorization framework to reduce the computational cost. The comparative evaluation results show that the proposed method outperforms conventional methods in terms of both objective and subjective evaluation.

OBJECTIVE: To find relationship between sleep breathing sounds and various polysomnographic features. METHOD: Patients who underwent PSG with audio recording using a microphone were included. After noise reduction preprocessing, the data were segmented into 5-second windows and sound features were extracted. A simple linear regression of various sound features were performed with various PSG features including apnea hypopnea index (AHI), apnea index (AI), hypopnea index (HI), Oxygen desaturation index (ODI), and respiratory arousal (RA). Also prediction of PSG features from sleep breathing sounds had been established using machine learning. RESULT: A total of 116 subjects were included. A total of 508 sound features were extracted. Among them, Derivative of Area Method of Moments Overall Standard Deviation was the most distinguishing feature that explained AHI (R2= 0.628), AI (R2=0.620), ODI (R2= 0.641), SpO2 CONCLUSION: Similar prediction values were acquired with simple linear regression and machine learning. Especially machine learning resulted in a better performance regarding to AI and Time of SpO2

We reviewed the clinical characteristics and treatment outcomes of patients with glomus tympanicum tumors (GTTs) presenting with pulsatile tinnitus (PT). We explored whether transcanal sound recording-spectro-temporal analysis (TSR-STA) usefully evaluated changes in PT. The medical records of 13 patients who underwent surgical removal of GTTs were reviewed retrospectively. Two patients underwent preoperative endovascular embolization. Changes in PT, pre- and postoperative audiometry data, TSR-STA results, and clinical outcomes were evaluated. PT was the chief complaint in eight patients (61.5%) and resolved immediately after surgical intervention in all. Two patients exhibited ipsilateral, pseudo-low-frequency hearing loss (PLFHL); surgical GTT removal elicited postoperative improvements in the ipsilesional low-frequency hearing thresholds. Five patients underwent TSR-STA using previously described methods. TSR-STA revealed definite rise-and-fall patterns; surgical tumor removal abated this pattern in one patient, but, for the other four, the patterns did not change greatly post-intervention. Thus, GTT-related PT can be treated successfully (via surgical GTT removal) without complications. In selected cases, preoperative embolization reduces intraoperative hemorrhage. In PT patients with PLFHL, a detailed otoendoscopic examination of the middle ear is required to rule out a GTT. TSR-STA may usefully (and objectively) assess postoperative improvements in GTT-related PT.

OBJECTIVE A dominant sigmoid sinus with focal dehiscence or thinning (DSSD/T) of the overlying bony wall is a commonly encountered, but frequently overlooked, cause of vascular pulsatile tinnitus (VPT). Also, the pathophysiological mechanism of sound perception in patients with VPT remains poorly understood. In the present study, a novel surgical method, termed transmastoid SS-reshaping surgery, was introduced to ameliorate VPT in patients with DSSD/T. The authors reviewed a case series, analyzed the surgical outcomes, and suggested the pathophysiological mechanism of sound perception. The theoretical background underlying VPT improvement after transmastoid SS-reshaping surgery was also explored. METHODS Eight patients with VPT that was considered attributable to DSSD/T underwent transmastoid SS-reshaping surgery between February 2010 and February 2015. The mean postoperative follow-up period was 9.5 months (range 4–13 months). Transmastoid SS-reshaping surgery featured simple mastoidectomy, partial compression of the SS using harvested cortical bone chips, and reinforcement of the bony SS wall with bone cement. Perioperative medical records, imaging results, and audiological findings were comprehensively reviewed. RESULTS In 7 of the 8 patients (87.5%), the VPT abated immediately after surgery. Statistically significant improvements in tinnitus loudness and distress were evident on numeric rating scales. Three patients with preoperative ipsilesional low-frequency hearing loss exhibited postoperative improvements in their low-frequency hearing thresholds. No major postoperative complications were encountered except in the first subject, who experienced increased intracranial pressure postoperatively. This subsided after a revision operation for partial decompression of the SS. CONCLUSIONS Transmastoid SS-reshaping surgery may be a good surgical option in patients with DSSD/T, a previously unrecognized cause of VPT. Redistribution of severely asymmetrical blood flow, reinforcement of the bony SS wall with bone cement to reconstruct a soundproof barrier, and disconnection of a problematic sound conduction route via simple mastoidectomy silence VPT.

Although frequently classified as "objective tinnitus," in most cases vascular pulsatile tinnitus (VPT) is not equal to objective tinnitus because it is not easy to objectively document VPT. The present study was conducted to develop a novel transcanal sound recording and spectrotemporal analysis method for the objective and differential diagnosis of VPT.A case series with a control group.Tertiary referral center.Six VPT subjects with radiological abnormalities and six normal controls.The method was tested based on recordings obtained from the ipsilateral external auditory canal (EAC) using an insert microphone with the subject's head in four different positions. The recorded signals were first analyzed in the time domain, and short-time Fourier transform was performed to analyze the data in the time-frequency domain.From the temporal analysis, the ear canal signals recorded from the VPT subjects exhibited large peak amplitudes and periodic structures, whereas the signals recorded from the control subjects had smaller peak amplitudes and weaker periodicity. From the STA represented by two-dimensional spectrograms and three-dimensional waterfall diagrams, all of the VPT subjects demonstrated pulse-synchronous acoustic characteristics that were representative of their respective presumptive vascular pathologies, whereas the control subjects did not display such characteristics.The present diagnostic approach may provide additional information regarding the origins of VPT cases as well as an efficient and objective diagnostic method. Furthermore, this approach may aid in the determination of appropriate imaging modalities, treatment planning, and evaluation of treatment outcomes.

In this study, we propose an application for immersive virtual reality experiences, which integrates three-dimensional (3D) head-mounted displays (HMDs) with a precise indoor position tracking algorithm based on ultrasound. Our method provides a natural virtual reality experience with interaction by precisely matching the physical movements in the real world with those in the virtual environment, unlike other methods that require external input devices to move around in the virtual environment. Users can move within the assigned indoor space while carrying a wireless client device with the HMD, without the risk of colliding with obstacles or structures. The system is designed to provide the accurate 3D X, Y, and Z coordinate values of translational movements in real-time as well as the pitch, roll, and yaw values of rotational movements supported by the HMD, resulting in the six degrees of freedom required by immersive virtual reality. In addition, the system utilizes ultrasonic transducers in a grid format, which makes it simple to expand the position tracking coverage area, and supports simultaneous tracking of multiple users. Through experiments and a user study we show that the system obtains the accurate position of the moving objects and delivers a highly immersive virtual reality experience.

In this paper, we propose a cover song identification algorithm using a convolutional neural network (CNN). We first train the CNN model to classify any non-/cover relationship, by feeding a cross-similarity matrix that is generated from a pair of songs as an input. Our main idea is to use the CNN output–the cover-probabilities of one song to all other candidate songs–as a new representation vector for measuring the distance between songs. Based on this, the present algorithm searches cover songs by applying several ranking methods: 1. sorting without using the representation vectors; 2. the cosine distance between the representation vectors; and 3. the correlation between the vectors. In our experiment, the proposed algorithm significantly outperformed the algorithms used in recent studies, by achieving a mean average precision (MAP) of 93.18% in a dataset consisting of 3,300 cover-pairs and 496,200 non-cover-pairs.

This study aimed to determine the factors influencing patients’ choice of physician at the first visit through database analysis of a tertiary hospital in South Korea. We collected data on the first treatments performed by physicians who had treated patients for at least 3 consecutive years over 10 years (from 2003 to 2012) from the database of Seoul National University’s affiliated tertiary hospital. Ultimately, we obtained data on 524,012 first treatments of 319,004 patients performed by 115 physicians. Variables including physicians’ age and medical school and patients’ age were evaluated as influencing factors for the number of first treatments performed by each physician in each year using a Poisson regression through generalized estimating equations with a log link. The number of first treatments decreased over the study period. Notably, the relative risk for first treatments was lower among older physicians than among younger physicians (relative risk 0.96; 95% confidence interval 0.95 to 0.98). Physicians graduating from Seoul National University (SNU) also had a higher risk for performing first treatments than did those not from SNU (relative risk 1.58; 95% confidence interval 1.18 to 2.10). Finally, relative risk was also higher among older patients than among younger patients (relative risk 1.03; 95% confidence interval 1.01 to 1.04). This study systematically demonstrated that physicians’ age, whether the physician graduated from the highest-quality university, and patients’ age all related to patients’ choice of physician at the first visit in a tertiary university hospital. These findings might be due to Korean cultural factors.

A recommender system based on a positively-related item-graph targeted for novel and relevant recommendations is proposed.A live test was performed comparing the proposed system with a state-of-the-art matrix factorization algorithm.The proposed system consistently outperforms matrix factorization in finding items that are both novel and relevant.By finding novel and relevant items, the system addresses popularity bias commonly found in collaborative filtering-based recommender systems. Recommender systems have steadily advanced in their ability to filter out unnecessary information and deliver the most relevant data to users. Such recommender systems are being used commercially with popular methods being based on collaborative filtering. While collaborative filtering-based recommenders perform well in terms of accuracy, they lack the ability of finding fresh and novel items, due to the nature of its inner workings. We propose a new graph-based recommender system that uses only positively rated items in users' profiles to construct a highly-connected, undirected graph, with items as nodes and positive correlations as edges. Using the concept of entropy and the linked items in the graph, the proposed system can find recommendations that are both novel and relevant. We test the system on Last.fm data to recommend music to users and show that the proposed recommender system is indeed able to provide novel recommendations while keeping them relevant to the user profile, consistently outperforming a state-of-the-art matrix factorization-based recommender.

With the increasing amount of new motion-sensing hardware, improving the usability of gesture interfaces is becoming more important. However, people don't accept gesture interfaces as comfortably as traditional interfaces because they lack the tactile or haptic feedback of traditional input methods, such as mice or keyboards. However, having tactile feedback in gesture interfaces isn't possible unless users are wearing a device with actuators. Therefore, auditory feedback is an appropriate and unique alternative for assisting visual feedback in gesture interfaces. The authors propose various types of novel auditory feedback methods and explore their effects as secondary feedback on complementing visual feedback. They performed a user study for a menu selection task, and experimental results show that the proposed auditory feedback is significantly more efficient and effective than visual-only feedback or conventional auditory feedback in terms of time and accuracy.

With the increasing popularity of touch screen mobile devices, improving the usability and the user experience while inputting text on these devices is becoming increasingly important. Most conventional touch screen keyboards on mobile devices rely heavily on visual feedback, while auditory feedback seldom includes any useful information about what is being inputted by the user. The auditory feedback usually simply replicates the sounds produced by a physical keyboard. This paper describes the development of an enhanced auditory feedback mechanism for a Korean touch screen keyboard called the enhanced auditory feedback (EAF) mechanism. EAF has subtle phonetic auditory feedback generated using the acoustic phonetic features of human speech. While typing with EAF, users can acquire non-invasive auditory clues about the keys pressed. In this work, we compare conventional auditory feedback for a touch screen keyboard used in touch screen mobile devices with that of EAF and explore the possibility of using enhanced auditory feedback for touch screen keyboards.

Background Polysomnography (PSG) is the gold standard test for obstructive sleep apnea (OSA), but it incurs high costs, requires inconvenient measurements, and is limited by a one-night test. Thus, a repetitive OSA screening test using affordable data would be effective both for patients interested in their own OSA risk and in-hospital PSG. The purpose of this research was to develop a four-OSA severity classification model using a patient’s breathing sounds. Methods Breathing sounds were recorded from 83 subjects during a PSG test. There was no exclusive experimental protocol or additional recording instruments use throughout the sound recording procedure. Based on the Apnea-Hypopnea Index (AHI), which indicates the severity of sleep apnea, the subjects’ sound data were divided into four-OSA severity classes. From the individual sound data, we proposed two novel methods which were not attempted in previous OSA severity classification studies. First, the total transition probability of approximated sound energy in time series, and second, the statistical properties derived from the dimension-reduced cyclic spectral density. In addition, feature selection was conducted to achieve better results with a more relevant subset of features. Then, the classification model was trained using support vector machines and evaluated using leave-one-out cross-validation. Results The overall results show that our classification model is better than existing multiple OSA severity classification method using breathing sounds. The proposed method demonstrated 79.52% accuracy for the four-class classification task. Additionally, it demonstrated 98.0% sensitivity, 75.0% specificity, and 92.78% accuracy for OSA subject detection classification with AHI threshold 5. Conclusions The results show that our proposed method can be used as part of an OSA screening test, which can provide the subject with detailed OSA severity results from only breathing sounds.

In this paper, we present an extension of robust principal component analysis (RPCA) with weighted \(l_{1}\)-norm minimization for singing voice separation. While the conventional RPCA applies a uniform weight between the low-rank and sparse matrices, we use different weighting parameters for each frequency bin in a spectrogram by estimating the variance ratio between the singing voice and accompaniment. In addition, we incorporate the results of vocal activation detection into the formation of the weighting matrix, and use it in the final decomposition framework. From the experimental results using the DSD100 dataset, we found that proposed algorithm yields a meaningful improvement in the separation performance compared to the conventional RPCA.

Recommending appropriate music to users has always been a difficult task. In this paper, we propose a novel method in recommending music by analyzing the textual input of users. To this end, we mine a large corpus of documents from a Korean radio station's online bulletin board. Each document, written by the listener, is composed of a song request associated with a brief, personal story. We assume that such stories are closely related with the background of the song requests and thus, our system performs text analysis to recommend songs that were requested from other similar stories. We evaluate our system using conventional metrics along with a user evaluation test. Results show that there is close correlation between document similarity and song similarity, indicating the potential of using text as a source to recommending music.

Summary ObjectivesWe aimed to determine the characteristics of quantitative metrics for nursing narratives documented in electronic nursing records and their association with hospital admission traits and diagnoses in a large data set not limited to specific patient events or hypotheses. MethodsWe collected 135,406,873 electronic, structured coded nursing narratives from 231,494 hospital admissions of patients discharged between 2008 and 2012 at a tertiary teaching institution that routinely uses an electronic health records system. The standardized number of nursing narratives (i.e., the total number of nursing narratives divided by the length of the hospital stay) was suggested to integrate the frequency and quantity of nursing documentation. ResultsThe standardized number of nursing narratives was higher for patients aged ≥ 70 years (median = 30.2 narratives/day, interquartile range [IQR] = 24.0–39.4 narratives/day), long (≥ 8 days) hospital stays (median = 34.6 narratives/day, IQR = 27.2–43.5 narratives/day), and hospital deaths (median = 59.1 narratives/day, IQR = 47.0–74.8 narratives/day). The standardized number of narratives was higher in “pregnancy, childbirth, and puerperium” (median = 46.5, IQR = 39.0–54.7) and “diseases of the circulatory system” admissions (median = 35.7, IQR = 29.0–43.4). ConclusionsDiverse hospital admissions can be consistently described with nursing-documentderived metrics for similar hospital admissions and diagnoses. Some areas of hospital admissions may have consistently increasing volumes of nursing documentation across years. Usability of electronic nursing document metrics for evaluating healthcare requires multiple aspects of hospital admissions to be considered. Citation: Kim K, Jeong S, Lee K, Park H-A, Min YH, Lee JY, Kim Y, Yoo S, Doh G, Ahn S. Metrics for electronicnursing-record-based narratives: cross-sectional analysis.

Pulsatile tinnitus (PT) is often an initial presenting symptom of dural arteriovenous fistula (dAVF), but it may be overlooked or diagnosed late if not suspected on initial diagnostic work-up. Here, we assess anatomical features, treatment outcomes, and clinical implications of patients with PT due to dAVF. Of 220 patients who were diagnosed with dAVF between 2003 and 2014, 30 (13.6%) presented with only PT as their initial symptom. The transverse-sigmoid sinus (70.0%) was the most common site, followed by the hypoglossal canal (10.0%) and the middle cranial fossa (6.7%) on radiologic evaluation. Regarding venous drainage patterns, sinus or meningeal venous drainage pattern was the most common type (73.3%), followed by sinus drainage with a cortical venous reflux (26.7%). PT disappeared completely in 21 (80.8%) of 26 patients who underwent therapeutic intervention with transarterial embolization of the fistula, improved markedly in 3 (11.5%), and remained the same in 2 (7.7%). In conclusion, considering that PT may be the only initial symptom in more than 10% of dAVF, not only otolaryngologists but also neurologists and neurosurgeons should meticulously evaluate patients with PT. In most cases, PT originating from dAVF can be cured with transarterial embolization regardless of location and venous drainage pattern.

Feature learning for music applications has recently received considerable attention from many researchers. This paper reports on the sparse feature learning algorithm for musical instrument identification, and in particular, focuses on the effects of the frame sampling techniques for dictionary learning and the pooling methods for feature aggregation. To this end, two frame sampling techniques are examined that are fixed and proportional random sampling. Furthermore, the effect of using onset frame was analyzed for both of proposed sampling methods. Regarding summarization of the feature activation, a standard deviation pooling method is used and compared with the commonly used max- and average-pooling techniques. Using more than 47 000 recordings of 24 instruments from various performers, playing styles, and dynamics, a number of tuning parameters are experimented including the analysis frame size, the dictionary size, and the type of frequency scaling as well as the different sampling and pooling methods. The results show that the combination of proportional sampling and standard deviation pooling achieve the best overall performance of 95.62% while the optimal parameter set varies among the instrument classes.

Objective Although vascular pulsatile tinnitus (VPT) has been classified as “objective”, VPT is not easily recognizable or documentable in most cases. In response to this, we have developed transcanal sound recording (TSR) and spectro-temporal analysis (STA) for the objective diagnosis of VPT. By refining our initial method, we were able to apply TSR/STA to post-treatment outcome evaluation, as well as pre-treatment objective diagnosis. Methods TSR was performed on seven VPT patients and five normal controls before and after surgical or interventional treatment. VPT was recorded using an inserted microphone with the subjects placed in both upright and supine positions with 1) a neutral head position, 2) head rotated to the tinnitus side, 3) head rotated to the non-tinnitus side, and 4) a neutral position with ipsi-lesional manual cervical compression. The recorded signals were analyzed in both time and time-frequency domains by performing a short-time Fourier transformation. Results The pre-treatment ear canal signals of all VPT patients demonstrated pulse-synchronous periodic structures and acoustic characteristics that were representative of their presumptive vascular pathologies, whereas those the controls exhibited smaller peaks and weak periodicities. Compared with the pre-treatment signals, the post-treatment signals exhibited significantly reduced peak- and root mean square amplitudes upon time domain analysis. Additionally, further sub-band analysis confirmed that the pulse-synchronous signal of all subjects was not identifiable after treatment and, in particular, that the signal decrement was statistically significant at low frequencies. Moreover, the post-treatment signals of the VPT subjects revealed no significant differences when compared to those of the control group. Conclusion We reconfirmed that the TSR/STA method is an effective modality to objectify VPT. In addition, the potential role of the TSR/STA method in the objective evaluation of treatment outcomes in patients with VPT was proven. Further studies incorporating a larger sample size and more refined recording techniques are warranted.

In this paper, we propose a novelmethod to search for precise locations of paired note onset and offset in a singing voice signal. In comparison with the existing onset detection algorithms,our approach differs in two key respects. First, we employ Correntropy, a generalized correlation function inspired from Reyni's entropy, as a detection function to capture the instantaneous flux while preserving insensitiveness to outliers. Next, a novel peak picking algorithm is specially designed for this detection function. By calculating the fitness of a pre-defined inverse hyperbolic kernel to a detection function, it is possible to find an onset and its corresponding offset simultaneously. Experimental results show that the proposed method achieves performance significantly better than or comparable to other state-of-the-art techniques for onset detection in singing voice., Comment: 2014 IEEE International Conference on Acoustics, Speech and Signal Processing, 5 pages, 5 figures

We consider a problem of retrieving the extreme value among sensed data under deadline constraints in wireless sensor networks with potential applications to alarm systems. The sensed data is mapped to a score which we adopt as a unified measure of the relative urgency of the data. The objective is to retrieve the data with the maximum score. We propose fully distributed schemes for contention based medium access and data combining. The proposed medium access scheme uses a randomized back-off which is controlled based on the score of the data to be transmitted. Data combining techniques are proposed to further suppress unnecessary traffic and reduce contention. The key observation is that one should aggressively prioritize packets with high score, up to an extent that does not incur excessive contention in channel access. Designed to capture such aspect, the proposed scheme is shown to substantially decrease the latency of the retrieval.

We describe an acoustic chord transcription system that uses symbolic data to train hidden Markov models and gives best-of-class frame-level recognition results. We avoid the extremely laborious task of human annotation of chord names and boundaries-which must be done to provide machine learning models with ground truth-by performing automatic harmony analysis on symbolic music files. In parallel, we synthesize audio from the same symbolic files and extract acoustic feature vectors which are in perfect alignment with the labels. We, therefore, generate a large set of labeled training data with a minimal amount of human labor. This allows for richer models. Thus, we build 24 key-dependent HMMs, one for each key, using the key information derived from symbolic data. Each key model defines a unique state-transition characteristic and helps avoid confusions seen in the observation vector. Given acoustic input, we identify a musical key by choosing a key model with the maximum likelihood, and we obtain the chord sequence from the optimal state path of the corresponding key model, both of which are returned by a Viterbi decoder. This not only increases the chord recognition accuracy, but also gives key information. Experimental results show the models trained on synthesized data perform very well on real recordings, even though the labels automatically generated from symbolic data are not 100% accurate. We also demonstrate the robustness of the tonal centroid feature, which outperforms the conventional chroma feature.

Separating vocal and accompaniment signals from a monaural music signal is a challenging task. Recently, robust principal component analysis (RPCA) has been proposed for use in the magnitude spectrogram domain to separate the low-rank and sparse residual matrices, which are assumed to represent the accompaniment and vocal signals, respectively. In this paper, we propose two extended methods based on the RPCA algorithm for more effective vocal separation. First, we extend the conventional RPCA and propose to use in the spectrogram decomposition framework Schatten p- and lp-norms, which are generalized versions of the nuclear norm and l1-norm used in RPCA, respectively. Second, we apply proper scale compression to the magnitude spectrogram, making it a more appropriate representation for the decomposition. Experiments using the MIR-1K dataset show that the proposed methods yield significantly better separation performance than the conventional RPCA.

In recent years, there has been a lot of work in transcribing polyphonic music using non-negative spectrogram factorization. However, most of them focus on transcribing audio signal into the occurrence of notes, onset and pitch of notes. In this paper, a concept for automatic transcription of frequency modulated musical expressions such as vibrato, glissando is proposed. To transcribe those musical expressions from polyphonic music signal, hidden Markov model constrained shift-invariant probabilistic latent component analysis is used. From a impulse distribution which reveals the frequency variation of each note, each expression can be modelled in accordance with designed rules. Experiments showed that the impulse distribution can be used to transcribe expressions from polyphonic music signals.

Recommending music that satisfies the user's taste has been a challenging problem. Previous works on music recommendation system focused on the user's purchase history or the content of the music. In this paper, we propose a music recommendation system purely based on analyzing textual input of the users. We first mine a large corpus of Korean radio episodes, which is written by the listener. Each episode is composed of a personal story and a song request which we assume to be somehow related to the story. We then performing probabilistic Latent Semantic Analysis (pLSA) to find similar documents and recommend music that are associated to those documents. We evaluate our system by computing the mean reciprocal rank and mean average precision, which are both conventional metrics in evaluating information retrieval systems. The result shows that music similarity and document similarity are closely correlated, and thus it is possible to recommend music purely based on text analysis.

Recently unsupervised learning algorithms have been successfully used to represent data in many of machine recognition tasks. In particular, sparse feature learning algorithms have shown that they can not only discover meaningful structures from raw data but also outperform many hand-engineered features. In this paper, we apply the sparse feature learning approach to acoustic scene classification. We use a sparse restricted Boltzmann machine to capture manyfold local acoustic structures from audio data and represent the data in a high-dimensional sparse feature space given the learned structures. For scene classification, we summarize the local features by pooling over audio scene data. While the feature pooling is typically performed over uniformly divided segments, we suggest a new pooling method, which first detects audio events and then performs pooling only over detected events, considering the irregular occurrence of audio events in acoustic scene data. We evaluate the learned features on the IEEE AASP Challenge development set, comparing them with a baseline model using mel-frequency cepstral coefficients (MFCCs). The results show that learned features outperform MFCCs, event-based pooling achieves higher accuracy than uniform pooling and, furthermore, a combination of the two methods performs even better than either one used alone.

Collaborative filtering, being a popular method for generating recommendations, produces satisfying results for users by providing extremely relevant items. Despite being popular, however, this method is prone to many problems. One of these problems is popularity bias, in which the system becomes skewed towards items that are popular amongst the general user population. These 'obvious' items are, technically, extremely relevant items but fail to be novel. In this paper, we maintain using collaborative filtering methods while still managing to produce novel yet relevant items. This is achieved by utilizing the long-tailed distribution of listening behavior of users, in which their playlists are biased towards a few songs while the rest of the songs, those in the long tail, have relatively low play counts. In addition, we also apply a link analysis method to users and define links between them to create an increasingly fine-grained approach in calculating weights for the recommended items. The proposed recommendation method was available online as a user study in order to measure the relevancy and novelty of the recommended items. Results show that the algorithm manages to include novel recommendations that are still relevant, and shows the potential for a new way of generating novel recommendations.

We describe a system for automatic chord transcription from the raw audio using genre-specific hidden Markov models trained on audio-from-symbolic data. In order to avoid enormous amount of human labor required to manually annotate the chord labels for ground-truth, we use symbolic data such as MIDI files to automate the labeling process. In parallel, we synthesize the same symbolic files to provide the models with the sufficient amount of observation feature vectors along with the automatically generated annotations for training. In doing so, we build different models for various musical genres, whose model parameters reveal characteristics specific to their corresponding genre. The experimental results show that the HMMs trained on synthesized data perform very well on real acoustic recordings. It is also shown that when the correct genre is chosen, simpler, genre-specific model yields performance better than or comparable to that of more complex model that is genre-independent. Furthermore, we also demonstrate the potential application of the proposed model to the genre classification task.

A novel approach for obtaining labeled training data is presented to directly estimate the model parameters in a supervised learning algorithm for automatic chord recognition from the raw audio. To this end, harmonic analysis is first performed on symbolic data to generate label files. In paral-lel, we synthesize audio data from the same symbolic data, which are then provided to a machine learning algorithm along with label files to estimate model parameters. Experimental results show higher performance in frame-level chord recognition than the previous approaches.

In prior work, we examined the possibility of sound generation from colon tissue scan data using vocal synthesis models. In this work, we review key results and present extensions to the prior work. Sonification entails the mapping of data values to sound synthesis parameters such that informative sounds are produced by the chosen sound synthesis model. We review the physical equations and technical highlights of a vocal synthesis model developed by Cook. Next we present the colon tissue scan data gathered, and discuss processing steps applied to the data. Finally, we review preliminary results from a simple sonification map. New findings regarding perceptual distance of vowel sounds are presented.