Abstract
Introduction
During positive airway pressure (PAP) therapy for sleep apnea syndromes, the machine detected apnea hypopnea index (AHI) is an important method for clinicians to evaluate the ...beneficial effects of PAP. There are concerns about the accuracy of this detection, which also confounds a related question-how common and severe are residual events on PAP. Our study aimed for estimating the long term accuracy of machine detected AHI and the predictors.
Methods
Subjects with OSA who underwent a split night polysomnography were recruited prospectively. Those treated with PAP and tracked by the EncoreAnywhereTM system were analyzed. The ones who stopped PAP within one month were excluded for this analysis. Compliance, therapy data and waveform data were analyzed. Machine detected versus manually scored events were compared at the 1st, 3rd, 6th and 12th month from PAP initiation, and logistic regression was done to explore the factors associated with a high AHI difference.
Results
One hundred and seventy-two patients with mean age of 58.79 ± 13.80 and 63.4% male were included. The differences between the machine detected AHI and manual scored AHI was 10.72 ±8.43 in the first month and were stable for up to 12 months. Male sex, large leak ≥ 1.5% of the whole night, titration arousal index ≥ 15 times/hour, and higher ratio of unstable breathing were associated with AHI difference ≥ 5 times/hour.
Conclusion
The limited agreement between machine detected AHI in the tracking system and manually scored AHI persists for up to 12 months. Gender, large leak, the amount of unstable breathing on PAP, and arousal index during the titration were factors associated with this inaccuracy.
Support (if any)
positive airway pressure, apnea hypopnea index; detection accuracy
Background: The objective of this study was to assess the quality and accuracy of visual backgrounds published in academic surgical journals. Visual backgrounds are commonly used to disseminate ...medical research findings. They distill the key messages of a research article, presenting them graphically in an engaging manner so that potential readers can decide whether to read the complete manuscript. Methods: We developed a Visual Background Assessment Tool based upon published guidelines. Seven reviewers underwent iterative training to apply the tool. We collected visual backgrounds published by 25 surgical journals from January 2017 to April 2021; those corresponding to systematic reviews without meta-analysis, conference backgrounds, narrative reviews, video backgrounds, or nonclinical research were excluded. Included visual backgrounds were scored on accuracy (as compared with written backgrounds) and design and were given a first impression score. Results: Across 25 surgical journals, 1325 visual backgrounds were scored. We found accuracy deficits in the reporting of study design (35.8%), appropriate icon use (49%), and sample size reporting (69.2%) as well as design deficits in element alignment (54.8%) and symmetry (36.1%). Overall scores ranged from 9 to 14 (out of 15), accuracy scores ranged from 4 to 8 (out of 8), and design scores ranged from 3 to 7 (out of 7). No predictors of visual background score were identified. Conclusion: Visual backgrounds vary widely in quality. As visual backgrounds become integrated with the traditional components of scientific publication, they must be held to similarly high standards. We propose a checklist to be used by authors and journals to standardize the quality of visual backgrounds.
Introduction Snoring is an indicator of obstructive sleep apnea (OSA), which contributes to cardiovascular disease and mortality. To better study snoring, audio-based snore detection methods using ...different feature representations have been proposed. However, there is a gap in (1) baseline comparisons of different deep learning features, and (2) analysis of the robustness of snore detection in the presence of subject variation. Through an ablation study, we quantified the effect of features. As a measure of robustness to subject variation, we employed a leave-one-subject-out scheme. Methods We used 1D raw signals or 2D Mel-frequency-cepstrum-coefficients (MFCC) of the signals as inputs to fully connected, convolutional, long-short-term-memory (LSTM) cell-based recurrent, very deep networks (VGG) or combinations of them. The classifiers were support-vector-machines (SVM) or neural networks. The ablation study consists of seven modular combinations of the elements mentioned above. For training, we used 81,207 snore and non-snore 5s- segments from the snore channel of polysomnography (PSG) data obtained from 19 subjects. A leave-one-subject-out scheme, in which each subject is tested using the training data from other subjects, is used to simulate subject variation. We then measure the variation in performance (F1-score) over different subjects using the standard deviation (SD). Results Features learned from 2D convolutional, LSTM, and very deep network (VGG) significantly improve the classification accuracy and robustness of snore detection. Applying these findings, we developed a 2D convolutional LSTM network model that combines spectral and temporal features, resulting in the highest accuracy (mean F1-score = 0.8812) and the second-best robustness. Very deep convolutional networks (VGG-SVM) has the most robust performance (SD of F1-score = 0.0568). Conclusion We provide a baseline comparison to understand the effect of feature representation on snore classification. Besides accuracy, we introduce robustness as another performance metric. Methods with the best accuracy do not necessarily give the best robustness. Features extracted from 2D-convolutional and LSTM network results in the best accuracy, but those from very deep convolutional networks (VGG) have the best robustness. Support (If Any) Supported by Philips Respironics.
We investigate a specific diversity phase for phase diversity (PD) phase retrieval, which possesses higher accuracy than common PD, especially for large-scale and high-frequency wavefront sensing. ...The commonly used PD algorithm employs the image intensities of the focused plane and one defocused plane to build the error metric. Unlike the commonly used PD, we explore a bisymmetric defocuses diversity phase, which employs the image intensities of two symmetrical defocused planes to build the error metric. This kind of diversity phase, named PD-BD (bisymmetric defocuses phase diversity), is analysed with the Cramer-Rao lower bound (CRLB). Statistically, PD-BD shows smaller CRLBs than the commonly used PD, which indicates stronger capacity of phase retrieval. Numerical simulations also verify that PD-BD has higher accuracy of phase retrieval than the commonly used PD when dealing with large-scale and high-frequency wavefront aberrations. To further affirm that PD-BD possesses higher accuracy of wavefront sensing than PD, we also perform a simple verification experiment.
BackgroundThe Clinch Token Transfer Test (C3t) and Step & Stroop Test (SST) are newly developed dual task assessments for Huntington’s Disease (HD). As these are formed of a number of items they ...produce numerous variables and it is unknown which of these have greatest discriminative ability for each assessment.AimsTo use machine learning classifiers to assess the discriminative ability of the tests and to determine which individual aspects are most important to retain.MethodsControls (n=27) and manifest HD (n=36) participants performed the C3t and SST. The C3t records 30 variables including time to complete each item, number of errors made and task costs (the difference in performance between increasingly complex tasks). The SST records 20 variables including number of steps and Stroop accuracy. To determine the discriminatory power of the assessments two classifiers were constructed, one using variables from the C3t and another using variables from the SST. A feature selection algorithm was used to determine which assessment variables were most important to retain.ResultsThe best performing SST classifier had a mean accuracy of 84% using the number of steps in the baseline Step task, the number of correct answers in the Stroop Congruent Baseline and the number of correct answers in the Stroop Incongruent Baseline. The best performing C3t classifier had a mean accuracy of 88% using the time taken for the C3t Dual Task, the time taken for the C3t Triple Task and the task cost (time) between Baseline and Dual Tasks.ConclusionsThis study suggests that the C3t and SST are reasonably suitable for distinguishing between manifest HD and controls. Furthermore, the assessment complexity can be reduced as the optimal models required a fraction of the scores recorded. Future work will seek to a) reduce the complexity of the tests and b) explore potential test enhancements that may bolster classifier performance.
In recent years, significant advancements have been made in pose estimation methods. These methods can be broadly divided into two categories: device-based and device-free. Device-based methods, such ...as virtual reality data gloves and marker-based motion capture, are known for their accuracy. However, these require specific equipment, making them less accessible for the public. On the other hand, device-free methods need no device to recognize human movement. They usually use cameras and estimation algorithms to recognize the body parts. Owing to evolving artificial intelligence (AI) technology, estimation accuracy has increased and we adopt one of the device-free methods to develop two interactive games, “Brain Wall” and “Touch de Pose”. “Brain Wall” challenges players to imitate a silhouette, scoring them based on the accuracy of their pose compared to the silhouette using the Intersection over Union metric. The game encourages competitiveness and participation through a leader board system. “Touch de Pose” allows players to choose a pose theme and the players are required to position specific body parts within the displayed circles on the screen. “Touch de Pose” also includes real-time evaluation of the player’s pose and combines their image with themed background images generated by generative AI, some of which demonstrate the limitations of the technology. The games were showcased at a university open campus event, receiving overwhelmingly positive feedback: 95.6% satisfaction for “Brain Wall” and 98.1% for “Touch de Pose”. Our study shows that pose estimation-based interactive games show immense potential in generating interest in technology for those who are unfamiliar with information and communication technology. Additionally, the full-body engagement aspect of these games could also play a role in promoting physical activity as a regular habit.
Category:
Ankle; Basic Sciences/Biologics; Sports; Trauma
Introduction/Purpose:
Diagnosis of subtle instability of the distal tibiofibular syndesmosis is challenging. In surgically treated rotational ...malleolar fractures, instability is typically assessed with the intraoperative Cotton test. However, this test can be unreliable due to its dynamic nature and uncontrolled distraction force. The Tap test is an alternative test where a cortical tap is advanced through the fibula with a progressive, stable, and unidirectional distraction force. The objective of this cadaveric study was to compare the DTFS widening when using the Cotton and Tap tests as diagnostic tools for coronal plane syndesmotic instability.
Methods:
Ten below-knee cadaveric specimens were tested in intact non-stressed, intact stressed, injured non-stressed, and injured stressed conditions, with stressed conditions utilizing both Cotton and Tap tests. In injured conditions, the syndesmotic ligamentous complex was sectioned (anterolateral longitudinal approach). Perfect fluoroscopic Mortise images were acquired for all conditions. For the Tap test, a 2.5 drill bit was used to drill a hole through both distal fibular cortices. A blunt-edged 3.5mm cortical tap was advanced toward the tibia. For the Cotton test, a lateral distraction force was applied to the distal fibula with a towel clamp. Two observers measured Tibiofibular Clear Space (TFCS) 1cm proximal to the ankle joint line. Intra and interobserver reliabilities were assessed by Intraclass Correlation Coefficient (ICC). Syndesmotic TFCS values for all conditions were compared by paired Wilcoxon. Diagnostic performance of the Cotton and Tap tests was assessed (a relative increase of TFCS>2mm). P-values <0.05 were considered significant.
Results:
The intraclass correlation coefficient (ICC) for intraobserver and interobserver reliability was respectively, 0.96 and 0.78.TFCS measurements were similar in intact non-stressed, intact stressed (both Cotton and Tap tests) and injured non- stressed conditions: intact non-stressed, 3.5mm (CI, 3.0 to 3.9mm); intact stressed, 3.6mm (CI, 3.1 to 4.1mm) (Cotton test) and 4.0mm (CI, 3.5 to 4.5mm) (Tap test); injured non-stressed, 3.8mm (CI, 3.3 to 4.3mm). TFCS was significantly increased (p<0.0001) in injured and stressed ankles for both Cotton and Tap tests, with values of respectively, 6.2mm (CI, 5.8 to 6.7mm) and 6.1mm (CI, 5.7 to 6.6mm). The Cotton test had 73.3% sensitivity, 100% specificity, and 86.7% diagnostic accuracy. The Tap test had 70% sensitivity, 90% specificity, and 80% diagnostic accuracy.
Conclusion:
Our cadaveric study compared the Cotton and Tap tests for detection of coronal plane syndesmotic instability. Both tests demonstrated similar increases in TFCS measurements in stressed injured conditions when compared to intact and injured non-stressed conditions. Additionally, both tests demonstrated similar diagnostic accuracy for coronal plane syndesmotic instability, with slight favor for the Cotton test. In our experience, the Cotton test can be unreliable due to the difficulty in applying a steady distraction force while maintaining a perfect Mortise view. We recommend the Tap test as a more stable, controlled, and reproducible intraoperative diagnostic test for coronal plane syndesmotic instability.
The LOF data anomaly detection method has some defects, such as the value of
k
has great influence on the accuracy of detection results, and the selection of
k
value usually adopts trial method, ...which consumes a lot of calculation time. Therefore, this paper proposes an anomaly detection method for LOF data based on sample parameter selection, Tagged according to the sample data set point of normal and abnormal point, the adaptive selection of
k
value and outlier detection, so as to improve the accuracy of data outlier detection and calculation speed, and through the example of meteorological data outliers detection showed that LOF abnormal data points based on sample parameter selection method in the detection accuracy and reliability are improved significantly.