Wavelet-based neural network analysis of ophthalmic artery Doppler signals

Protocol for a systematic review and meta-analysis of the diagnostic accuracy of artificial intelligence for grading of ophthalmology imaging modalities

BACKGROUND

With the rise of artificial intelligence (AI) in ophthalmology, the need to define its diagnostic accuracy is increasingly important. The review aims to elucidate the diagnostic accuracy of AI algorithms in screening for all ophthalmic conditions in patient care settings that involve digital imaging modalities, using the reference standard of human graders.

METHODS

This is a systematic review and meta-analysis. A literature search will be conducted on Ovid MEDLINE, Ovid EMBASE, and Wiley Cochrane CENTRAL from January 1, 2000, to December 20, 2021. Studies will be selected via screening the titles and abstracts, followed by full-text screening. Articles that compare the results of AI-graded ophthalmic images with results from human graders as a reference standard will be included; articles that do not will be excluded. The systematic review software DistillerSR will be used to automate part of the screening process as an adjunct to human reviewers. After the full-text screening, data will be extracted from each study via the categories of study characteristics, patient information, AI methods, intervention, and outcomes. Risk of bias will be scored using Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) by two trained independent reviewers. Disagreements at any step will be addressed by a third adjudicator. The study results will include summary receiver operating characteristic (sROC) curve plots as well as pooled sensitivity and specificity of artificial intelligence for detection of any ophthalmic conditions based on imaging modalities compared to the reference standard. Statistics will be calculated in the R statistical software.

DISCUSSION

This study will provide novel insights into the diagnostic accuracy of AI in new domains of ophthalmology that have not been previously studied. The protocol also outlines the use of an AI-based software to assist in article screening, which may serve as a reference for improving the efficiency and accuracy of future large systematic reviews.

TRIAL REGISTRATION

PROSPERO, CRD42021274441.

Investigation of the relationship between anxiety and heart rate variability in fibromyalgia: A new quantitative approach to evaluate anxiety level in fibromyalgia syndrome

BACKGROUND

Fibromyalgia syndrome (FMS) is identified by widespread musculoskeletal pain, sleep disturbance, nonrestorative sleep, fatigue, morning stiffness and anxiety. Anxiety is very common in Fibromyalgia and generally leads to a misdiagnosis. Self-rated Beck Anxiety Inventory (BAI) and doctor-rated Hamilton Anxiety Inventory (HAM-A) are frequently used by specialists to determine anxiety that accompanies fibromyalgia. However, these semi-quantitative anxiety tests are still subjective as the tests are scored using doctor-rated or self-rated scales.

METHOD

In this study, we investigated the relationship between heart rate variability (HRV) frequency subbands and anxiety tests. The study was conducted with 56 FMS patients and 34 healthy controls. BAI and HAM-A test scores were determined for each participant. ECG signals were then recruited and 71 HRV subbands were obtained from these ECG signals using Wavelet Packet Transform (WPT). The subbands and anxiety tests scores were analyzed and compared using multilayer perceptron neural networks (MLPNN).

RESULTS

The results show that a HRV high frequency (HF) subband in the range of 0.15235Hz to 0.40235Hz, is correlated with BAI scores and another HRV HF subband, frequency range of 0.15235Hz to 0.28907Hz is correlated with HAM-A scores. The overall accuracy is 91.11% for HAM-A and 90% for BAI with MLPNN analysis.

CONCLUSION

Doctor-rated or self-rated anxiety tests should be supported with quantitative and more objective methods. Our results show that the HRV parameters will be able to support the anxiety tests in the clinical evaluation of fibromyalgia. In other words, HRV parameters can potentially be used as an auxiliary diagnostic method in conjunction with anxiety tests.

Wavelet entropy of Doppler ultrasound blood velocity flow waveforms distinguishes nitric oxide-modulated states

Wavelet entropy assesses the degree of order or disorder in signals and presents this complex information in a simple metric. Relative wavelet entropy assesses the similarity between the spectral distributions of two signals, again in a simple metric. Wavelet entropy is therefore potentially a very attractive tool for waveform analysis. The ability of this method to track the effects of pharmacologic modulation of vascular function on Doppler blood velocity waveforms was assessed. Waveforms were captured from ophthalmic arteries of 10 healthy subjects at baseline, after the administration of glyceryl trinitrate (GTN) and after two doses of N(G)-nitro-L-arginine-methyl ester (L-NAME) to produce vasodilation and vasoconstriction, respectively. Wavelet entropy had a tendency to decrease from baseline in response to GTN, but significantly increased after the administration of L-NAME (mean: 1.60 ± 0.07 after 0.25 mg/kg and 1.72 ± 0.13 after 0.5 mg/kg vs. 1.50 ± 0.10 at baseline, p < 0.05). Relative wavelet entropy had a spectral distribution from increasing doses of L-NAME comparable to baseline, 0.07 ± 0.04 and 0.08 ± 0.03, respectively, whereas GTN had the most dissimilar spectral distribution compared with baseline (0.17 ± 0.08, p = 0.002). Wavelet entropy can detect subtle changes in Doppler blood velocity waveform structure in response to nitric-oxide-mediated changes in arteriolar smooth muscle tone.

Classification of sleep apnea through sub-band energy of abdominal effort signal using Wavelets + Neural Networks

Detection and classification of sleep apnea syndrome (SAS) is a critical problem. In this study an efficient method for classification sleep apnea through sub-band energy of abdominal effort using a particularly designed hybrid classifier as Wavelets + Neural Network is proposed. The Abdominal respiration signals were separated into spectral sub-band energy components with multi-resolution Discrete Wavelet Transform (DWT). The energy content of these spectral components was applied to the input of the artificial neural network (ANN). The ANN was configured to give three outputs dedicated to SAS cases; obstructive sleep apnea (OSA), central sleep apnea (CSA) and mixed sleep apnea (MSA). Through the network, satisfactory results that rewarding 85.62% mean accuracy in classifying SAS were obtained.

Combining eigenvector methods and support vector machines for detecting variability of Doppler ultrasound signals

In this paper, the multiclass support vector machines (SVMs) with the error correcting output codes (ECOC) were presented for detecting variabilities of the multiclass Doppler ultrasound signals. The ophthalmic arterial (OA) Doppler signals were recorded from healthy subjects, subjects suffering from OA stenosis, subjects suffering from ocular Behcet disease. The internal carotid arterial (ICA) Doppler signals were recorded from healthy subjects, subjects suffering from ICA stenosis, subjects suffering from ICA occlusion. Methods of combining multiple classifiers with diverse features are viewed as a general problem in various application areas of pattern recognition. Because of the importance of making the right decision, better classification procedures for Doppler ultrasound signals are searched. Decision making was performed in two stages: feature extraction by eigenvector methods and classification using the SVMs trained on the extracted features. The research demonstrated that the multiclass SVMs trained on extracted features achieved high accuracy rates.

Combining neural network models for automated diagnostic systems

This paper illustrates the use of combined neural network (CNN) models to guide model selection for diagnosis of internal carotid arterial (ICA) disorders. The ICA Doppler signals were decomposed into time-frequency representations using discrete wavelet transform and statistical features were calculated to depict their distribution. The first level networks were implemented for the diagnosis of ICA disorders using the statistical features as inputs. To improve diagnostic accuracy, the second level network was trained using the outputs of the first level networks as input data. The CNN models achieved accuracy rates which were higher than that of the stand-alone neural network models.