Data Mining - Concepts and Techniques.

Importance Deep learning is a family of computational methods that allow an algorithm to program itself by learning from a large set of examples that demonstrate the desired behavior, removing the need to specify rules explicitly. Application of these methods to medical imaging requires further assessment and validation. Objective To apply deep learning to create an algorithm for automated detection of diabetic retinopathy and diabetic macular edema in retinal fundus photographs. Design and Setting A specific type of neural network optimized for image classification called a deep convolutional neural network was trained using a retrospective development data set of 128 175 retinal images, which were graded 3 to 7 times for diabetic retinopathy, diabetic macular edema, and image gradability by a panel of 54 US licensed ophthalmologists and ophthalmology senior residents between May and December 2015. The resultant algorithm was validated in January and February 2016 using 2 separate data sets, both graded by at least 7 US board-certified ophthalmologists with high intragrader consistency. Exposure Deep learning–trained algorithm. Main Outcomes and Measures The sensitivity and specificity of the algorithm for detecting referable diabetic retinopathy (RDR), defined as moderate and worse diabetic retinopathy, referable diabetic macular edema, or both, were generated based on the reference standard of the majority decision of the ophthalmologist panel. The algorithm was evaluated at 2 operating points selected from the development set, one selected for high specificity and another for high sensitivity. Results The EyePACS-1 data set consisted of 9963 images from 4997 patients (mean age, 54.4 years; 62.2% women; prevalence of RDR, 683/8878 fully gradable images [7.8%]); the Messidor-2 data set had 1748 images from 874 patients (mean age, 57.6 years; 42.6% women; prevalence of RDR, 254/1745 fully gradable images [14.6%]). For detecting RDR, the algorithm had an area under the receiver operating curve of 0.991 (95% CI, 0.988-0.993) for EyePACS-1 and 0.990 (95% CI, 0.986-0.995) for Messidor-2. Using the first operating cut point with high specificity, for EyePACS-1, the sensitivity was 90.3% (95% CI, 87.5%-92.7%) and the specificity was 98.1% (95% CI, 97.8%-98.5%). For Messidor-2, the sensitivity was 87.0% (95% CI, 81.1%-91.0%) and the specificity was 98.5% (95% CI, 97.7%-99.1%). Using a second operating point with high sensitivity in the development set, for EyePACS-1 the sensitivity was 97.5% and specificity was 93.4% and for Messidor-2 the sensitivity was 96.1% and specificity was 93.9%. Conclusions and Relevance In this evaluation of retinal fundus photographs from adults with diabetes, an algorithm based on deep machine learning had high sensitivity and specificity for detecting referable diabetic retinopathy. Further research is necessary to determine the feasibility of applying this algorithm in the clinical setting and to determine whether use of the algorithm could lead to improved care and outcomes compared with current ophthalmologic assessment.

/pdf/development-and-validation-of-a-deep-learning-algorithm-for-4ql9aa9p1e.pdf

Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs

Heart rate variability (HRV) is a reliable reflection of the many physiological factors modulating the normal rhythm of the heart. In fact, they provide a powerful means of observing the interplay between the sympathetic and parasympathetic nervous systems. It shows that the structure generating the signal is not only simply linear, but also involves nonlinear contributions. Heart rate (HR) is a nonstationary signal; its variation may contain indicators of current disease, or warnings about impending cardiac diseases. The indicators may be present at all times or may occur at random-during certain intervals of the day. It is strenuous and time consuming to study and pinpoint abnormalities in voluminous data collected over several hours. Hence, HR variation analysis (instantaneous HR against time axis) has become a popular noninvasive tool for assessing the activities of the autonomic nervous system. Computer based analytical tools for in-depth study of data over daylong intervals can be very useful in diagnostics. Therefore, the HRV signal parameters, extracted and analyzed using computers, are highly useful in diagnostics. In this paper, we have discussed the various applications of HRV and different linear, frequency domain, wavelet domain, nonlinear techniques used for the analysis of the HRV.

Heart rate variability: a review

Kubios HRV - Heart rate variability analysis software

Deep convolutional neural network for the automated detection and diagnosis of seizure using EEG signals.

The unpredictability of the occurrence of epileptic seizures contributes to the burden of the disease to a major degree. An automatic system that detects seizure onsets would allow patients or the people near them to take appropriate precautions, and could provide more insight into these phenomena, thereby revealing important clinical information. Thus, various methods have been proposed to predict the onset of seizures based on EEG recordings. A seemingly promising approach involves nonlinear features motivated by the higher order spectra (HOS). The goal in this paper is to find the different HOS features for normal, pre-ictal (background) and epileptic EEG signals. This may help in the detection of seizure onset as early as possible with maximal accuracy. In this work, 300 EEG data, each belonging to the three classes, are studied. Our results show that the HOS based measures show unique ranges for the different classes with high confidence level (p = 0.002).

Analysis of epileptic EEG signals using higher order spectra

Application of higher order statistics for atrial arrhythmia classification

Electrocardiogram is the P-QRS-T wave representing the cardiac depolarization and re-polarization, recorded at the body surface. The subtle changes in amplitude and duration of these waves indicate various pathological conditions. It is very difficult to decipher minute changes in the ECG wave by naked eye. Hence a computer aided diagnosis tool to classify various cardiac diseases will assist the doctors in their ECG reading. In this paper, five types of ECG beats (ANSI/AAMI EC57:1998 standard) of MIT-BIH arrhythmia database were automatically classified. Our proposed methodology involves computation of Discrete Cosine Transform (DCT) coefficients from the segmented beats of ECG, which were then subjected for principal component analysis for dimensionality reduction. Then the clinically significant principal components were fed to (i) feed forward neural network, (ii) least square support vector machine with different kernel functions, and (iii) Probabilistic Neural Network (PNN) for automatic classification. We have obtained the highest average sensitivity of 98.69%, specificity of 99.91%, and classification accuracy of 99.52% with the developed knowledge based system. The developed system is clinically ready to deploy for mass screening programs.

Characterization of ECG beats from cardiac arrhythmia using discrete cosine transform in PCA framework

The unpredictability of the occurrence of epileptic seizures contributes to the burden of the disease to a major degree An automatic system that detects seizure onsets would allow patients or the people near them to take appropriate precautions, and could provide more insight into these phenomena, thereby revealing important clinical information Thus, various methods have been proposed to predict the onset of seizures based on EEG recordings A seemingly promising approach involves nonlinear features motivated by the higher order spectra (HOS) The goal in this paper is to find the different HOS features for normal, pre-ictal (background) and epileptic EEG signals This may help in the detection of seizure onset as early as possible with maximal accuracy In this work, 300 EEG data, each belonging to the three classes, are studied Our results show that the HOS based measures show unique ranges for the different classes with high confidence level (p = 0002)

The ECG is a representative signal containing information about the condition of the heart. The shape and size of the P-QRS-T wave, the time intervals between its various peaks, etc., may contain useful information about the nature of disease afflicting the heart. However, the human observer can not directly monitor these subtle details. Besides, since biosignals are highly subjective, the symptoms may appear at random in the time scale. Therefore, the heart rate variability signal parameters, extracted and analysed using computers, are highly useful in diagnostics. This paper deals with the classification of certain diseases using correlation dimension (CD) and detrended fluctuation analysis (DFA).

Choo Min Lim

Papers

Analysis of epileptic EEG signals using higher order spectra

Application of higher order statistics for atrial arrhythmia classification

Characterization of ECG beats from cardiac arrhythmia using discrete cosine transform in PCA framework

Analysis of epileptic EEG signals using higher order spectra

Heart rate variability analysis using correlation dimension and detrended fluctuation analysis