A New Shapley-Based Feature Selection Method in a Clinical Decision Support System for the Identification of Lung Diseases.

The aim of this study is to propose a new feature selection method based on the class-based contribution of Shapley values. For this purpose, a clinical decision support system was developed to assist doctors in their diagnosis of lung diseases from lung sounds. The developed systems, which are based on the Decision Tree Algorithm (DTA), create a classification for five different cases: healthy and disease (URTI, COPD, Pneumonia, and Bronchiolitis) states. The most important reason for using a Decision Tree Classifier instead of other high-performance classifiers such as CNN and RNN is that the class contributions of Shapley values can be seen with this classifier. The systems developed consist of either a single DTA classifier or five parallel DTA classifiers each of which is optimized to make a binary classification such as healthy vs. others, COPD vs. Others, etc. Feature sets based on Power Spectral Density (PSD), Mel Frequency Cepstral Coefficients (MFCC), and statistical characteristics extracted from lung sound recordings were used in these classifications. The results indicate that employing features selected based on the class-based contribution of Shapley values, along with utilizing an ensemble (parallel) system, leads to improved classification performance compared to performances using either raw features alone or traditional use of Shapley values.

A new imbalance-aware loss function to be used in a deep neural network for colorectal polyp segmentation.

Colorectal cancers may occur in colon region of human body because of late detection of polyps. Therefore, colonoscopists often use colonoscopy device to view the entire colon in their routine practice to remove polyps by excisional biopsy. The aim of this study is to develop a new imbalance-aware loss function, i.e., omni-comprehensive loss, to be used in deep neural networks to overcome both imbalanced dataset and the vanishing gradient problem in identifying the related regions of a polyp. Another reason of developing a new loss function is to be able to produce a more comprehensive one that has evaluation capabilities of region-based, shape-aware, and pixel-wise distribution loss approaches at once. To measure the performance of the new loss function, two scenarios have been conducted. First, an 18-layer residual network as backbone with UNet as the decoder is implemented. Second, a 34-layer residual network as the encoder and a UNet as the decoder is designed. For both scenarios, the results of using popular imbalance-aware losses are compared with those of using our proposed new loss function. During training and 5-fold cross validation steps, multiple publicly available datasets are used. In addition to original data in these datasets, their augmented versions are also created by flipping, scaling, rotating and contrast-limited adaptive histogram equalization operations. As a result, our proposed new custom loss function produced the best performance metrics compared with the popular loss functions.

A software tool for interactive generation, representation, and systematical storage of transfer functions for 3D medical images.

As being a tool that assigns optical parameters, i.e. color, transparency, used in interactive visualization, transfer functions have very important effects on the quality of volume rendered medical images. However, finding accurate transfer functions is a very difficult, tedious, and time consuming task because of the variety of all possibilities. By addressing this problem, a software module, which can be easily plugged into any visualization program, is developed based on the specific expectations of medical experts. Its design includes both a new user interface to ease the interactive generation of the volume rendered medical images and a volumetric histogram based method for initial generation of transfer functions. In addition, a novel file system has been implemented to represent 3D medical images using transfer functions based on the DICOM standard. For evaluation of the system by various medical experts, the software is installed into a DICOM viewer. Based on the feedback obtained from the medical experts, several improvements are made, especially to increase the flexibility of the program. The final version of the implemented system shortens the transfer function design process and is applicable to various application areas.

Combining classical HRV indices with wavelet entropy measures improves to performance in diagnosing congestive heart failure.

In this study, best combination of short-term heart rate variability (HRV) measures are sought for to distinguish 29 patients with congestive heart failure (CHF) from 54 healthy subjects in the control group. In the analysis performed, in addition to the standard HRV measures, wavelet entropy measures are also used. A genetic algorithm is used to select the best ones from among all possible combinations of these measures. A k-nearest neighbor classifier is used to evaluate the performance of the feature combinations in classifying these two groups. The results imply that two combinations of all HRV measures, both of which include wavelet entropy measures, have the highest discrimination power in terms of sensitivity and specificity values.

A study on fuzzy C-means clustering-based systems in automatic spike detection.

In this study, different systems based on the fuzzy C-means (FCM) clustering algorithm are utilized for the detection of epileptic spikes in electroencephalogram (EEG) records. The systems are constructed as either single or two-stages. In contrast to single-stage systems, the two-stage system comprises a pre-classifier stage realized by a neural network. The FCM based two-stage system is also compared to a similar system implemented using the K-means clustering algorithm. The results imply that an FCM based two-stage system should be preferred as the spike detection system.

Computational model of an adaptive rhythm generator within the olivocerebellar system.

The aim of this study is to investigate via computation whether the olivocerebellar system is capable of functioning as an adaptive rhythm generator. For this purpose, a detailed and physiologically realistic computational model of the olivocerebellar system is developed, based on the known intrinsic cell and network topological properties of this brain system. The present network, where individual cells are modelled by leaky integrate-and-fire units, converts the irregular spikes produced by the olivary cells into a precise rhythmic signal at the output. The simulation results reveal that the computational model, which normally does not exhibit any rhythmic activity, could be switched into a new mode in which it functions as a rhythm generator producing pulses within three different frequency ranges corresponding to alpha, beta, or gamma bands, respectively. In either mode of operation, the firing rates of all simulated cell types are observed to match real data. The results of this study therefore support the experimental findings of researchers who argue that a biological clock producing rhythmic pulses within different temporal ranges is located within the cerebellum.

Automatic detection of epileptiform events in EEG by a three-stage procedure based on artificial neural networks.

This paper introduces a three-stage procedure based on artificial neural networks for the automatic detection of epileptiform events (EVs) in a multichannel electroencephalogram (EEG) signal. In the first stage, two discrete perceptrons fed by six features are used to classify EEG peaks into three subgroups: 1) definite epileptiform transients (ETs); 2) definite non-ETs; and 3) possible ETs and possible non-ETs. The pre-classification done in the first stage not only reduces the computation time but also increases the overall detection performance of the procedure. In the second stage, the peaks falling into the third group are aimed to be separated from each other by a nonlinear artificial neural network that would function as a postclassifier whose input is a vector of 41 consecutive sample values obtained from each peak. Different networks, i.e., a backpropagation multilayer perceptron and two radial basis function networks trained by a hybrid method and a support vector method, respectively, are constructed as the postclassifier and then compared in terms of their classification performances. In the third stage, multichannel information is integrated into the system for contributing to the process of identifying an EV by the electroencephalographers (EEGers). After the integration of multichannel information, the overall performance of the system is determined with respect to EVs. Visual evaluation, by two EEGers, of 19 channel EEG records of 10 epileptic patients showed that the best performance is obtained with a radial basis support vector machine providing an average sensitivity of 89.1%, an average selectivity of 85.9%, and a false detection rate (per hour) of 7.5.

A simple and low-cost Internet-based teleconsultation system that could effectively solve the health care access problems in underserved areas of developing countries.

In many developing countries including Turkey, telemedicine systems are not in wide use due to the high cost and complexity of the required technology. Lack of these systems however has serious implications on patients who live in rural areas. The objective of this paper is to present a simple and economically affordable alternative to the current systems that would allow experts to easily access the medical data of their remote patients over the Internet. The system is developed in client-server architecture with a user-friendly graphical interface and various services are implemented as dynamic web pages based on PHP. The other key features of the system are its powerful security features and platform independency. An academic prototype is implemented and presented to the evaluation of a group of physicians. The results reveal that the system could find acceptance from the medical community and it could be an effective means of providing quality health care in developing countries.