Computational Intelligence-Based Cell Nuclei Segmentation from Pap Smear Images

Automated Segmentation of cell nuclei in Pap smear images plays an important role in the cervical cancer cell analysis systems to make a correct diagnosis decision. The aim of this chapter is to detail about the variety of computational intelligence and image processing approaches developed and used for the nuclei segmentation. In additional, the threshold based segmentation problem is treated as an optimization problem with an objective of preserving both the size and volume of the cell nuclei and also to segment the nuclei region from the original microscopic Pap smear image with the help of Particle Swarm Optimization (PSO) and Ant Colony Optimization techniques (ACO). Experimental results are shown, compared in quantitative and qualitative manner as well as the main advantages and limitations of each algorithm are explained.

Mining of Medical Trends Using Social Networks

Recent work in machine learning and natural language processing has studied the content of health related information in tweets and demonstrated the potential for extracting useful public health information from their aggregation. Social intelligence derived from health content has become of significant importance for various applications, including post-marketing drug surveillance, competitive intelligence, medicine reviews and to assess health-related opinions and sentiments. Further, the quantity of medical information in the media such as tweets on Twitter, Facebook or medical blogs is growing at an exponential rate. Medical data such as health records, drug data, etc. has become major candidates for Big Data analysis and thus exploring this content has become a necessity for organizations. However, the volume, velocity, variety, and quality of online health information present challenges, necessitating enhanced facilitation mechanisms for medical social computing. The objective of this chapter is to discuss the possibility of mining medical trends using Social Networks.

A Review on Brain Imaging Techniques for BCI Applications

Evolution has endowed human race with the most adroit brain, and to harness its potential to the fullest the concept of brain computer interface (BCI) has emerged. One of the most crucial components of BCI is the technique of brain imaging. The first approach in the field of brain imaging was to measure the electrical and magnetic activity of the brain, the techniques being known as Electroencephalography and Magnetoencephalography. Striving for furtherance, researchers came up with another alternative known as Magnetic Resonance Imaging. But it being confined to only structural imaging, the functional aspects of brain were mapped using functional magnetic resonance imaging. A similar but comparatively newer neuroimaging modality is Functional Near Infrared Spectroscopy. Transcranial Magnetic Stimulation neuro-physiological technique is based on the principle of electromagnetic induction. Based on nuclear medicine the brain imaging technologies that are widely explored in the world of BCI are Positron Emission Tomography and Single Positron Emission Tomography.

Biomedical Image Processing Overview

Over the years, there are different sorts of medical imaging have been developed. Where the most known are: X-ray, computed tomography (CT), nuclear medicine imaging (PET, SPECT), ultrasound and magnetic resonance imaging (MRI), each one has its different utilities. Various studies in biomedical informatics present a process to analyze images for extracting the hidden information which can be used after that. Image analysis combines several fields that are classified into two categories; the process of low-level, that requires very little information about the content image and the process of high-level, which may need information about the image content. The topic of this chapter is to introduce the different techniques for medical image processing and to present many research studies in this domain. It includes four stages, firstly, we introduce the most important medical imaging modalities and secondly, we outline the main process of biomedical image.

Molecular Dynamics Simulations for Biological Systems

Molecular dynamics simulation is an important tool to capture the dynamicity of biological molecule and the atomistic insights. These insights are helpful to explore biological functions. Molecular dynamics simulation from femto seconds to milli seconds scale give a large ensemble of conformations that can reveal many biological mysteries. The main focus of the chapter is to throw light on theories, requirement of molecular dynamics for biological studies and application of molecular dynamics simulations. Molecular dynamics simulations are widely used to study protein-protein interaction, protein-ligand docking, effects of mutation on interactions, protein folding and flexibility of the biological molecules. This chapter also deals with various methods/algorithms of protein tertiary structure prediction, their strengths and weaknesses.

Medical Image Mining Using Fuzzy Connectedness Image Segmentation

This chapter presents novel approach fuzzy connectedness image segmentation with geometric moments (FCISGM) for digital imaging and communications in medicine (DICOM) image mining. As most of the medical imaging data is exchanged in DICOM format, this chapter focuses on the various methodologies available for DICOM image feature extraction and mining. The comparison of existing medical image mining approaches with the proposed FCISGM approach is provided in this chapter. After carrying out exhaustive results it has been found that proposed FCISGM method gives more precise results and requires minimum number of computations compare to other medical image mining approaches resulting in improved relevant outcomes.

A Simple Non-Invasive Automated Heart Rate Monitoring System Using Facial Images

In this chapter we will discuss a simple non invasive automated heart rate monitoring method. Commonly heart rate is measured by using heart rate monitor devices. Many patients do not feel comfortable when they use contact devices for diagnostic purposes. Our algorithm gives a non-invasive way of heart rate measurement. The first step is to record a video. After 5 frames of the video are captured, the face is detected. A total of 300 frames will be used for further processing. At this stage, ROI (part of forehead) will be cropped out automatically. All image frames are in RGB color model, so these will be separated into 3 channels. For analysis, graph normalization is applied, which uses mean and standard deviation. Fast Fourier transform is used to plot the power spectrum of the traces. This power spectrum will have a peak if the heart rate is detected. We used RGB, HSI, YCbCr, YIQ, and CIE LAB color models for analysis. The best result is achieved with RGB color model followed by CIELab. The average accuracy is 95.32%.

Electromyography-Based Functional Electrical Stimulation (FES) in Rehabilitation

Loss or impairment in the ability of muscle movement or sensation is called Paralysis which is caused by disruption of communication of nerve impulses along the pathway from the brain to the muscles. One of the principal reasons causing paralysis is Spinal Cord Injury (SCI) and Neurological rehabilitation by using neuro-prostheses, based on Functional Electrical Stimulation (FES) is extensively used for its treatment. Impaired muscles are activated by applying small amplitude electrical current. Electromyography (EMG), the recording of biosignals generated by muscle activity during the application of FES can be used as the control signal for FES based rehabilitative devices. This method is predominantly used for restoring upper extremity functioning (wrist, hand, elbow, etc.), standing, walking (speed, pattern) in stroke patients. FES, collaborated with conventional methods, has the potential to be utilized as a useful tool for rehabilitation and restoration of muscle strength, metabolic responses etc. in paralyzed patients.

Prediction and Detection of Epileptic Seizure

Epilepsy is one of the common neurological disorders characterized by a sudden and recurrent malfunction of the brain that is termed “seizure”, affecting around 65 million individuals worldwide. Epileptic seizure may lead to many injuries such as fractures, submersion, burns, motor vehicle accidents and even death. It is highly possible to prevent these unwanted situations if we can predict/detect electrical changes in brain that occur prior to onset of actual seizure. When building a prediction model, the goal should be to make a model that accurately classifies preictal period (prior to a seizure onset) from interictal (period between seizures when non-seizure syndrome is observed) period. On the hand, for the detection we need to make a model that can classify ictal (actual seizure period) from non-ictal/interictal period. This chapter describes the seizure detection and prediction techniques with its background, features, recent developments, and future trends.

Data Mining-Based CBIR System

Multimedia mining primarily involves information analysis and retrieval based on implicit knowledge. The ever increasing digital image databases on the internet has created a need for using multimedia mining on these databases for effective and efficient retrieval of images. Contents of an image can be expressed in different features such as Shape, Texture and Intensity-distribution (STI). Content Based Image Retrieval (CBIR) is the efficient retrieval of relevant images from large databases based on features extracted from the image. The emergence and proliferation of social network sites such as Facebook, Twitter and LinkedIn and other multimedia networks such as Flickr has further accelerated the need of efficient CBIR systems. Analyzing this huge amount of multimedia data to discover useful knowledge is a challenging task. Most of the existing systems either concentrate on a single representation of all features or linear combination of these features. The need of the day is New Image Mining techniques need to be explored and a self-adaptable CBIR system needs to be developed.