Search

The focus of this research is on automated identification of the quality of human induced pluripotent stem cell (iPSC) colony images. iPS cell technology is a contemporary method by which the patient's cells are reprogrammed back to stem cells and are differentiated to any cell type wanted. iPS cell technology will be used in future to patient specific drug screening, disease modeling, and tissue repairing, for instance. However, there are technical challenges before iPS cell technology can be used in practice and one of them is quality control of growing iPSC colonies which is currently done manually but is unfeasible solution in large-scale cultures. The monitoring problem returns to image analysis and classification problem. In this paper, we tackle this problem using machine learning methods such as multiclass Support Vector Machines and several baseline methods together with Scaled Invariant Feature Transformation based features. We perform over 80 test arrangements and do a thorough parameter value search. The best accuracy (62.4%) for classification was obtained by using a k-NN classifier showing improved accuracy compared to earlier studies.

Inner ear balance problems are common worldwide and are often difficult to diagnose. In this study we examine the classification of patients with inner ear balance problems and controls (people not suffering from inner ear balance problems) based on data derived from the stabilogram signals and using machine learning algorithms. This paper is a continuation for our earlier paper where the same dataset was used and the focus was medically oriented. Our collected dataset consists of stabilogram (a force platform response) data from 30 patients suffering from Ménière's disease and 30 students called controls. We select a wide variety of machine learning algorithms from traditional baseline methods to state-of-the-art methods such as Least-Squares Support Vector Machines and Random Forests. We perform extensive and carefully made parameter value searches and we are able to achieve 88.3% accuracy using k -nearest neighbor classifier. Our results show that machine learning algorithms are well capable of separating patients and controls from each other. [ABSTRACT FROM AUTHOR]

Matching digital fingerprint, face or iris images, biometric verification of persons has advanced. Notwithstanding the progress, this is no easy computational task because of great numbers of complicated data. Since the 1990s, eye movements previously only applied to various tests of medicine and psychology are also studied for the purpose of computer interfaces. Such a short one-dimensional measurement signal contains less data than images and may therefore be simpler and faster to recognize. Using saccadic eye movements we developed a computational verification method to reliably distinguish a legitimate person or a subject in general from others. We tested features extracted from signals recorded from saccade eye movements. We used saccades of 19 healthy subjects and 21 otoneurological patients recorded with electro-oculography and additional 40 healthy subjects recorded with a videocamera system. Verification tests produced high accuracies. [ABSTRACT FROM AUTHOR]