An intelligent system based on fuzzy probabilities for medical diagnosis- a study in aphasia diagnosis.

BACKGROUND: Aphasia diagnosis is particularly challenging due to the linguistic uncertainty and vagueness, inconsistencies in the definition of aphasic syndromes, large number of measurements with imprecision, natural diversity and subjectivity in test objects as well as in opinions of experts who diagnose the disease. METHODS: Fuzzy probability is proposed here as the basic framework for handling the uncertainties in medical diagnosis and particularly aphasia diagnosis. To efficiently construct this fuzzy probabilistic mapping, statistical analysis is performed that constructs input membership functions as well as determines an effective set of input features. RESULTS: Considering the high sensitivity of performance measures to different distribution of testing/training sets, a statistical t-test of significance is applied to compare fuzzy approach results with NN results as well as author's earlier work using fuzzy logic. The proposed fuzzy probability estimator approach clearly provides better diagnosis for both classes of data sets. Specifically, for the first and second type of fuzzy probability classifiers, i.e. spontaneous speech and comprehensive model, P-values are 2.24E-08 and 0.0059, respectively, strongly rejecting the null hypothesis. CONCLUSIONS: THE TECHNIQUE IS APPLIED AND COMPARED ON BOTH COMPREHENSIVE AND SPONTANEOUS SPEECH TEST DATA FOR DIAGNOSIS OF FOUR APHASIA TYPES: Anomic, Broca, Global and Wernicke. Statistical analysis confirms that the proposed approach can significantly improve accuracy using fewer Aphasia features.

A hierarchical fuzzy rule-based approach to aphasia diagnosis.

Aphasia diagnosis is a particularly challenging medical diagnostic task due to the linguistic uncertainty and vagueness, inconsistencies in the definition of aphasic syndromes, large number of measurements with imprecision, natural diversity and subjectivity in test objects as well as in opinions of experts who diagnose the disease. To efficiently address this diagnostic process, a hierarchical fuzzy rule-based structure is proposed here that considers the effect of different features of aphasia by statistical analysis in its construction. This approach can be efficient for diagnosis of aphasia and possibly other medical diagnostic applications due to its fuzzy and hierarchical reasoning construction. Initially, the symptoms of the disease which each consists of different features are analyzed statistically. The measured statistical parameters from the training set are then used to define membership functions and the fuzzy rules. The resulting two-layered fuzzy rule-based system is then compared with a back propagating feed-forward neural network for diagnosis of four Aphasia types: Anomic, Broca, Global and Wernicke. In order to reduce the number of required inputs, the technique is applied and compared on both comprehensive and spontaneous speech tests. Statistical t-test analysis confirms that the proposed approach uses fewer Aphasia features while also presenting a significant improvement in terms of accuracy.

The spontaneous electrical activity of neurons in leech ganglia.

Using the newly developed voltage-sensitive dye VF2.1.Cl, we monitored simultaneously the spontaneous electrical activity of ∼80 neurons in a leech ganglion, representing around 20% of the entire neuronal population. Neurons imaged on the ventral surface of the ganglion either fired spikes regularly at a rate of 1-5 Hz or fired sparse spikes irregularly. In contrast, neurons imaged on the dorsal surface, fired spikes in bursts involving several neurons. The overall degree of correlated electrical activity among leech neurons was limited in control conditions but increased in the presence of the neuromodulator serotonin. The spontaneous electrical activity in a leech ganglion is segregated in three main groups: neurons comprising Retzius cells, Anterior Pagoda, and Annulus Erector motoneurons firing almost periodically, a group of neurons firing sparsely and randomly, and a group of neurons firing bursts of spikes of varying durations. These three groups interact and influence each other only weakly.