Adaptive environment classification system for hearing aids.

An adaptive sound classification framework is proposed for hearing aid applications. The long-term goal is to develop fully trainable instruments in which both the acoustical environments encountered in daily life and the hearing aid settings preferred by the user in each environmental class could be learned. Two adaptive classifiers are described, one based on minimum distance clustering and one on Bayesian classification. Through unsupervised learning, the adaptive systems allow classes to split or merge based on changes in the ongoing acoustical environments. Performance was evaluated using real-world sounds from a wide range of acoustical environments. The systems were first initialized using two classes, speech and noise, followed by a testing period when a third class, music, was introduced. Both systems were successful in detecting the presence of an additional class and estimating its underlying parameters, reaching a testing accuracy close to the target rates obtained from best-case scenarios derived from non-adaptive supervised versions of the classifiers (about 3% lower performance). The adaptive Bayesian classifier resulted in a 4% higher overall accuracy upon splitting adaptation than the minimum distance classifier. Merging accuracy was found to be the same in the two systems and within 1%-2% of the best-case supervised versions.

GHQ-28 and cGHQ-28: implications of two scoring methods for the GHQ in a primary care setting.

BACKGROUND: The General Health Questionnaire allows for distinguishing patients suffering from psychiatric problems from those in good mental health. There is some controversy as to the best scoring method. Two different methods have been proposed: GHQ and cGHQ. The present study aims at verifying whether there is confirmation of the hypothesis that the population resulting from the cGHQ calculation presents greater chronicity. METHOD: A total of 40 general practitioners and 1145 of their patients in the Greater Montreal area, Québec, Canada participated in the original study. A sub-group of 313 patients were followed over a 3-month period. The patients were asked to complete a set of questionnaires, including the GHQ-28. The traditional method of calculating the GHQ score and the alternative cGHQ method were used. RESULTS: Two discordant groups of patients were identified: Group 1 (acute cases only), positive on the GHQ, but negative on the cGHQ (n = 124), and Group 2 (chronic cases only), negative on the GHQ and positive on the cGHQ (n = 64). Patients in Group 2 were older (p<0.001), were more worried about their health (p = 0.06) and their health problem had lasted for a longer time (p = 0.02). Patients in Group 2 were twice as likely to remain positive on the GHQ and/or the cGHQ as those in Group 1 in the 1- and 3-month follow-ups. CONCLUSIONS: The results suggest the presence of notable differences between the two groups, generally explainable by the hypothesis of greater chronicity in the cGHQ patients. The results confirm the impact of the choice of calculation method on the composition of the two groups. Finally, the authors suggest using the two calculation methods simultaneously.