Automatic identification of schizophrenia employing EEG records analyzed with deep learning algorithms.

Electroencephalography is a method of detecting and analyzing electrical activity in the brain. This electrical activity can be recorded and processed to aid in the clinical diagnosis of mental disorders. In this study, a novel system for classifying schizophrenia patients from EEG recordings is presented. The developed algorithm decomposes the EEG signals into a system of radial basis functions using the method of fuzzy means. This decomposition helps to obtain the information from the various electrodes of the EEG and allows separating between healthy controls and patients with schizophrenia. The proposed method has been compared with classical machine learning algorithms, such as, K-Nearest Neighbor, Adaboost, Support Vector Machine, and Bayesian Linear Discriminant Analysis. The results show that the proposed method obtains the highest values in terms of balanced accuracy, recall, precision and F1 score, close to 93 % in all cases. The model developed in this study can be implemented in brain activity analysis systems that help in the prediction of patients with schizophrenia.

Early detection of neonatal hearing loss by otoacoustic emissions and auditory brainstem response over 10 years of experience.

OBJECTIVES: A number of different screening protocols for detecting neonatal hearing loss currently exist. We present our 10 years of experience with using auditory brainstem response (ABR) complementary to otoacoustic emissions (OAEs) in the three phases hearing screening process in our hospital. Furthermore, we want to demonstrate the usefulness of these screening techniques used in combination, that remain valid to identify cases of neonatal hearing loss and meet the well-established program quality criteria for these screening protocols. METHODS: Data were collected retrospectively from patient record forms completed on 9698 newborns from 2007 to 2017. The screening protocol for neonatal hearing loss in our centre is carried out in three phases. First phase, prior to discharge from the hospital, consists of carrying out the OAE evaluation on the newborn. Second phase is carried out in the paediatric consultation department. There, the newborns who did not pass the first phase are again studied with OAE. If this phase is not passed either, the child is referred to a third phase for the realization of ABR, in the clinical neurophysiology service. Newborns with risk factors for hearing loss, identified in the first phase, also go on to this third phase. When this hearing threshold exceeds 30 dB, it is considered abnormal. Cases with abnormal ABR, has a re-test conducted within the next six months from the initial ABR assessment. RESULTS: A total of 9390 (97.1%) OAEs were performed during first phase, with 8245 newborns (87.8%) passing the screening test, while 1145 children (12.1%) presented an abnormal OAE and were included in the second screening phase. Second phase involving a repeat OAE examination performed on 1077 newborns (94%). In this second phase, 941 newborns (87.3%) passed the test. Nevertheless, 136 newborns (12.6%) failing the retest and were referred to continue on to phase three. Furthermore, 181 newborns (1.8%) presented high-risk factors at birth and were also included in this third phase. However, in the registries of children referred to this phase, only 255 (80%) ABR evaluations were confirmed. In total, 227 newborns (2.3%) were missed from the first to third phases of the screening process. According to the database of the clinical neurophysiology service, ABRs evaluations were performed in 352 newborns referred between December 2007 and December 2017. Of this sample, 38.9% were boys and 61.1% were girls. From among cases underwent ABR, 34% of newborns did not pass the OAEs. The most common risk factor was prematurity (with admission to the neonatal intensive care unit for more than five days), affecting 28%. Abnormal ABRs waveforms were found in 43.9%, with 12.3% having a sensorineural hearing loss, 26.5% showing mixed hearing loss and, conductive hearing loss being present in 61.9%. Considering sensorineural hearing loss and other types of severe hearing loss, affected patients constituted only 1.7% of the total number of individuals studied. Finally, regarding quality control of the program participation in the first phase of care included 97.2% of all newborns, yielding a third phase referral rate of 2.9%, confirmation of a diagnosis before the fourth month of life in more than 90% of cases with an average of 3.4 months of age, and a hearing impairment detection rate as an outcome indicator of 4.5%. CONCLUSIONS: Our data are similar to those of previous studies on screening for hearing loss in newborns. We have demonstrated the advantages of carrying out this protocol in three phases using the otoacoustic emissions together with auditory brainstem response, diagnostic tools that remain as a Gold Standard. Also, we want to highlight and demonstrate the importance of interdisciplinary coordination between the paediatric and clinical neurophysiology services in the implementation of this screening protocol. The foregoing has allowed us to comply with the proposed quality indicators, reaching coverage percentages of more than 95%, confirming the diagnosis of hearing loss within the first six months of life and making timely referrals to benefit the newborns with hearing impairment by way of treatment and follow-up in the early stages of development, avoiding future disabilities.