Which Color Channel Is Better for Diagnosing Retinal Diseases Automatically in Color Fundus Photographs?

Color fundus photographs are the most common type of image used for automatic diagnosis of retinal diseases and abnormalities. As all color photographs, these images contain information about three primary colors, i.e., red, green, and blue, in three separate color channels. This work aims to understand the impact of each channel in the automatic diagnosis of retinal diseases and abnormalities. To this end, the existing works are surveyed extensively to explore which color channel is used most commonly for automatically detecting four leading causes of blindness and one retinal abnormality along with segmenting three retinal landmarks. From this survey, it is clear that all channels together are typically used for neural network-based systems, whereas for non-neural network-based systems, the green channel is most commonly used. However, from the previous works, no conclusion can be drawn regarding the importance of the different channels. Therefore, systematic experiments are conducted to analyse this. A well-known U-shaped deep neural network (U-Net) is used to investigate which color channel is best for segmenting one retinal abnormality and three retinal landmarks.

A study on transvelar coupling for non-nasalized sounds.

Previous studies have found that the velum in speech production may not only serve as a binary switch with on-off states for nasal and non-nasal sounds, but also partially alter the acoustic characteristics of non-nasalized sounds. The present study investigated the unique functions of the velum in the production of non-nasalized sounds by using morphological, mechanical, and acoustical measurements. Magnetic resonance imaging movies obtained from three Japanese speakers were used to measure the behaviors of the velum and dynamic changes in the pseudo-volume of the pharyngeal cavity during utterances of voiced stops and vowels. The measurements revealed no significant enlargements in the supraglottal cavity as subjects uttered voiced stops. It is found that the velum thickness varied across utterances in a way that depended on vowels, but not on consonants. The mechanical and acoustical observations in the study suggested that the velum is actively controlled to augment the voice bars of voiced stops, and nostril-radiated sound is one of the most important sources for voice bars, just as is laryngeal wall vibration. This study also proposed a two-layer diaphragm model that simulates transvelar coupling during the production of non-nasalized speech sounds. The simulation demonstrated that the model accurately represented the basic velar functions involved in speech production.

A visual stethoscope to detect the position of the tracheal tube.

BACKGROUND: Advancing a tracheal tube into the bronchus produces unilateral breath sounds. We created a Visual Stethoscope that allows real-time fast Fourier transformation of the sound signal and 3-dimensional (frequency-amplitude-time) color rendering of the results on a personal computer with simultaneous processing of 2 individual sound signals. The aim of this study was to evaluate whether the Visual Stethoscope can detect bronchial intubation in comparison with auscultation. METHODS: After induction of general anesthesia, the trachea was intubated with a tracheal tube. The distance from the incisors to the carina was measured using a fiberoptic bronchoscope. While the anesthesiologist advanced the tracheal tube from the trachea to the bronchus, another anesthesiologist auscultated breath sounds to detect changes of the breath sounds and/or disappearance of bilateral breath sounds for every 1 cm that the tracheal tube was advanced. Two precordial stethoscopes placed at the left and right sides of the chest were used to record breath sounds simultaneously. Subsequently, at a later date, we randomly entered the recorded breath sounds into the Visual Stethoscope. The same anesthesiologist observed the visualized breath sounds on the personal computer screen processed by the Visual Stethoscope to examine changes of breath sounds and/or disappearance of bilateral breath sound. We compared the decision made based on auscultation with that made based on the results of the visualized breath sounds using the Visual Stethoscope. RESULTS: Thirty patients were enrolled in the study. When irregular breath sounds were auscultated, the tip of the tracheal tube was located at 0.6 +/- 1.2 cm on the bronchial side of the carina. Using the Visual Stethoscope, when there were any changes of the shape of the visualized breath sound, the tube was located at 0.4 +/- 0.8 cm on the tracheal side of the carina (P < 0.01). When unilateral breath sounds were auscultated, the tube was located at 2.6 +/- 1.2 cm on the bronchial side of the carina. The tube was also located at 2.3 +/- 1.0 cm on the bronchial side of the carina when a unilateral shape of visualized breath sounds was obtained using the Visual Stethoscope (not significant). CONCLUSIONS: During advancement of the tracheal tube, alterations of the shape of the visualized breath sounds using the Visual Stethoscope appeared before the changes of the breath sounds were detected by auscultation. Bilateral breath sounds disappeared when the tip of the tracheal tube was advanced beyond the carina in both groups.

[A pilot study to develop a screening system and maltreatment index for elementary schoolchildren focusing on daily habits and dental caries].

OBJECTIVE: In recent years, the number of consultations for maltreatment cases has been increasing in Japan. The characteristics of maltreated children are important factors for their identification. In an earlier paper, we analysed the daily habits related to maltreated children. In the present study, the objectives were to assess the relative weights of each factor and score the daily habits and treatment percentages for dental caries, and also to develop a screening system to find maltreated children. METHODS: The subjects of this study were 57 elementary schoolchildren who were given temporary protection by the child social welfare authorities. The control group comprised 575 elementary schoolchildren. Multiple logistics regression analysis was used to calculate scores for each item derived from partial correlation coefficients. RESULTS: Two versions of a maltreatment index for elementary schoolchildren (MIES) were developed, a daily habits version (H) and daily habits and treatment percentage of dental caries version (HD) for the 1st to 3rd grades (G1-3) and the 4th to 6th grades (G4-6). The MIES for G1-3H was: Do you brush your teeth before bedtime (3 points); Do you know that the first molars are important for chewing (2); Have you used a fluoride gel or mouth rinse (2); Do you have any hobbies except video games (2); Do you know the differences between permanent teeth and milk teeth (1). The MIES for G1-3HD were: Have you used a fluoride gel or mouth rinse (4); Do you brush your teeth before bedtime (3); Do you wash your hands after you play outside (2); Is the number of untreated deciduous teeth larger than the number of treated deciduous teeth (1). The MIES for G4-6H were: Do you wash your hands after you play outside (4); Do you brush your teeth after breakfast (3); Do you brush your teeth before bedtime (3). The G4-6HD were: Do you brush your teeth before bedtime (3); Do you wash your hands after you play outside (2); Is the number of untreated permanent teeth greater than the number of treated permanent teeth (1). The availability (sensitivity + specificity) of MIES HD was higher than MIES H when the original values were substituted into MIES. CONCLUSIONS: A screening system for finding maltreated children for MIES was designed. The precision of MIES could be improved by adding the treatment percentage of dental caries to daily habits.