RESUMEN
In this study, we conducted a collaborative study on the classification between silicone oil droplets and protein particles detected using the flow imaging (FI) method toward proposing a standardized classifier/model. We compared four approaches, including a classification filter composed of particle characteristic parameters, principal component analysis, decision tree, and convolutional neural network in the performance of the developed classifier/model. Finally, the points to be considered were summarized for measurement using the FI method, and for establishing the classifier/model using machine learning to differentiate silicone oil droplets and protein particles.
Asunto(s)
Aceites de Silicona , Siliconas , Tamaño de la Partícula , ProteínasRESUMEN
A 47-year-old man was admitted to the emergency room, half a day after having fallen down on his right cheek drunkenly onto a concrete block. Physical examination revealed that the contusion was limited to the right side of his face, only around the cheek, without trauma to the neck, chest or abdomen. But wide ranging tactile crepitus with severe swelling was present on his face and neck due to widely spread emphysema. Computed tomography (CT) scan revealed some fractures of maxillary antrum, facial and cervical emphysema spreading to the lower part of mediastinum. After a conservative treatment, he recovered without any severe systematic complication. It was found that the facial and cervical emphysema and pneumomediastimum completely disappeared on the follow-up CT scan, 18 days after the event.
Asunto(s)
Enfermedades Maxilomandibulares/cirugía , Seno Maxilar/cirugía , Enfisema Mediastínico/cirugía , Fracturas Craneales/cirugía , Humanos , Enfermedades Maxilomandibulares/complicaciones , Enfermedades Maxilomandibulares/diagnóstico por imagen , Masculino , Seno Maxilar/diagnóstico por imagen , Enfisema Mediastínico/diagnóstico por imagen , Persona de Mediana Edad , Fracturas Craneales/complicaciones , Fracturas Craneales/diagnóstico por imagen , Tomografía Computarizada por Rayos X , Resultado del TratamientoRESUMEN
Generally, the characterization of a metal layer formed on a planar substrate has been achieved using scanning electron microscopy and transmission electron microscopy. These techniques provide details of the surface and/or the cross-section of a planar structure with high resolution. However, the evaluation of sphere-like structures is troublesome owing to the necessity to observe a sample from various angles and/or to calculate the yield from many values obtained for many samples, since the conventional methods can observe a sample only from one direction. We have developed a simple evaluation method for a thin metal layer on plastic microbeads based on its light-scattering properties using dark-field microscopy coupled with a spectrometer. The light-scattering intensity of gold-nanoparticle-coated microbeads depends significantly on the gold coverage. We believe that our study is significant because it describes the development and evaluation of the surface coverage of a thin metal layer on a sphere-like microstructure.
RESUMEN
BACKGROUND: An increasing number of old people, and their medical requirements, cannot be managed by their families in their homes, which has been the traditional and prevalent practice in Japan. The number of people with dementia is increasing and behavioral and psychological symptoms of dementia (BPSD) make care difficult. In the present study, we investigated management techniques for BPSD in long-term care facilities in Japan by using the data mining method, which looks at the reported behaviors of care providers. METHODS: First, interviews were conducted with 15 care providers to develop items for a questionnaire. These data were analyzed qualitatively and synthesized with criteria from the professional literature. The resulting self-report questionnaire on techniques used to manage different symptoms of dementia was completed by 275 care providers. We applied the association rule as a data mining method examining 15 management techniques related to 13 BPSD. RESULTS: Analysis identified four types of management techniques: (i) emotional and behavior-concordance techniques; (ii) acceptance and supportive techniques; (iii) restraining techniques; and (iv) avoidance techniques. Different management techniques, and combinations of techniques, were found to vary in use and effectiveness with different BPSD. CONCLUSIONS: Good management techniques for many BPSD have been developed and are being implemented by care providers. The present study has the potential to inform researchers and care providers in Japan about the types of management techniques in current use, as well as areas of potential need for staff training.