Toshio Modegi

Association between shape of sclera and myopic retinochoroidal lesions in patients with pathologic myopia.

PURPOSEnThe purpose of the study was to analyze the shape of the sclera determined by swept-source optical coherence tomography (OCT) and to determine the relationship between the shape and the myopic retinochoroidal lesions.nnnMETHODSnWe studied 488 eyes of 272 patients with high myopia (refractive error ≥-8.00 diopters [D] or axial length >26.5 mm) and 43 emmetropic eyes of 43 controls (refractive error ≤±3 D). An image of the sclera was obtained by a swept-source OCT prototype instrument that uses a wavelength sweeping laser centered on 1 μm wavelength with an A-scan repetition rate of 100,000 Hz. The scans were 12 mm radial scans centered on the fovea. Seventy eyes were also examined by three-dimensional magnetic resonance imaging (3D MRI) to obtain the contour of the outer surface of the eyes. The main outcome measures, visibility of the entire sclera layer, scleral thickness, scleral contour, and location of the most protruded point of the globe, were obtained by swept-source OCT and 3D MRI.nnnRESULTSnThe entire thickness of the sclera was observed in 278 of 488 (57.0%) highly myopic eyes, but the outer border was not observed in any of the emmetropic eyes. The mean subfoveal scleral thickness was 227.9 ± 82.0 μm in the highly myopic eyes. The sclera was thickest at 3000 μm nasal to the fovea. The curvatures of the inner scleral surface of highly myopic eyes could be divided into curvatures that sloped toward the optic nerve, those that were symmetrical and centered on the fovea, those that were asymmetrical, and those that were irregular. Patients with irregular curvature were significantly older and had significantly longer axial lengths than those with other curvatures. Myopic fundus lesions were present significantly more frequently in the eyes with irregular curvature. All of the eyes whose scleral curvature sloped toward the optic nerve had nasally distorted shape in the 3D MRI images, and all eyes with temporally dislocated shape had irregular curvature.nnnCONCLUSIONSnIn vivo evaluations of the sclera in highly myopic eyes by swept-source OCT can provide important information on deformations of the sclera and how such deformities are related to myopic fundus lesions.

Quantitative Analyses of High-Resolution 3D MR Images of Highly Myopic Eyes to Determine Their Shapes

PURPOSEnWe analyzed the symmetry and pointedness of the posterior segment of highly myopic eyes.nnnMETHODSnWe studied 234 eyes of 117 patients with bilateral high myopia (refractive error ≤-8.00 diopters [D]) and 40 eyes of 20 patients with emmetropia (refractive error between -1.0 and +1.0 D). Volume renderings of high-resolution magnetic resonance (MR) images were performed to obtain 3D images of the eye. To analyze the symmetry and pointedness of the posterior surface, a software was developed to measure the area and angle of a fan-shaped segment formed by selected points on the MR images.nnnRESULTSnAll of the emmetropic eyes were symmetrical in the horizontal and sagittal planes with no deformity. In highly myopic eyes, the shape was symmetrical in the horizontal plane in 146 eyes (62.4%) and in the sagittal plane in 162 (69.2%). The shape of the posterior pole was pointed (angle of fan-shaped segment <150°) in 45.7% and blunted (angle ≥150°) in 54.3% of highly myopic eyes. The most common shape was symmetrical in the horizontal and sagittal planes, and the posterior surface was blunt. The shape of the two eyes of the same individual was the same in 61 of 117 patients (52.1%). In 56 patients whose two eyes had different shapes, the most frequent pattern was a difference in the pointedness (51.8%).nnnCONCLUSIONSnQuantitative assessments of the shape of eyes were useful in determining the pattern of eye shape deformity specific to pathologic myopia.

Proposals of MIDI coding and its application for audio authoring

We propose applying MIDI technology to coding of physiologic sounds such as heart sounds used in medical diagnosis for constructing medical audio databases. Furthermore, we extend our encoding algorithm in order to apply it to other types of sounds such as bird sounds and human singing. We overview our proposed encoding algorithms, which consist of a real time coding method and a high precision coding method. Then we present several coded examples applied for three typical categories: heart sounds, bird sounds and singing sounds. Finally, as our conclusion and future work, we propose an effective audio authoring application concept utilizing MIDI coding, which will be able to provide for conventional waveform editing, as much flexibility as MIDI editing.

Application of MIDI technique for medical audio signal coding

Proposes applying MIDI technology to the coding of physiological sounds, such as heart sounds or lung sounds, used in medical diagnosis for constructing medical audio databases. We have implemented our proposed method and checked its processing speed, the storage size of the encoded data and the quality of the decoded sound. Although our proposed encoding method cannot reproduce the perfect original wave pattern, we confirmed that diagnosing points in source sounds are not degraded by our proposed method. In this paper, we describe our proposed encoding and decoding algorithms, and we present several specific coding examples of both heart and lung sounds on typical clinical cases.

Small object recognition techniques based on structured template matching for high-resolution satellite images

We are developing infrastructure tools of wide-area monitoring used for such as disaster damaged areas or traffic conditions, using Earth observation satellite images. Especially, we are focusing on developing a small object recognition tool for satellite images, which enables extract automobile patterns in high-resolution satellite images such as QuickBird panchromatic images, for example. Although, resolution of optical sensors installed in the current earth observation satellites has been highly advanced, their pixel resolution is not enough for identifying each small object such as an automobile by the currently available pattern matching techniques. Whereas, the pattern matching calculation load of high-resolution images becomes bigger, it will take tremendous time for searching whole objects included in a slice of satellite images. In order to overcome these problems, we propose a structured template matching technique for recognizing small objects in satellite images, which consists of a micro-template matching, clustered micro-template matching and macro-template matching. In this paper, we describe an abstract of our proposed method and present its experimental results.

Nearly Lossless Audio Watermark Embedding Techniques to be Extracted Contactlessly by Cell Phone

We have proposed a novel audio watermarking technique, which embeds a set of bitstream data by changing two-channel stereo locations of lower frequency components in an embedding target audio signal. This method features nearly lossless embedding, robustness against lossy data compression or analogue conversion, and enables contactless asynchronous detection of embedded watermarks through speaker and microphone devices without the original audio signals. Then, we propose several extended monaural embedding methods, which embed data by changing level balances between two sets of lower frequency components, which are divided by either frequency or temporal dimensions. These methods also support nearly lossless embedding and enable watermark detection from only the left channel signal by a single monaural microphone, and we have confirmed detection capability by a cell phone. In this paper, we describe abstracts of our proposed watermark embedding and extracting algorithms, experimental results of watermark extraction precision.

MIDI encoding method based on variable frame-length analysis and its evaluation of coding precision

The MIDI interface was originally designed for electronic musical instruments but we consider that this music-note based coding concept can be extended to general acoustic signal description. We propose applying MIDI technology to coding of biomedical auscultation sound signals such as heart sounds for medical records and telemedicine. We have tried to extend our encoding algorithm in order to apply it to vocal sounds. Currently, we are trying to improve the coding precision based on variable frame-length frequency analysis considering auditory parameters in humans towards general-purpose audio coding applications. We present our improved high-precision MIDI encoding algorithm for reproducing high-quality vocal sounds, and present an evaluation of its coding precision by an automatic notation experiment of short piano sounds.

Structured description method for general acoustic signals using XML format

The MIDI interface is originally designed for electronic musical instruments but we consider this music-note based coding concept can be extended for general acoustic signal description. We proposed applying the MIDI technology to coding of biomedical auscultation sound signals such as heart sounds for retrieving medical records and performing telemedicine. Then we have tried to extend our encoding algorithm and improve the coding precision based on Generalized Harmonic Analysis in order to apply to vocal sounds. Currently, we can separate vocal sounds included in popular songs and encode given musical sounds into a multiple-channel MIDI format. This can reproduce complete musical signals including singing sounds using a GMstandard MIDI tone generator. In this paper, we present our novel automatic XML transcription method devised on our extended MIDI coding research project: “Structured Description of Acoustic Information,” including two coded examples.

Very low bit-rate audio coding technique using MIDI representation

The MIDI interface is originally designed for electronic musical instr uments but we consider this music-note based coding concept can be extended for general acoustic signal description. We proposed applying the MIDI technology to coding of bio-medical auscultation sound signals such as heart sounds for retrieving medical records and performing telemedicine. Then we have tried to extend our encoding algorithm and improve the coding precision based on Generalized Harmonic Analysis in order to apply to vocal sounds. Currently, we can separate vocal sounds included in popular songs and encode both vocal sounds and background instrumental sounds into separate MIDI channels. Using a GM-standard MIDI tone generator, we can playback this multi-channel MIDI encoded data with vocal singing sounds. In this paper, we present our MIDI encoding algorithm for reproducing general audio signals including vocal sounds, and propose its application for very low bit-rate audio compression.

High-precision MIDI encoding method including decoder control for synthesizing vocal sounds

MIDI interface is originally designed for electronic musical instrument but we consider this music-note based coding concept can be extended for general acoustic signal description. We proposed applying the MIDI technology to coding of bio-medical auscultation sound signals such as heart sounds for medical records and tele-medicine. Then we have tried to extend our encoding algorithm in order to apply to vocal sounds. Currently, we are trying to improve the coding precision and also control a decoder device in order to make it generate half-tone sounds undefined on the standard musical chromatic-scale. In this paper. we present our newly proposed high-precision MIDI encoding method which can control standard GM tone generators and make them produce high-quality vocal sounds.