Short utterance recognition using a network with minimum training

A feedforward network is used to recognize short, digitized, isolated utterances. A high, multispeaker recognition rate is achieved with a small vocabulary with a single training utterance. This approach makes use of the pattern recognition property of the network architecture to classify different temporal patterns in the multidimensional feature space. The network recognizes the utterances without the need of segmentation, phoneme identification, or time alignment. We train the network with four words spoken by one single speaker. The network is then able to recognize 20 tokens spoken by 5 other speakers. We repeat the above training and testing procedure using a different speaker's utterances for training each time. The overall accuracy is 97.5%. We compare this approach to the traditional dynamic programming (DP) approach, and find that DP with slope constraints of 0 and 1 achieve 98.5% and 85% accuracies respectively. Finally we validate out statistics by training and testing the network of a four-word subset of the Texas Instruments (Tl) isolated word database. The accuracy with this vocabulary exceeds 96%. By doubling the size of the training set, the accuracy is raised to 98%. Using a suitable threshold, we are able to raise the accuracy of one network from 87% to 98.5%. Thresholding applied to all networks would then raise the overall accuracy to well over 99%.

This technique is especially promising because of the low overhead and computational requirements, which make it suitable for a low cost, portable, command recognition type of application.     

Role of technetium-99m planar bone scanning in the evaluation of low back pain

The records of 1018 patients with low back pain in a tertiary spine referral practice were reviewed. One hundred thirty-nine out of 1018 (13.6%) underwent technetium-99m planar bone scanning as part of their investigation. Seventy-three out of 139 scans (52%) showed increased uptake in some area, but only 27 out of 139 (19.4%) showed increased uptake specifically in the low back. Scans consistently yielded no findings with reference to the back when the prescan diagnosis was spinal stenosis, lumbar pain syndrome, herniated nucleus pulposus, or postlaminectomy syndrome. Some scans gave positive findings in patients with a diagnosis of degenerative disc disease, pseudarthrosis, spondylolisthesis, fracture, infection, metabolic disorder, or tumor. Positive scans were generally obtained early after presentation (within 3 months) and negative scans obtained later (after 6 months), suggesting that clinical suspicion is still the main indication for early scanning. Planar bone scanning was helpful in both diagnosis and therapeutic decisionmaking in many conditions.     

Migration and chemical modification of silicone in women with breast prostheses

Studies using magnetic resonance spectroscopy in human volunteers to evaluate their livers in vivo and to analyze their blood in vitro suggest that there are measurable amounts of silicon compounds in the blood of some women with implants and that there is migration of silicone to other organs such as the liver.     

Dopamine transporters, D2 receptors, and glucose metabolism in corticobasal degeneration.

Alterations in presynaptic and postsynaptic dopaminergic system and cerebral glucose metabolism in corticobasal degeneration (CBD) were assessed to evaluate the potential usefulness of different imaging methods for CBD. (123)I-FP-CIT/(123)I-beta-CIT SPECT and (123)I-IBZM SPECT as well as (18)F-FDG PET were performed in eight CBD patients. Decreased presynaptic dopamine transporter binding was found in all CBD patients while D2 receptor binding was reduced in only one patient. (18)F-FDG PET displayed a contralateral hypometabolism in cortical and subcortical areas in seven out of eight patients. Our results demonstrate that glucose metabolism and DAT are reduced, while D2 receptors may be frequently preserved in CBD.