-
IEEE Transactions on Image Processing :... Feb 2023Most recent methods for RGB (red-green-blue)-thermal salient object detection (SOD) involve several floating-point operations and have numerous parameters, resulting in...
Most recent methods for RGB (red-green-blue)-thermal salient object detection (SOD) involve several floating-point operations and have numerous parameters, resulting in slow inference, especially on common processors, and impeding their deployment on mobile devices for practical applications. To address these problems, we propose a lightweight spatial boosting network (LSNet) for efficient RGB-thermal SOD with a lightweight MobileNetV2 backbone to replace a conventional backbone (e.g., VGG, ResNet). To improve feature extraction using a lightweight backbone, we propose a boundary boosting algorithm that optimizes the predicted saliency maps and reduces information collapse in low-dimensional features. The algorithm generates boundary maps based on predicted saliency maps without incurring additional calculations or complexity. As multimodality processing is essential for high-performance SOD, we adopt attentive feature distillation and selection and propose semantic and geometric transfer learning to enhance the backbone without increasing the complexity during testing. Experimental results demonstrate that the proposed LSNet achieves state-of-the-art performance compared with 14 RGB-thermal SOD methods on three datasets while improving the numbers of floating-point operations (1.025G) and parameters (5.39M), model size (22.1 MB), and inference speed (9.95 fps for PyTorch, batch size of 1, and Intel i5-7500 processor; 93.53 fps for PyTorch, batch size of 1, and NVIDIA TITAN V graphics processor; 936.68 fps for PyTorch, batch size of 20, and graphics processor; 538.01 fps for TensorRT and batch size of 1; and 903.01 fps for TensorRT/FP16 and batch size of 1). The code and results can be found from the link of https://github.com/zyrant/LSNet.
PubMed: 37022901
DOI: 10.1109/TIP.2023.3242775 -
Asian Pacific Journal of Cancer... Oct 2021Liquid-based preparation (LBP) cytology is commonly used in most laboratories these days due to its convenience and reliable results for the cervical cancer screening...
UNLABELLED
Liquid-based preparation (LBP) cytology is commonly used in most laboratories these days due to its convenience and reliable results for the cervical cancer screening program. The PathTezt™ Liquid-based Pap smear is a second-generation LBP, which uses a filter-based concentration technique in processing the sample.
OBJECTIVE
This study was done to evaluate the cellular fixation, morphology, quality of smear in gynae cytology, and diagnostic interpretation of cervical cytological smears produced by the PathTezt liquid-based processor.
MATERIALS AND METHODS
A total of 400 pap smear samples were taken and processed using the PathTezt 2000 processor. The slides were evaluated in terms of sample adequacy, percentage of the circle covered by epithelial cells, cellular distribution, obscuring factors, and cell fixation.
RESULTS
About 95.25% (381) of the samples were satisfactory for the evaluation. In 19 (4.75%) of the samples, epithelial cells covered less than 50% of the circle. A sample with good cellular distribution was seen in 92% of the cases, while 354 (88.5%) samples showed minimal inflammatory background. Almost all the smears (95.75%) had no erythrocytes in the background. All smears showed good quality fixation features toward nuclear, cytoplasm, and microorganisms. The total performance rate was 99%.
CONCLUSION
Although the PathTezt liquid-based processor is still new compared to other first-generation LBP, the smears produced by this method were of high quality and it was cost-effective.
Topics: Adult; Aged; Cervix Uteri; Cost-Benefit Analysis; Epithelial Cells; Female; Humans; Malaysia; Middle Aged; Papanicolaou Test; Uterine Cervical Neoplasms; Vaginal Smears; Young Adult
PubMed: 34711003
DOI: 10.31557/APJCP.2021.22.10.3261 -
Journal of the American Academy of... Mar 2021Cochlear implant (CI) recipients frequently experience difficulty understanding speech over the telephone and rely on hearing assistive technology (HAT) to improve...
BACKGROUND
Cochlear implant (CI) recipients frequently experience difficulty understanding speech over the telephone and rely on hearing assistive technology (HAT) to improve performance. Bilateral inter-processor audio streaming technology using nearfield magnetic induction is an advanced technology incorporated within a hearing aid or CI processor that can deliver telephone audio signals captured at one sound processor to the sound processor at the opposite ear. To date, limited data exist examining the efficacy of this technology in CI users to improve speech understanding on the telephone.
PURPOSE
The primary objective of this study was to examine telephone speech recognition outcomes in bilateral CI recipients in a bilateral inter-processor audio streaming condition (DuoPhone) compared with a monaural condition (i.e., telephone listening with one sound processor) in quiet and in background noise. Outcomes in the monaural and bilateral conditions using either a telecoil or T-Mic2 technology were also assessed. The secondary aim was to examine how deactivating microphone input in the contralateral processor in the bilateral wireless streaming conditions, and thereby modifying the signal-to-noise ratio, affected speech recognition in noise.
RESEARCH DESIGN
A repeated-measures design was used to evaluate speech recognition performance in quiet and competing noise with the telephone signal transmitted acoustically or via the telecoil to the ipsilateral sound processor microphone in monaural and bilateral wireless streaming listening conditions.
STUDY SAMPLE
Nine bilateral CI users with Advanced Bionics HiRes 90K and/or CII devices were included in the study.
DATA COLLECTION AND ANALYSIS
The effects of phone input (monaural [DuoPhone Off] vs. bilateral [DuoPhone on]) and processor input (T-Mic2 vs. telecoil) on word recognition in quiet and noise were assessed using separate repeated-measures analysis of variance. Effect of the contralateral device mic deactivation on speech recognition outcomes for the T-Mic2 DuoPhone conditions was assessed using paired Student's -tests.
RESULTS
Telephone speech recognition was significantly better in the bilateral inter-processor streaming conditions relative to the monaural conditions in both quiet and noise. Speech recognition outcomes were similar in quiet and noise when using the T-Mic2 and telecoil in the monaural and bilateral conditions. For the acoustic DuoPhone conditions using the T-Mic2, speech recognition in noise was significantly better when the microphone of the contralateral processor was disabled.
CONCLUSION
Inter-processor audio streaming allows for bilateral listening on the telephone and produces better speech recognition in quiet and in noise compared with monaural listening conditions for adult CI recipients.
Topics: Adult; Cochlear Implantation; Cochlear Implants; Hearing; Humans; Speech Perception; Telephone
PubMed: 33873219
DOI: 10.1055/s-0041-1722982 -
PloS One 2021Gordon Moore famously observed that the number of transistors in state-of-the-art integrated circuits (units per chip) increases exponentially, doubling every 12-24...
Gordon Moore famously observed that the number of transistors in state-of-the-art integrated circuits (units per chip) increases exponentially, doubling every 12-24 months. Analysts have debated whether simple exponential growth describes the dynamics of computer processor evolution. We note that the increase encompasses two related phenomena, integration of larger numbers of transistors and transistor miniaturization. Growth in the number of transistors per unit area, or chip density, allows examination of the evolution with a single measure. Density of Intel processors between 1959 and 2013 are consistent with a biphasic sigmoidal curve with characteristic times of 9.5 years. During each stage, transistor density increased at least tenfold within approximately six years, followed by at least three years with negligible growth rates. The six waves of transistor density increase account for and give insight into the underlying processes driving advances in processor manufacturing and point to future limits that might be overcome.
Topics: Electronics; Humans; Miniaturization; Transistors, Electronic
PubMed: 34407116
DOI: 10.1371/journal.pone.0256245 -
Otolaryngologic Clinics of North America Oct 2019The work-up and management of sensorineural hearing loss in children has been an area of rapid evolution. With the availability of genetic and cytomegalovirus testing,... (Review)
Review
The work-up and management of sensorineural hearing loss in children has been an area of rapid evolution. With the availability of genetic and cytomegalovirus testing, the diagnostic process is continuously refined. Aural rehabilitation should be provided to children in a timely manner. At present, the main surgical options for the treatment of sensorineural hearing loss are bone conduction sound processors and cochlear implants. Investigations into modalities such as auditory brainstem implants are ongoing. With further technological and medical advancements, the evaluation and management of pediatric sensorineural hearing loss will undoubtedly continue to change.
Topics: Auditory Brain Stem Implants; Bone Conduction; Child; Cochlear Implants; Hearing Loss, Sensorineural; Humans
PubMed: 31303327
DOI: 10.1016/j.otc.2019.05.004 -
Biomedical Engineering Letters May 2022Conventional spike sorting and motor intention decoding algorithms are mostly implemented on an external computing device, such as a personal computer. The innovation of...
Conventional spike sorting and motor intention decoding algorithms are mostly implemented on an external computing device, such as a personal computer. The innovation of high-resolution and high-density electrodes to record the brain's activity at the single neuron level may eliminate the need for spike sorting altogether while potentially enabling in vivo neural decoding. This article explores the feasibility and efficient realization of in vivo decoding, with and without spike sorting. The efficiency of neural network-based models for reliable motor decoding is presented and the performance of candidate neural decoding schemes on sorted single-unit activity and unsorted multi-unit activity are evaluated. A programmable processor with a custom instruction set architecture, for the first time to the best of our knowledge, is designed and implemented for executing neural network operations in a standard 180-nm CMOS process. The processor's layout is estimated to occupy 49 mm of silicon area and to dissipate 12 mW of power from a 1.8 V supply, which is within the tissue-safe operation of the brain.
PubMed: 35529345
DOI: 10.1007/s13534-022-00217-z -
American Journal of Clinical Pathology Jul 2022Tissue carryovers are contaminants of surgical pathology cases in which extraneous tissue is incorporated into tissue blocks. Carryovers occur most frequently at the...
OBJECTIVES
Tissue carryovers are contaminants of surgical pathology cases in which extraneous tissue is incorporated into tissue blocks. Carryovers occur most frequently at the grossing or embedding stations, but little is published about them. We sought to analyze their transmission during transit to the histology lab.
METHODS
Cassettes of friable donor tissue were mixed with cassettes of spongy recipient tissue in formalin-filled containers and agitated by shipment via pneumatic tube. The tissue cassettes were processed, embedded as blocks, and cut as usual. Liquid samples were prepared from the submission containers as well as from workstation submission containers and histology tissue processor waste.
RESULTS
A high rate of contamination (14.9%) was observed under these artificial conditions. Friable donor tissue, including urothelium and colorectal adenocarcinoma, were promiscuous contaminants, as were placental villi. Fluid from submission containers showed viable tumor cells and fragments, which were also present in workstation submission containers and in tissue processor waste fluid.
CONCLUSIONS
This study implicates liquid transport media as a possible avenue of contamination during submission and transportation of tissue cassettes for histologic processing. Attention should be given to the friability of submitted tissue and physical agitation of the cassettes in transit. Such contaminants may be present in the fluid in tissue submission bins and in tissue processor fluid.
Topics: Female; Formaldehyde; Humans; Neoplasms; Paraffin Embedding; Placenta; Pregnancy; Specimen Handling; Tissue Fixation
PubMed: 35195717
DOI: 10.1093/ajcp/aqac014 -
Nature Communications Dec 2022Tensor analytics lays the mathematical basis for the prosperous promotion of multiway signal processing. To increase computing throughput, mainstream processors...
Tensor analytics lays the mathematical basis for the prosperous promotion of multiway signal processing. To increase computing throughput, mainstream processors transform tensor convolutions into matrix multiplications to enhance the parallelism of computing. However, such order-reducing transformation produces data duplicates and consumes additional memory. Here, we propose an integrated photonic tensor flow processor (PTFP) without digitally duplicating the input data. It outputs the convolved tensor as the input tensor 'flows' through the processor. The hybrid manipulation of optical wavelengths, space dimensions, and time delay steps, enables the direct representation and processing of high-order tensors in the optical domain. In the proof-of-concept experiment, an integrated processor manipulating wavelengths and delay steps is implemented for demonstrating the key functionalities of PTFP. The multi-channel images and videos are processed at the modulation rate of 20 Gbaud. A convolutional neural network for video action recognition is demonstrated on the processor, which achieves an accuracy of 97.9%.
PubMed: 36577748
DOI: 10.1038/s41467-022-35723-2 -
Optics Letters Aug 2022We propose and experimentally demonstrate an optical processor for a binarized neural network (NN). Implementation of a binarized NN involves multiply-accumulate...
We propose and experimentally demonstrate an optical processor for a binarized neural network (NN). Implementation of a binarized NN involves multiply-accumulate operations, in which positive and negative weights should be implemented. In the proposed processor, the positive and negative weights are realized by switching the operations of a dual-drive Mach-Zehnder modulator (DD-MZM) between two quadrature points corresponding to two binary weights of +1 and -1, and the multiplication is also performed at the DD-MZM. The accumulation operation is realized by dispersion-induced time delays and detection at a photodetector (PD). A proof-of-concept experiment is performed. A binarized convolutional neural network (CNN) accelerated by the optical processor at a speed of 32 giga floating point operations/s (GFLOPS) is tested on two benchmark image classification tasks. The large bandwidth and parallel processing capability of the processor has high potential for next generation data computing.
Topics: Equipment Design; Neural Networks, Computer; Vision, Ocular
PubMed: 35913340
DOI: 10.1364/OL.464214 -
The Journal of International Advanced... Nov 2023RONDO 2 is a lightweight, compact, wirelessly charged, and fully integrated single-unit speech processor. Single-unit processors provide an effective and convenient...
BACKGROUND
RONDO 2 is a lightweight, compact, wirelessly charged, and fully integrated single-unit speech processor. Single-unit processors provide an effective and convenient alternative to behind-the-ear processors for adults. Therefore, the aim of this study was to investigate if RONDO 2 is suitable for and did not compromise the hearing performance of young children in everyday life.
METHODS
Thirteen children aged -lt;4 years were fitted with the RONDO 2 speech processor at the first activation of the cochlear implant. They were evaluated with the LittlEARS® Auditory Questionnaire, LittlEARS® Early Speech Production Questionnaire, and the Speech, Spatial, and Qualities of Hearing Scale 12 pre-implantation. In addition to these measures, they were evaluated with the Audio Processor Satisfaction Questionnaire post-implantation. Duration of daily use and troubleshooting data were acquired. Evaluation occurred at 4 time points: before implantation and 1, 3, and 12 months post-initial activation.
RESULTS
Ten out of 13 children continued using RONDO 2 after the study. Twelve months after implantation, they used it on average 11.6 hours per day and had an average Audio Processor Satisfaction Questionnaire score of 9.1 out of 10. Average hearing performance scores continuously improved throughout the follow-up period across measures. Twelve months after implantation, the mean scores were 30.1 out of 35 for the LittlEARS® Auditory Questionnaire, 19.9 out of 27 for the LittlEARS® Early Speech Production Questionnaire, and 7.4 out of 10 for the Speech, Spatial, and Qualities of Hearing Scale 12.
CONCLUSION
Participants demonstrated high levels of satisfaction and good hearing performance with RONDO 2, which indicates that this single-unit processor could be a viable and comfortable alternative to behind-the-ear processors in young children, although larger controlled experiments are warranted.
Topics: Adult; Child; Humans; Child, Preschool; Cochlear Implants; Cochlear Implantation; Hearing; Speech Perception; Hearing Tests
PubMed: 38088315
DOI: 10.5152/iao.2023.22979