New possibilities exist for the development of novel hardware/software platforms havin g fast data acquisition capability with low power requirements. One application is a high speed Adaptive Design for Information (ADI) system that combines the advantages of feature-based data compression, low power nanometer CMOS technology, and stream computing . We have developed a compressive sensing (CS) algorithm which linearly reduces the data at the analog front end, an approach which uses analog designs and computations instead of smaller feature size transistors for higher speed and lower power. A level-crossing sampling approach replaces Nyquist sampling. With an in-memory design, the new compressive sensing based instrumentation performs digitization only when there is enough variation in the input and when the random selection matrix chooses this input.
Advances in integrated circuit technologies have led to the integration of medical sensor front ends with data processing circuits, i.e., mobile platform design for wearable sensors. We discuss design methodologies for wearable sensor nodes and their applications in m-Health. From the user perspective, flexibility, comfort, appearance, fashion, ease-of-use, and visibility are key form factors. From the technology development point of view, high accuracy, low power consumption, and high signal to noise ratio are desirable features. From the embedded software design standpoint, real time data analysis algorithms, application and database interfaces are the critical components to create successful wearable sensor-based products.
In health care applications, we obtain, manage, store and communicate using high quality, large volume of image data through integrated devices. In this paper we propose several promising methods that can assist physicians in image data process and communication. We design a new semi-automated segmentation approach for radiological images, such as CT and MRI to clearly identify the areas of interest. This approach combines the advantages from both the region-based method and boundary-based methods. It has three key steps compose: coarse segmentation by using fuzzy affinity and homogeneity operator, image division and reclassification using the Voronoi Diagram, and refining boundary lines using the level set model.
Proc. SPIE. 8853, Medical Applications of Radiation Detectors III
KEYWORDS: Infrared imaging, Digital signal processing, Image compression, Image restoration, Image acquisition, Medical imaging, Digital imaging, Mammography, High dynamic range imaging, Compressed sensing
Modern imaging modalities, such as Computed Tomography (CT), Digital Breast Tomosynthesis (DBT) or Magnetic
Resonance Tomography (MRT) are able to acquire volumetric images with an isotropic resolution in micrometer (um) or
millimeter (mm) range. When used in interactive telemedicine applications, these raw images need a huge storage unit,
thereby necessitating the use of high bandwidth data communication link. To reduce the cost of transmission and enable
archiving, especially for medical applications, image compression is performed. Recent advances in compression
algorithms have resulted in a vast array of data compression techniques, but because of the characteristics of these images,
there are challenges to overcome to transmit these images efficiently. In addition, the recent studies raise the low dose
mammography risk on high risk patient. Our preliminary studies indicate that by bringing the compression before the
analog-to-digital conversion (ADC) stage is more efficient than other compression techniques after the ADC. The linearity
characteristic of the compressed sensing and ability to perform the digital signal processing (DSP) during data conversion
open up a new area of research regarding the roles of sparsity in medical image registration, medical image analysis (for
example, automatic image processing algorithm to efficiently extract the relevant information for the clinician), further Xray
dose reduction for mammography, and contrast enhancement.
In this paper, we present a design of a multi optical modalities blood glucose monitor. The Monte Carlo tissues optics
simulation with typical human skin model suggests the SNR ratio for a detector sensor is 104 with high sensitivity that
can detect low blood sugar limit at 1 mMole/dL ( <20 mg/dL). A Bayesian filtering algorithm is proposed for multisensor
fusion to identify whether e user has the danger of having diabetes. The new design has real time response (on the
average of 2 minutes) and provides great potential to perform real time monitoring for blood glucose.
Conference Committee Involvement (1)
Micro- and Nanotechnology Sensors, Systems, and Applications IX