Binarization and cleanup of handwritten text from carbon copy medical form images

This paper presents a methodology for separating handwritten foreground pixels, from background pixels, in carbon copied medical forms. Comparisons between prior and proposed techniques are illustrated. This study involves the analysis of the New York State (NYS) Department of Health (DoH) Pre-Hospital Care Report (PCR) [Western Regional Emergency Medical Services, Bureau of Emergency Medical Services, New York State (NYS) Department of Health (DoH), Prehospital Care Report v4.] which is a standard form used in New York by all Basic and Advanced Life Support pre-hospital health care professionals to document patient status in the emergency environment. The forms suffer from extreme carbon mesh noise, varying handwriting pressure sensitivity issues, and smudging which are further complicated by the writing environment. Extraction of handwriting from these medical forms is a vital step in automating emergency medical health surveillance systems.     

Automatic recognition of handwritten medical forms for search engines

A new paradigm, which models the relationships between handwriting and topic categories, in the context of medical forms, is presented. The ultimate goals are: (1) a robust method which categorizes medical forms into specified categories, and (2) the use of such information for practical applications such as an improved recognition of medical handwriting or retrieval of medical forms as in a search engine. Medical forms have diverse, complex and large lexicons consisting of English, Medical and Pharmacology corpus. Our technique shows that a few recognized characters, returned by handwriting recognition, can be used to construct a linguistic model capable of representing a medical topic category. This allows (1) a reduced lexicon to be constructed, thereby improving handwriting recognition performance, and (2) PCR (Pre-Hospital Care Report) forms to be tagged with a topic category and subsequently searched by information retrieval systems. We present an improvement of over 7% in raw recognition rate and a mean average precision of 0.28 over a set of 1,175 queries on a data set of unconstrained handwritten medical forms filled in emergency environments. This work was supported by the National Science Foundation.     

Unconstrained handwritten document retrieval

With the ever-increasing growth of the World Wide Web, there is an urgent need for an efficient information retrieval system that can search and retrieve handwritten documents when presented with user queries. However, unconstrained handwriting recognition remains a challenging task with inadequate performance thus proving to be a major hurdle in providing robust search experience in handwritten documents. In this paper, we describe our recent research with focus on information retrieval from noisy text derived from imperfect handwriting recognizers. First, we describe a novel term frequency estimation technique incorporating the word segmentation information inside the retrieval framework to improve the overall system performance. Second, we outline a taxonomy of different techniques used for addressing the noisy text retrieval task. The first method uses a novel bootstrapping mechanism to refine the OCR’ed text and uses the cleaned text for retrieval. The second method uses the uncorrected or raw OCR’ed text but modifies the standard vector space model for handling noisy text issues. The third method employs robust image features to index the documents instead of using noisy OCR’ed text. We describe these techniques in detail and also discuss their performance measures using standard IR evaluation metrics.     

Synthetic handwritten CAPTCHAs

CAPTCHAs (completely automated public Turing test to tell computers and humans apart) are in common use today as a method for performing automated human verification online. The most popular type of CAPTCHA is the text recognition variety. However, many of the existing printed text CAPTCHAs have been broken by web-bots and are hence vulnerable to attack. We present an approach to use human-like handwriting for designing CAPTCHAs. A synthetic handwriting generation method is presented, where the generated textlines need to be as close as possible to human handwriting without being writer-specific. Such handwritten CAPTCHAs exploit the differential in handwriting reading proficiency between humans and machines. Test results show that when the generated textlines are further obfuscated with a set of deformations, machine recognition rates decrease considerably, compared to prior work, while human recognition rates remain the same.     

Reading handwritten phrases on U.S. census forms

Commercial form-reading systems for extraction of data from forms do not meet acceptable accuracy requirements on forms filled out by hand. Several important form-processing applications involve the automated reading of handwritten responses. U.S. Census forms are a case in point. A database of form images containing actual responses received by the U.S. Census Bureau was made available by National Institute of Standards and Technology (NIST) in December 1993. A number of factors combine to make the task of reading these forms a challenging one. The quality of form images is often poor, and the handwritten responses are very loosely constrained in terms of writing style, format of response, and choice of text. The sizes of the lexicons provided are large (10,000-50,000 entries) and yet the coverage is incomplete (60%-70%). In this article we discuss our approach to automate the task of reading the census forms. The subtasks of field extraction and phrase recognition are described and multiclassifier control strategies for phrase recognition are presented. The error rate of the system when no rejects are allowed is 59%, with a lower bound of 40% being imposed by the incomplete coverage of the lexicon. The article concludes with a discussion of experimental results and directions for future research. © 1996 John Wiley & Sons, Inc.     

Role of Post-translational Modifications in Alzheimer's Disease

The global burden of Alzheimer's disease (AD) is growing. Valiant efforts to develop clinical candidates for treatment have continuously met with failure. Currently available palliative treatments are temporary and there is a constant need to search for reliable disease pathways, biomarkers and drug targets for developing diagnostic and therapeutic tools to address the unmet medical needs of AD. Challenges in drug-discovery efforts raise further questions about the strategies of current conventional diagnosis; drug design; and understanding of disease pathways, biomarkers and targets. In this context, post-translational modifications (PTMs) regulate protein trafficking, function and degradation, and their in-depth study plays a significant role in the identification of novel biomarkers and drug targets. Aberrant PTMs of disease-relevant proteins could trigger pathological pathways, leading to disease progression. Advancements in proteomics enable the generation of patterns or signatures of such modifications, and thus, provide a versatile platform to develop biomarkers based on PTMs. In addition, understanding and targeting the aberrant PTMs of various proteins provide viable avenues for addressing AD drug-discovery challenges. This review highlights numerous PTMs of proteins relevant to AD and provides an overview of their adverse effects on the protein structure, function and aggregation propensity that contribute to the disease pathology. A critical discussion offers suggestions of methods to develop PTM signatures and interfere with aberrant PTMs to develop viable diagnostic and therapeutic interventions in AD.     

Production of okara and soy protein concentrates using membrane technology

Microfiltration (MF) membranes with pore sizes of 200 and 450 nm and ultrafiltration (UF) membranes with molecular weight cut off of 50, 100, and 500 kDa were assessed for their ability to eliminate nonprotein substances from okara protein extract in a laboratory cross-flow membrane system. Both MF and UF improved the protein content of okara extract to a similar extent from approximately 68% to approximately 81% owing to the presence of protein in the feed leading to the formation of dynamic layer controlling the performance rather than the actual pore size of membranes. Although normalized flux in MF-450 (117 LMH/MPa) was close to UF-500 (118 LMH/MPa), the latter was selected based on higher average flux (47 LMH) offering the advantage of reduced processing time. Membrane processing of soy extract improved the protein content from 62% to 85% much closer to the target value. However, the final protein content in okara (approximately 80%) did not reach the target value (90%) owing to the greater presence of soluble fibers that were retained by the membrane. Solubility curve of membrane okara protein concentrate (MOPC) showed lower solubility than soy protein concentrate and a commercial isolate in the entire pH range. However, water absorption and fat-binding capacities of MOPC were either superior or comparable while emulsifying properties were in accordance with its solubility. The results of this study showed that okara protein concentrate (80%) could be produced using membrane technology without loss of any true proteins, thus offering value addition to okara, hitherto underutilized. Practical Application: Okara, a byproduct obtained during processing soybean for soymilk, is either underutilized or unutilized in spite of the fact that its protein quality is as good as that of soy milk and tofu. Membrane-processed protein products have been shown to possess superior functional properties compared to conventionally produced protein products. However, the potential of membrane technology has not been exploited for the recovery of okara protein. Our study showed that protein content of okara extract could be improved from approximately 68% to approximately 81% without losing any true proteins in the process.     

Rapid thermal annealing effects on the photoluminescence properties of molecular beam epitaxy-grown GaIn(N)As quantum wells with Ga(N)As spacers and barriers

This article describes the effects of rapid thermal annealing (RTA) on the photoluminescence (PL) emission from a series of GaIn(N)As quantum wells. Indium compositions of both 20% and 32% were examined with nominal N compositions of 1% or 2%. The N location was varied within our quantum structure, which can be divided into three regions: (1) quantum well, (2) Ga(N)As spacer layers at the barrier-to-well interface and well-to-barrier interface, and (3) barriers surrounding each quantum well. Eight combinations of samples were examined with varying In content, Ga(N)As spacer layer thickness, N content, and N location in the structure. In the best cases, the presence of these Ga(N)As spacer layers improves the PL properties, due to annealing, with a reduction in the emission wavelength blueshift by ～400 Å, a reduction of the decrease in the full-width at half-maximum (FWHM) by ～5 meV, and a threefold reduction of the increase in integrated intensity. It was also observed that relocating N from the quantum wells to the barriers produces a comparable emission wavelength both before and after annealing. Our results further show that the composition of incorporated N in the material is most influential during the stages of RTA in which relatively small amounts of thermal energy is present from our lower annealing times and temperatures. Hence, we believe a low thermal-energy anneal is responsible for the recovery of the plasma-related crystal damage that was incurred during its growth. However, the In composition in the quantum well is most influential during the latter stages of thermal annealing, at increased times and temperatures, where the wavelength blueshift was roughly independent of the amount of incorporated N. As a result, our investigations into the effects of RTA on the PL properties support other reports that suggest the wavelength blueshift is not due to N diffusion.