Leo J. Tick

Natural Language Processing and the Representation of Clinical Data

OBJECTIVE Develop a representation of clinical observations and actions and a method of processing free-text patient documents to facilitate applications such as quality assurance. DESIGN The Linguistic String Project (LSP) system of New York University utilizes syntactic analysis, augmented by a sublanguage grammar and an information structure that are specific to the clinical narrative, to map free-text documents into a database for querying. MEASUREMENTS Information precision (I-P) and information recall (I-R) were measured for queries for the presence of 13 asthma-health-care quality assurance criteria in a database generated from 59 discharge letters. RESULTS I-P, using counts of major errors only, was 95.7% for the 28-letter training set and 98.6% for the 31-letter test set. I-R, using counts of major omissions only, was 93.9% for the training set and 92.5% for the test set.

The Estimation of “Transfer Functions” of Quadratic Systems

In recent years the use of stationary random inputs as a forcing function to experimentally determine the transfer function of linear systems has become widespread. The procedure involves the measurement of spectra and crossspectra between the input and output and the formation of the proper ratio. There are two basic reasons for using random testing functions. For many situations, particularly mechanical ones, these inputs are easier to generate than say steps or impulses and frequently they can be made to more closely approximate the in-service input. (The latter attribute is an advantage since the linearity of the system may not be complete but may be sufficiently so over the operating range of interest). A simple measure of linearity is immediately available i.e., the coherency. Although many experimenters try to generate a Gaussian stationary process for the forcing function this is not necessary from an expected value veiwpoint. Actually the probability structure plays no role in the general logic of the detailed procedure since only second moment characteristics of the process are relevant to the expected value calculations. However, a Gaussian input can be convenient since sometimes the evaluation of the variability of the estimates is simplified. In studying higher order systems by driving them with a random forcing function this use of a Gaussian process makes the calculations of the expected values considerably easier. In this paper we shall extend the spectral techniques of transfer function estimation of linear systems to time invariant quadratic systems when a stationary Gaussian process is used as a driving function.

A Method for Electrocardiogram Wave-Pattern Estimation Example: Left Ventricular Hypertrophy

Canine antisera to rabbit and hog renin were coupled to several fluorescein dyes. The antiserum was immunologically adsorbed once each with rabbit-liver and with bone-marrow powder, thereby removing excess dye and eliminating nonspecific staining. Mounted frozen-dried sections of kidneys from sodium- deficient rabbits (in which hyperplasia and hypergranulation of juxtaglomerular [JG] cells were present), from control rabbits, from a sodium-deficient dog, and from sodium-deficient rats were treated with the adsorbed labeled antiserum. Ultraviolet microscopy (direct technique) revealed an intense yel low-green fluorescence sharply limited to the granules of the juxtagloinerular cells in all sections studied from kidneys of rabbit and dog. JG granules in rats did not fluoresce, an observation in accord with the species spe cificity of renin. The indirect (“sandwich”) method was also employed (with adsorbed, labeled, rabbit antiserum to canine globulin), and although slight staining of gloinerular epithelium resulted, that in the JG granules was far more intense. In our hands, the faint glornerular staining was not blocked by prior treatment with unlabeled renin. Staining of JG granules in any kidney paralleled the intensity of JG granulation by light microscopy and the amount of extractable renin in the same kidney. JG-granule staining (direct and indirect techniques) was blocked by neutralization of the antirenin with renin. Heterogenous anti sera (to insulin; to human gamma globulin) failed to stain. Other rabbit tissues (heart, lung, liver) similarly treated with labeled antirenin never stained. If this work can be repeated with immunologically pure renin, the evidence presented here, together with previously published studies from this and other laboratories, will establish beyond any reasonable doubt that the source of renin in the kidney is the juxtaglomerular cell, as postulated first by Goormagh tigh nearly a quarter of a century ago.

Sampling Rates and Appearance of Stationary Gaussian Processes

The expected number of maxima and level crossings of a continuous stationary Gaussian process and the discrete process obtained by sampling the continuous one are evaluated and compared. The ratio of these two values as a function of sampling rate for two classes of spectra is calculated. It is shown that the rule of thumb of sampling at twice the Nyquist frequency is a good one.