Bootstrap Error-adjusted Single-sample Technique Research Articles

Acoustic-resonance spectrometry as a process analytical technology for rapid and accurate tablet identification.

This research was performed to test the hypothesis that acoustic-resonance spectrometry (ARS) is able to rapidly and accurately differentiate tablets of similar size and shape. The US Food and Drug Administration frequently orders recalls of tablets because of labeling problems (eg, the wrong tablet appears in a bottle). A high-throughput, nondestructive method of online analysis and label comparison before shipping could obviate the need for recall or disposal of a batch of mislabeled drugs, thus saving a company considerable expense and preventing a major safety risk. ARS is accurate and precise as well as inexpensive and nondestructive, and the sensor, is constructed from readily available parts, suggesting utility as a process analytical technology (PAT). To test the classification ability of ARS, 5 common household tablets of similar size and shape were chosen for analysis (aspirin, ibuprofen, acetaminophen, vitamin C, and vitamin B12). The measures of successful tablet identification were intertablet distances in nonparametric multidimensional standard deviations (MSDs) greater than, 3 and intratablet MSDs less than 3, as calculated from an extended bootstrap erroradjusted single sample technique. The average intertablet MSD was 65.64, while the average intratablet MSD from cross-validation was 1.91. Tablet mass (r2=0.977), thickness (r2=0.977), and density (r2=0.900) were measured very accurately from the AR spectra, each with less than 10% error. Tablets were identified correctly with only 250 ms data collection time. These results demonstrate that ARS effectively identified and characterized the 5 types of tablets and could potentially serve as a rapid high-throughput online pharmaceutical sensor.

A Process Analytical Technology approach to near-infrared process control of pharmaceutical powder blending: Part II: Qualitative near-infrared models for prediction of blend homogeneity

The successful implementation of near-infrared spectroscopy (NIRS) in process control of powder blending requires constructing an inclusive spectral database that reflects the anticipated voluntary or involuntary changes in processing conditions, thereby minimizing bias in prediction of blending behavior. In this study, experimental design was utilized as an efficient way of generating blend experiments conducted under varying processing conditions such as humidity, blender speed and component concentration. NIR spectral data, collected from different blending experiments, was used to build qualitative models for prediction of blend homogeneity. Two pattern recognition algorithms: Soft Independent Modeling of Class Analogies (SIMCA) and Principal Component Modified Bootstrap Error-adjusted Single-sample Technique (PC-MBEST) were evaluated for qualitative analysis of NIR blending data. Optimization of NIR models, for the two algorithms, was achieved by proper selection of spectral processing, and training set samples. The models developed were successful in predicting blend homogeneity of independent blend samples under different processing conditions. © 2005 Wiley-Liss, Inc. and the American Pharmacists Association

Hyperspectral techniques in analysis of oral dosage forms.

Pharmaceutical oral dosage forms are used in this paper to test the sensitivity and spatial resolution of hyperspectral imaging instruments. The first experiment tested the hypothesis that a near-infrared (IR) tunable diode-based remote sensing system is capable of monitoring degradation of hard gelatin capsules at a relatively long distance (0.5 km). Spectra from the capsules were used to differentiate among capsules exposed to an atmosphere containing 150 ppb formaldehyde for 0, 2, 4, and 8 h. Robust median-based principal component regression with Bayesian inference was employed for outlier detection. The second experiment tested the hypothesis that near-IR imaging spectrometry of tablets permits the identification and composition of multiple individual tablets to be determined simultaneously. A near-IR camera was used to collect thousands of spectra simultaneously from a field of blister-packaged tablets. The number of tablets that a typical near-IR camera can currently analyze simultaneously was estimated to be approximately 1300. The bootstrap error-adjusted single-sample technique chemometric-imaging algorithm was used to draw probability-density contour plots that revealed tablet composition. The single-capsule analysis provides an indication of how far apart the sample and instrumentation can be and still maintain adequate signal-to-noise ratio (S/N), while the multiple-tablet imaging experiment gives an indication of how many samples can be analyzed simultaneously while maintaining an adequate S/N and pixel coverage on each sample.

Multispectral imaging of tablets in blister packaging.

This experiment tested the hypothesis that using near-infrared (IR) imaging spectrometry on tablets through blister packs permits the identification and composition of multiple individual tablets to be determined simultaneously. Aspirin was selected for this study because its breakdown mechanism is well understood. Near-IR cameras were used to collect thousands of spectra simultaneously from a field of packaged aspirin tablets. Tablets were selected by a principal component analysis selection algorithm. Graphs of the columns of the transformation matrix showed that salicylic acid and acetylsalicylic acid in the samples were modeled by the principal components. The bootstrap error-adjusted single-sample technique chemometric-imaging algorithm was used to draw probability-density contour plots that revealed tablet composition. Choice of color was used to represent constituent identity, whereas intensity represented concentration. The percentage of usable pixels in the indium antimonide (InSb) array was 99.9%. The SEP was 0.06% of the tablet mass for both water uptake and salicylic acid production. The number of tablets that a typical near-IR camera can currently analyze simultaneously was also estimated to be approximately 1300.

Detection of Subpopulations in Near-Infrared Reflectance Analysis

In typical near-infrared multivariate statistical analyses, samples with similar spectra produce points that cluster in a certain region of spectral hyperspace. These clusters can vary significantly in shape and size due to variation in sample packings, particle-size distributions, component concentrations, and drift with time. These factors, when combined with discriminant analysis using simple distance metrics, produce a test in which a result that places a particular point inside a particular cluster does not necessarily mean that the point is actually a member of the cluster. Instead, the point may be a member of a new, slightly different cluster that overlaps the first. A new cluster can be created by factors like low-level contamination or instrumental drift. An extention added to part of the BEAST (Bootstrap Error-Adjusted Single-sample Technique) can be used to set nonparametric probability-density contours inside spectral clusters as well as outside, and when multiple points begin to appear in a certain region of cluster-hyperspace the perturbation of these density contours can be detected at an assigned significance level. The detection of false samples both within and beyond 3 SDs of the center of the training set is possible with this method. This procedure is shown to be effective for contaminant levels of a few hundred ppm in an over-the-counter drug capsule, and is shown to function with as few as one or two wavelengths, suggesting its application to very simple process sensors.

Quantile BEAST Attacks the False-Sample Problem in Near-Infrared Reflectance Analysis

The multiple linear regression approach typically used in near-infrared calibration yields equations in which any amount of reflectance at the analytical wavelengths leads to a corresponding composition value. As a result, when the sample contains a component not present in the training set, erroneous composition values can arise without any indication of error. The Quantile BEAST (Bootstrap Error-Adjusted Single-sample Technique) is described here as a method of detecting one or more “false” samples. The BEAST constructs a multidimensional form in space using the reflectance values of each training-set sample at a number of wavelengths. New samples are then projected into this space, and a confidence test is executed to determine whether the new sample is part of the training-set form. The method is more robust than other procedures because it relies on few assumptions about the structure of the data; therefore, deviations from assumptions do not affect the results of the confidence test.

Detection of capsule tampering by near-infrared reflectance analysis

Abstract : The growing incidence of product tampering has brought to attention the need for a rapid, reliable, inexpensive, noninvasive and nondestructive method of screening. Such a method, based on near-infrared reflectance analysis (NIRA), is presented here for the detection of adulterated non prescription drugs. The method relies upon a nonparametric clustering algorithm known as the BEAST (Bootstrap Error-Adjusted Single-sample Technique). Specially-designed sampling reflectors have been constructed to enable rapid and convenient measurement of capsules. A right-circular conical reflector has been found to be optimal for this purpose; the capsules fit directly into the reflector and need not be opened for analysis. A variety of foreign substances have been successfully detected in capsules by probing the capsule contents directly through the gelatin walls; these substances include Fe2O3, Al shavings, NaF, As2O3 NaCN and KCN. The NIRA response for KCN is linear down to a detection limit of 2.6 mg (0.4% of capsule weight). An incidental advantage in the use of the conical reflector is that the response is dependent not only upon the mass of adulterant but its location within the capsule. Keywords: Sensors, Near infrared reflectance analysis, Capsule tampering, Fault detection.