OPUS/IDENT

Already a single miss-labeled or contaminated raw material can lead to significant costs e.g. due to necessary re-working of a product, disposal costs or loss of valuable manufacturing time. OPUS IDENT is a powerful tool to easily build identification libraries including many different materials.

Advanced algorithms and hierarchical classification structure allow both to establish the identity of a material and to confirm that it lies within the acceptable statistical range of variation, but can be distinguished from closely-related materials. The complete IDENT method is enclosed in a single method for easy implementation in QC environments.

IDENT Key Features:

• Easy Setup even with hundreds of materials.
• Hierarchical Structure and multi-step evaluation through the library structure ensure identification of chemically similar materials.
• One common Library for all materials measured with the same analysis technique, e.g. one for solids and one for liquids.
• One Click Creation of suggested sub-library structure for overlapping groups.
• Internal Validation using cross validation to separate groups for individual materials.
• External Test Set using independent positive and negative samples to prove the method performance.
• Multiple Algorithms, spectral pre-treatments and factorization available as well as guidance for factor selection and score plots in 2D and 3D.

• Class Assignment in case of non-separable groups due to high degree of chemical similarity or non-detectable spectral differences.
• Multi Evaluation: Identification can be followed automatically by other evaluation steps, such as conformity tests and quantification cascades.
• Protection of Method Files for usage with dedicated instruments and electronic signature in compliance with 21CFR11 requirements.

Conformity Test

Being part of the OPUS/IDENT package the conformity test uses the spectral variation between different lots of raw materials having been determined acceptable to define pass/fail thresholds at each spectral data point. A new raw material lot is deemed acceptable if its NIR spectrum lies within the individual material dependent established threshold. Any deviation in the NIR spectral intensity beyond the established threshold will give cause for the rejection of the questionable raw material.

Conformity Test Key Features:

• Easy Method Setup with spectra tables, parameter page and graphical result plots.
• Validation of Methods during optimization with an external Test Spectra data set.
• Three different statistical Metrics with adjustable threshold sensitivities.
• Multi Evaluation: Conformity test can be automatically performed after an identification step and followed by quantification cascades.