Analytica Chimica Acta 787 (2013) 173–180 Contents lists available at SciVerse ScienceDirect Analytica Chimica Acta j ourna l ho mepage: www.elsevier.com/locate/aca A new approach to determine the homogeneity in hyperspectral imaging considering the particle size J.G. Rosas a , S. Armenta a , J. Cruz b , M. Blanco a,∗ a Department of Chemistry, Faculty of Sciences, Universitat Autònoma de Barcelona, E-08193 Bellaterra, Barcelona, Spain b Escola Universitària Salesiana de Sarrià (EUSS), Passeig Sant Joan Bosco, 74, 08017 Barcelona, Spain h i g h l i g h t s • Homogeneity assessment of pow- der mixtures by a new methodology based in the macropixel analysis. • The macropixel size selection is based in a statistical criterion and considers the particle size. • A relationship between H% Poole and controlled homogeneity of simulated images is established. • The homogeneity of binary and ternary mixtures of pharmaceuticals has been evaluated. g r a p h i c a l a b s t r a c t a r t i c l e i n f o Article history: Received 22 March 2013 Received in revised form 16 May 2013 Accepted 23 May 2013 Available online 6 June 2013 Keywords: Macropixel Homogeneity index Homogeneity Chemical imaging a b s t r a c t Determining homogeneity of a mixture using hyperspectral-imaging (HSI) techniques is at this moment an interesting research subject in pharmaceutical industry. A new methodology based in the macropixel analysis technique for the homogeneity assessment in powder blend mixtures has been developed. The appropriate selection of the macropixel size for evaluation of the homogeneity is a topic under discussion. We propose that the macropixel size selection would be based in a statistical criterion called “represen- tative sample size” criterion. Moreover, it has been demonstrated that the macropixel size should be a function of the particle size and, thus, the particle size has been included in the formula to calculate the macropixel size. The proposed equations allow to accurately calculate the homogeneity index (H% Poole). By means of using simulated chemical images, a relationship between the H% Poole index and the controlled homogeneity was established. The method has been applied to determine the homogene- ity of binary and ternary powder blends mixtures of different pharmaceuticals discriminating between homogeneous and non-homogeneous samples. © 2013 Elsevier B.V. All rights reserved. 1. Introduction Under current good manufacturing practices (cGMPs), an appli- cant is required to perform a test or examination on each commercial batch of all products to monitor the output and vali- date the performance of processes that could be responsible for ∗ Corresponding author. Tel.: +34 93 581 1367. E-mail addresses: sergio.armenta@uv.es (S. Armenta), marcel.blanco@uab.es (M. Blanco). causing variability, which includes adequacy of mixing to ensure uniformity and homogeneity [1]. Homogeneity, in the pharmaceu- tical field, can be defined as the uniform distribution of the active pharmaceutical ingredient (API) between withdrawn samples [2]. In the pharmaceutical industry, standards on mixture homogene- ity are based on three criteria that must be fulfilled for each of the active ingredients after sampling of 10 samples: (1) Each individual value (x i ) must lie in a -10%/+10% interval around the real value () (European pharmacopeia) or the mean value (x m ) [3] and (2) The coefficient of variation CV = s/x m must be less than 5%. In this point, it should be differentiated two types of homogeneity analysis 0003-2670/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.aca.2013.05.040