EUROCON 2005 Serbia & Montenegro, Belgrade, November 22-24, 2005 Abstract — We propose a new algorithm to analyze cell migration. Sequences of frames are automatically recorded from standard (unmarked) cell cultures by means of phase- contrast microscopes equiped with video acquisition systems. This algorithm is able to automatically follow the locations in the reverse time of many cells during sequences covering relatively long periods of time such as 1 to 3 days. We then recombine the obtained cell tracks to detect mitoses and build a “mitotic tree”. Several features are extract to characterize cell population motility and proliferation. As illustration the method is tested on U373 astrocytoma cell line. Keywords — cell division, cell motility, image processing, mean-shift - tracking, video microscopy. I. INTRODUCTION N vitro cell displacement has been shown to be a useful parameter to consider in several biological applications, such as the study of cell migration and its variations under different culture conditions or drug actions (e.g., see [1- 5]). In order to analyze cell migration, it is usual to automatically record sequences of frames by means of microscopes equipped with video acquisition systems. The locations of each cell have to be tracked during entire sequences and qualitative and quantitative features have to be computed (average speed, maximum displacement, etc). Manual or interactive computer-assisted tracking quickly becomes a tedious task if a large number of cells must be tracked during long periods (dozens of hours). This is why automation is becoming increasingly popular [5]. Unfortunately, some acquisition modalities, imposed by the tackled biological problem, give poor image quality for automatic image processing (e.g. phase-contrast imaging in our case), leading to nonobvious image analysis problems that remain an interesting issue. While some kind of assay (e.g. scratch-wound [1]) use global approach, this kind of assay does not consider individual cell locomotion O.Debeir, D. Milojevic, T. Leloup and P. Van Ham are with the Department of Logical and Numerical Systems, Faculty of Applied Sciences, CP 165/57, Université Libre de Bruxelles , 50 Av. F. Roosevelt - 1050 Brussels - Belgium (corresponding author Tel: 32 2 650 2691, Fax: 32 2 650 2298, email: odebeir@ulb.ac.be ). R.Kiss and C.Decaestecker are with Laboratory of Toxicology, Institute of Pharmacy; CP 205/1, Université Libre de Bruxelles,50 Av. F. Roosevelt - 1050 Brussels - Belgium (e-mail: rkiss@ulb.ac.be, cdecaes@ulb.ac.be). nor does it distinguish between the roles played by cell migration and cell proliferation. We promote individual cell migration assay because it is well adapted to characterize biological mechanisms involving individual cells (such as inflammatory and immune reactions, and tumor spreading and metastasis). We showed in previous works the usefulness of the approach to characterize cell motility [6] and chemotaxism [7]. We will focus in this work on a specific event occuring in live cell cycle : the mitosis. Figure 1 shows a typical image obtained in vitro by standard phase contrast microscopy (no fluorescence marker was used). Arrows 1 and 2 point to cell undergoing mitosis. Arrow 1 points to the beginning of a mitosis, characterized by a rounded shape cell with a bright halo, arrow 2 points two the “8” shaped cells occuring when both daughter cells split from each other (end of the mitosis). In order to detect these events, we adapted our previously described cell tracking method [6], used in the back to front direction, to tackle the mitosis detection. After this introduction, we will describe in part II the biological material used and the acquistion system, in part III we briefly recall the method used for celltracking and we describe the track combination that permits mitosis detection, part IV gives some obtained results and we give conclusions and future works in part V. Mitotic Tree Construction by Computer In Vitro Cell Tracking : a Tool for Proliferation and Motility Features Extraction O. Debeir, D. Milojevic, T. Leloup, P. Van Ham, R. Kiss and C. Decaestecker I Fig 1 Phase contrast microscopy image of in vitro U373 cell line 1. mitose begin (strong halo, round shaped) 2. end of a mitose (8 shaped cells) 1-4244-0049-X/05/$20.00 ©2005 IEEE 951