Portal de Periódicos da CAPES

HIV Lipodystrophy Case Definition using Artificial Neural Network Modelling

2003; SAGE Publishing; Volume: 8; Issue: 5 Linguagem: Inglês

10.1177/135965350300800511

2040-2058

John P. A. Ioannidis, Thomas A Trikalinos, Matthew Law, Andrew Carr, Andrew Carr, Dale J. Barr, DA Cooper, Sean Emery, Steven Grinspoon, John P. A. Ioannidis, R. Lewis, Matthew Law, Kenneth Lichtenstein, Justine Murray, Daniela Pizzuti, William G. Powderly, Willy Rozenbaum, Morris Schambelan, Rebekah Puls, Sean Emery, Antonia L. Moore, J. Philip Miller, Andrew Carr, WH Belloso, SA Ivalo, LO Clara, LA Barcan, LD Stern, AM Galich, MI Perman, Marcelo Losso, Adriana Durán, Javier Toibaro, David Baker, R D Vale, Robert M. McFarlane, Haley MacLeod, J Kidd, B Genn, Andrew Carr, Robert J Fielden, S. Mallal, Martyn A. French, Alan Cain, J. Skett, Donald M. Maxwell, Anne Mijch, Jennifer Hoy, Anna Pierce, C Mccormick, B. de Graaf, Julian Falutz, J Vatistas, L Dion, Julio Montaner, Marianne Harris, Peter Phillips, V. Montessori, Monica Valyi, Walter F. Stewart, Sharon Walmsley, L Casciaro, Jens Lundgren, Ove Andersen, A Gronholdt, I. Béguinot, P. Mercié, Geneviève Chêne, J. Reynes, Laurent Cotte, Willy Rozenbaum, Lella Naït‐Ighil, Laurence Slama, TH Nguyen, Christophe Rousselle, Viard Jp, Laurent Roudière, Aurélie Maignan, Marianne Burgard, Stefan Mauss, Guenther Schmutz, Saskia Scholten, Shinichi Oka, Hamish Fraser, Masayuki Ishihara, Kazuko Itoh, P Reiss, Marc van der Valk, P Leunissen, M A F Nievaard, Berthe van Eck-Smit, C can Kujik, Nicholas I. Paton, Béatrice Peperstraete, Farina Karim, Chamroeun Khim, Sean Wei Xiang Ong, José M. Gatell, Estebán Martínez, Ana Milinkovic, Duncan Churchill, C Timaeus, Toby M. Maher, Nancy E. Perry, A. J. Bray, Graeme Moyle, Christine Baldwin, Christopher Higgs, B Reynolds, Charles C. J. Carpenter, Linda Bausserman, T Fiore, Melissa DiSpigno, C Cohen, James Hellinger, Karlissa Foy, S Hubka, Brett E. Riccio, Wafaa El‐Sadr, S. Raghavan, N Chowdury, Bert de Vries, Stephen P. Miller, S. Hammer, Matthew M. Crawford, Stanley Chang, J Dobkin, Bianca Quagliarello, Dympna Gallagher, Mark Punyanitya, Harold A. Kessler, A. Tenorio, Siri L. Kjos, Judith Falloon, HC Lane, Donald Rock, Linda A. Ehler, Kenneth Lichtenstein, Todd McClain, Robert L. Murphy, P Milne, William G. Powderly, Judith A. Aberg, Michael K. Klebert, Michael Conklin, Derek Ward, L Green, B Stearn,

HIV/AIDS drug development and treatment

A case definition of HIV lipodystrophy has recently been developed from a combination of clinical, metabolic and imaging/body composition variables using logistic regression methods. We aimed to evaluate whether artificial neural networks could improve the diagnostic accuracy.The database of the case-control Lipodystrophy Case Definition Study was split into 504 subjects (265 with and 239 without lipodystrophy) used for training and 284 independent subjects (152 with and 132 without lipodystrophy) used for validation. Back-propagation neural networks with one or two middle layers were trained and validated. Results were compared against logistic regression models using the same information.Neural networks using clinical variables only (41 items) achieved consistently superior performance than logistic regression in terms of specificity, overall accuracy and area under the ROC curve. Their average sensitivity and specificity were 72.4 and 71.2%, as compared with 73.0 and 62.9% for logistic regression, respectively (area under the ROC curve, 0.784 vs 0.748). The discriminating performance of the neural networks was largely unaffected when built excluding 13 parameters that patients may not have readily available. The average sensitivity and specificity of the neural networks remained the same when metabolic variables were also considered (total 60 items) without a clear advantage against logistic regression (overall accuracy 71.8%). The performance of networks considering also body composition variables was similar to that of logistic regression (overall accuracy 78.5% for both).Neural networks may offer a means to improve the discriminating performance for HIV lipodystrophy, when only clinical data are available and a rapid approximate diagnostic decision is needed. In this context, information on metabolic parameters is apparently not helpful in improving the diagnosis of HIV lipodystrophy, unless imaging and body composition studies are also obtained.