Clinical validation of the M5L lung computer-assisted detection system

Introduction: Lung cancer is one of the leading causes of death in the world. Early diagnosis is crucial to limit mortality. In order to support radiologists in the diagnosis several Computer-Assisted Detection (CAD) systems were developed. Materials and methods: The M5L CAD is based on a multi-thread approach: it combines the results of two independent algorithms, based on Voxel-Based Neural Analysis (VBNA) and on Virtual Ant Colonies (lungCAM), and provides a framework for further extension to others. Results: Having demonstrated the algorithm generalization capabilities (the training classifier procedure only used 69 of the 1018 LIDC Its), the development team tackled the issue of making it available to the largest possible user community. Therefore, a Web/Cloud prototype was designed and implemented: Its are uploaded through a Web front-end interface and analyzed by the cloud-backend at the INFN-Torino Computing Centre. Conclusions: Clinical validation on oncological patients undergoing staging or restating has recently started at IRCCS Candiolo, Italy. Preliminary results on few cases confirm a sensitivity of the order of 80% with about 3FP/scan.