This paper performs an investigation into the development of an intelligent image-based automatic in vitro diagnostic system for infectious diseases using personal devices. The proposed framework of the image-based diagnostic system is demonstrated using the case study of Tuberculosis (TB)-specific antibody detection. The developed system, denoted as the iTB-test, is an intelligent bio-sensing system, comprised of a plasmonic Enzyme-Linked Immunosorbent Assay based colourimetric test in combination with an artificial intelligence-enabled image-based system. The presented system can separate the region of interest with 99.62% accuracy using clustering-based hybrid image processing algorithms, whereas the classification accuracy of antibody detection using a supervised machine learning technique is 100^% based on the experiments conducted for the case study.