A cost-efficient solution to improve the accuracy of radiotherapy given to lung cancer patients. radiotherapy is a common recourse for treating lung tumors, being considerably more localised than chemotherapy. it, however, faces the challenge of maintaining the focus of the radiation beam on the tumor. due to physiological factors like respiratory motion, peristaltic motion, cardiac motion, coughing etc, the tumor and its surrounding tissue inside the thorax or abdomen can move dynamically, with a displacement of upto 3 cm. the result of such movement, if ignored during radiotherapy, is the exposure of surrounding healthy tissue and organs, causing severe damage to them. our solution provides real-time surrogate visual tracking of lung tumours at all times during radiotherapy. to achieve this we perform simultaneous acquisition of two types of data, visual chest surface data and x-ray data of the tumour. to perform visual surface tracking we use an array of 4 cameras in conjunction with an adequate number of visible markers to capture the time evolution of the surface motion. the x-ray tracking of the tumour is performed using a ct scanner. using the time synchronised x-ray and visual surface tracking data sets, we have built a time and space correlation model of the surface of the chest with the x-ray tracker data using machine learning. the patient can be subjected to radiotherapy with beam steering (currently used in hospitals) directed by the surrogate measurements from the correlation model that we have developed. this provides a more accurate, and less expensive method to maintain the radiation focus than the standards used today. any innovation in the medical field demands extensive trials before adoption. we are seeking the cooperation of sgpgi, lucknow to provide the infrastructure and subjects for the calibration step.