gyti.techpedia@sristi.org
Non-invasive, Point-of-care Diagnostic System For Early Detection Of Oral Cancer Using Digital Infrared Thermal Imaging
Project Description :

Worldwide oral cancer is the 6th most prevalent form of cancer while in india it is the most common form of cancer. the prevalence of oral cancer is predominantly high amongst the socio-economically challenged population of the developing countries. specifically, 66% of the total cases of oral carcinoma are reported from the developing countries. increase in risk factors amongst such group is primarily due to the lack of oral hygiene and poor oral habits like chewing betel-nuts, tobacco etc. the seriousness of the disease is intense because oral cancer can be considered as silent killer as it does not manifest prominent symptoms in nascent stage. unfortunately, major fraction of patient population is unable to avail diagnosis for oral cancer due to scarcity of state-of-the-art infrastructure and paucity of expert oral and maxilla facial pathologist. today experienced oral pathologist direct patients for biopsy which is the gold standard for oral cancer detection. but biopsy is invasive, thus patients are usually reluctant to go for this test. all these bottlenecks together pose hindrance to early medical attention towards the disease. patients diagnosed at the inception of the disease may have better diagnosis than those detected at advanced stage. specifically, five year survival rate for early detected cases is as high as 75% as compared to only 20% for patients diagnosed at more advanced stage of the disease. presently, there is dearth of computer aided automated point-of-care screening for oral cancer. thus, there is a compelling requirement of screening facility for oral cancer. our project is the pioneering effort towards automating oral cancer screening and detection using digital infrared thermal imaging (diti) . diti is already used as promising thermal imaging modality for many medical diagnostics applications like breast cancer detection because of its merit of being non-invasive, non-ionizing and radiation hazard free nature. however, naively using diti suffers from reporting too many false positives. in this project, we have used intelligent image processing and machine learning paradigms to cater and minimize false positives. the key idea of this project is to analyse the facial thermograms and to design intelligent learning algorithms to discriminate normal subjects from prospective patients thus providing an economical yet reliable smart health care solution for on-spot, non-invasive and non-ionizing modality technique oral cancer screening. our vision is to make technology easily accessible to general population of india. the image acquisition system is very simple and requires minimal image acquisition expertise. we have automated most of the image acquisition protocols and thus, even a high school senior can be appointed for the task of image acquisition. currently, our system achieves 86.12% accuracy in detecting precancerous patients and 85.42% in detecting malignant patients with appreciable consistency; we are improving our model further using fusion of both clinical and thermal features. with advent of this technology, a patient can get access to on-spot screening system without the need to find a specialized doctor. once this technology matures and after rounds of clinical trial, we envision to organize screening camps in remote villages of india such that this screening may help to detect potential malignant/precancerous subjects who can be forwarded to doctors for further in depth treatment. patients are usually hesitant towards invasive diagnostic procedure which causes delay in starting early medications. it is to be noted, that our technology inhibits patients from undergoing any time consuming invasive procedures and thus they will be more interested in being diagnosed. we see this technology as the next step towards completely automated oral cancer detection; an advancement which will play a pivotal role in reduction of mortality rate by systematic early stage detection.

 
Other Photos :
 

No Updates

 
Project Details :
  • Date : Dec 28,2016
  • Innovator : Manashi Chakraborty
  • Team Members : Dr. Sudipta Mukhopadhyay,Prof. Swapna Bannerjee,Dr. Sourav Mukhopadhyay,Manashi Chakraborty,Prof. Jay Gopal Ray,Dr. Santanu Patsa,Dr. Nishat Anjum
  • Guide Name : Dr. Sudipta Mukhopadhyay Prof. Swapna Banerjee Dr. Sourav Mukhopadhyay Prof. Jay Gopal Ray
  • College :
  • Submission Year : 2017
  • Category : Medical Science & Technology
Share Project :