IIIT-H Develops Smartphone-based Oral Cancer Screening Tool

This AI-backed smartphone model claims to offer a cost-effective and efficient solution for early detection, early intervention, and a higher chance of survival for individuals at risk of this devastating disease.

The International Institute of Information Technology (IIIT-Hyderabad) and INAI, an applied AI research centre supported by Intel, the Telangana Government, and the Public Health Foundation of India, have partnered to unveil an AI-powered smartphone approach to the early screening of oral cancer.

Leveraging the capabilities of AI and the pervasiveness of smartphones, this model promises the potential to detect abnormal lesions before they evolve into malignant conditions.

Oral cancer, often associated with the consumption of tobacco and areca nuts, is a prevalent concern in many parts of the world. Early intervention is essential to prevent the progression of oral potentially malignant disorders (OPMDs) into cancer, making early detection a critical component in the fight against this disease.

Traditional clinical assessments of oral health are subjective and often depend on the expertise of the healthcare professional. Biopsies, while definitive, are limited by the availability of experts and diagnostic centres, resulting in undetected lesions. This is where the AI-backed smartphone model comes into play, which claims to provide a cost-effective, consistent, and easily accessible solution.

One of the key features of this approach is its utilisation of smartphone cameras to capture high-quality images of the oral cavity. This eliminates the need for an external light source, ensuring that imaging consistency is maintained across different settings. Furthermore, the user-friendly nature of the system allows healthcare workers in the field to operate it with minimal training.

The research, titled 'AI-Assisted Screening of Potentially Malignant Disorders,' was commissioned by INAI and IHub Data, and it focused on exploring the efficacy of deep learning models in recognising OPMDs through images of oral cavities and lesions captured during community outreach programs for early cancer detection.

Led by Vivek Talwar and Pragya Singh, the research team achieved an 'F1 score' of over 70%. The F1 score is an indicator that measures the model's accuracy, and such a score demonstrates the model's potential for early detection, leading to a reduction in the treatment cycle and improved patient prognosis.

"Early detection would help in attending to the problem early. The treatment cycle reduction and improved prognosis can be achieved by early detection," an IIIT-H spokesperson said.

To accelerate the deployment of this solution, the team recommends further innovations in capturing high-quality images and data transfer to the app.

As technology continues to evolve, this solution is expected to play a pivotal role in capturing the incidence rate of oral cancer in community setups.

Konala Varma, CEO of INAI, expressed his optimism regarding the impact of this innovation, saying, "Innovation in the quality of screening and future deployment through a mobile app will help improve the capturing of oral cancer incidence rate in community setups."

With the potential to revolutionise oral cancer screening, this AI-backed smartphone model claims to offer a cost-effective and efficient solution but also brings hope for early detection, early intervention, and, ultimately, a higher chance of survival for individuals at risk of this devastating disease.