Lucknow: Researchers at Lucknow University’s computer science department developed an artificial intelligence (AI)-based tool to help doctors diagnose throat cancer at an early stage. Researchers say the technology could make diagnosis faster, more accurate and affordable, especially in places where oncologists are not easily available.The research was led by LU computer centre incharge Prof Puneet Misra along with research scholar Mohammad Usman, Siddharth Chaurasia of Pragnanix Labs and Ginika Mahajan of Manipal University Jaipur.The study was published in ‘Scientific Reports’, an international journal published under the Nature Portfolio.The AI model, called MedLK-SCNet, studies images of the throat taken during endoscopy. It also uses Narrow Band Imaging (NBI), a special imaging technique that helps doctors see tiny blood vessels more clearly. By carefully analysing these images, the AI can identify signs of cancer that may be difficult for the human eye to spot in the early stages.Researchers said throat cancer is one of the most serious cancers affecting the voice box. It is common among people who use tobacco products such as cigarettes, ‘beedi’ and ‘gutkha’. Detecting the disease early is often difficult, and delays in diagnosis can make treatment harder.“Our aim was to create technology that can reach people beyond laboratories. MedLK-SCNet can support doctors in finding throat cancer earlier, making diagnosis quicker, more accurate and less expensive, especially in rural and underserved areas,” said Misra.Researchers believe the tool could help save lives by finding cancer earlier, reduce need for repeated tests, support healthcare workers in smaller hospitals and make advanced cancer diagnosis available to more people across the country.
