MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641109010 A) filed by Mlr Institute Of Technology; and Marri Laxman Reddy Institute Of Technology And Management on September 10, 2026, for Machine Learning Based Lung Cancer Detection.
Inventors include Mrs. N. Thulasi Chitra; Ms. Vasamsetti Akhila; Mr. Uppala Sai Teja; and Mr. Md. Ehsaan Uddin.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: Lung cancer still ranks as one of the most significant public health challenges worldwide and it is a major cause of death. Detecting lung diseases at an early stage and appropriately classifying them are very important factors in increasing the chances of survival and the positive results of therapy. But, due to the slight signs of early stages and the complicated form of cancerous tissues in CT or MRI scans, it is usually very difficult to find lung cancers at an early stage. The silent progression of the disease in the early stages is a major challenge for good management as it leads to difficulties in timely diagnosis and treatment of individuals. Recent advances in computer-aided diagnosis (CAD) systems have significantly increased the accuracy of diagnoses, especially through a synergy of machine learning, deep learning, and highly sophisticated image processing techniques. Many methods were published for the classification and prediction of lung cancer malignancies by using these technological advancements. In this work, an extensive investigation and systematic review are utilized. The goal is to assess the effectiveness of current methods, highlight their weaknesses, and even indicate the directions these methods could be improved in the future to provide more accurate diagnoses. The present invention uses Convolutional Neural Networks. It is able to scan MRI or CT images and make predictions on the spot with corresponding confidence levels, and it can also record diagnostic reports and produce images pinpointing the key areas visually. Preliminary results show a high level of accuracy in classification, with the Random Forest classifier standing out as the most effective one with a rate of 88. 5%. The modular architecture of this system makes it adaptable for future developments, such as linking with hospital information systems. 4 Claims and 1 Figures.
Disclaimer: Curated by HT Syndication.
