Mexican students create lung cancer detection system

Through image-processing and pattern recognition techniques, a group of Mexican students has developed a computer tool to carry out lung cancer pre-diagnoses at an early stage, according to Mexico’s National Polytechnic Institute (IPN) .

The system analyzes CT scan images of the chest and classifies them to conduct a pre-diagnosis that can help specialists in their clinical diagnoses. The young students used pattern recognition techniques ( artificial neural networks ) for the digital analysis of the scan images

Lung cancer is the second leading cause of death among Mexican men, while it’s the fourth cause of death among women.

The innovating system was created by Ximena Cortés, Isaac Aguirre, and Sergio Martínez Ávila , students of the IPN’s Superior Computer School.

For the development of the lung cancer detection system, the students worked with images of 1,400 patients obtained through the Lung Image Database Consortium Image Collection (LIDC-IDRI) from the United States.

They used the images to “train” a series of algorithms capable of differentiating between healthy and unhealthy patients.

Aguirre explained that they had used some of the images for the preparation of the algorithm, another for the evaluation process, which allows the system’s performance to be measured, and another fraction was used to conduct trials with images different from the ones used during the training process.

“ Our system reached a precision level of 95%, ” the student explained.

Meanwhile, Cortés explained that the interpretation of a thorax tomography was far from simple since said scans provide a great deal of information, which can make it difficult for a doctor to conduct a proper diagnose.

On the other hand, Martínez assured that the software tool was not meant to replace traditional methods or clinical trials, though it would prevent healthy patients from having to undergo lung biopsies or unnecessary invasive treatments.

“The program is capable of detecting even the smallest signs of cancer, which will allow specialists to conduct a biopsy only when needed,” he stated.

According to the U.S. National Cancer Institute , early cancer detection could help patients recover by up to 70% more than those diagnosed in late stages.

Mexican students create pantyliner that detects infections

dm