Convolutional neural network deep learning model accurately detects rectal cancer in endoanal ultrasounds.

BACKGROUND: Imaging is vital for assessing rectal cancer, with endoanal ultrasound (EAUS) being highly accurate in large tertiary medical centers. However, EAUS accuracy drops outside such settings, possibly due to varied examiner experience and fewer examinations. This underscores the need for an AI-based system to enhance accuracy in non-specialized centers. This study aimed to develop and validate deep learning (DL) models to differentiate rectal cancer in standard EAUS images. METHODS: A transfer learning approach with fine-tuned DL architectures was employed, utilizing a dataset of 294 images. The performance of DL models was assessed through a tenfold cross-validation. RESULTS: The DL diagnostics model exhibited a sensitivity and accuracy of 0.78 each. In the identification phase, the automatic diagnostic platform achieved an area under the curve performance of 0.85 for diagnosing rectal cancer. CONCLUSIONS: This research demonstrates the potential of DL models in enhancing rectal cancer detection during EAUS, especially in settings with lower examiner experience. The achieved sensitivity and accuracy suggest the viability of incorporating AI support for improved diagnostic outcomes in non-specialized medical centers.

Preliminary study of thermal density distribution and entropy analysis during cycling exercise stress test using infrared thermography.

Considerable differences related to the results of temperature changes acquired during exercise exist, and in many cases, these lead to poor correlation with physiological variables. In this preliminary study we investigated the temperature changes and the temperature distribution (entropy) of the torso during a graded cycling exercise stress test using thermal imaging and studied the correlation between the increase in pulmonary ventilation (VE) and the changes in the surface temperature of the anterior torso during exercise. Thermal images of the anterior torso were captured every 30 s during the exercise, while the resistance was gradually increased every minute until exhaustion. The thermal images were processed to obtain a mean temperature in the regions of interest (ROI) (chest, forehead, and abdomen). We also developed an algorithm to calculate the distribution of temperature and texture (entropy) within each ROI. No changes were found in absolute temperatures. However, the entropy of the chest surface area increased significantly throughout the exercise test, compared with baseline temperature at rest. This increase in entropy was significantly correlated with exercise duration and intensity (p < 0.001). Furthermore, a high correlation between the increase in VE and chest entropy during exercise was detected (r = 0.9515). No correlations were found between the increase in entropy and the abdomen or the forehead compared with the VE. The non-invasive IR thermal imaging during graded exercise, combined with advanced image processing, successfully correlates surface thermography with exercise duration and pulmonary ventilation.