RESUMEN
A large percentage of the world's population is affected by gastric diseases ranging from erosion and ulcer to serious ailments such as gastric cancer, which is mainly caused by Helicobacter pylori(H.pylori) infection. While most erosions and ulcers are benign, severe cases of gastric diseases can still develop into cancer. Thus, early screening and treatment of all gastric diseases are of great importance. Upper gastroscopy is one such common screening procedure that visualizes the patient's upper digestive system by inserting a camera attached to a rubber tube down the patient's digestive tracts, but since the procedure requires manual inspection of the video feed, it is prone to human errors. To improve the sensitivity and specificity of gastroscopies, we applied deep learning methods to develop an automated gastric disease detection system that detects frames of the video feed showing signs of gastric diseases. To this end, we collected data from images in anonymous patient case reports and gastroscopy videos to train and evaluate a convolutional neural network (CNN), and we used sliding window to improve the stability of our model's video performance. Our CNN model achieved 84.92% sensitivity, 88.26% specificity, and 85.2% F1-score on the test set, as well as 97% true positive rate and 16.2% false positive rate on a separate video test set.
RESUMEN
OBJECTIVE: Physicians' fears of being sued may lead to defensive medical practices, such as ordering nonindicated medical imaging. We investigated the association between states' medical malpractice tort reforms and neurologic imaging rates for patients seen in the emergency department with mild head trauma. METHODS: We assessed neurologic imaging among a national sample of 8588 women residing in 10 US states evaluated in an emergency setting for head injury between January 1, 1992, and December 31, 2001. We assessed the odds of imaging as it varied by the enactment of medical liability reform laws. RESULTS: The medical liability reform laws were significantly associated with the likelihood of imaging. States with laws that limited monetary damages (odds ratio [OR], 0.63; 95% confidence interval [CI], 0.40-0.99), mandated periodic award payments (OR, 0.64; 95% CI, 0.43-0.97), or specified collateral source offset rules (OR, 0.62; 95% CI, 0.40-0.96) had an approximately 40% lower odds of imaging, whereas states that had laws that limited attorney's contingency fees had significantly higher odds of imaging (OR, 1.5; 95% CI, 0.99-2.4), compared to states without these laws. When we used a summation of the number of laws in place, the greater the number of laws, the lower the odds of imaging. In the multivariate analysis, after adjusting for individual and community factors, the total number of laws remained significantly associated with the odds of imaging, and the effect of the individual laws was attenuated, but not eliminated. CONCLUSION: The tort reforms we examined were associated with the propensity to obtain neurologic imaging. If these results are confirmed in larger studies, tort reform might mitigate defensive medical practices.
