RESUMO
Breast cancer is the second most commonly diagnosed type of cancer among women as of 2021. Grading of histopathological images is used to guide breast cancer treatment decisions and a critical component of this is a mitotic score, which is related to tumor aggressiveness. Manual mitosis counting is an extremely tedious manual task, but automated approaches can be used to overcome inefficiency and subjectivity. In this paper, we propose an automatic mitosis and nuclear segmentation method for a diverse set of H&E breast cancer pathology images. The method is based on a conditional generative adversarial network to segment both mitoses and nuclei at the same time. Architecture optimizations are investigated, including hyper parameters and the addition of a focal loss. The accuracy of the proposed method is investigated using images from multiple centers and scanners, including TUPAC16, ICPR14 and ICPR12 datasets. In TUPAC16, we use 618 carefully annotated images of size 256×256 scanned at 40×. TUPAC16 is used to train the model, and segmentation performance is measured on the test set for both nuclei and mitoses. Results on 200 held-out testing images from the TUPAC16 dataset were mean DSC = 0.784 and 0.721 for nuclear and mitosis, respectively. On 202 ICPR12 images, mitosis segmentation accuracy had a mean DSC = 0.782, indicating the model generalizes well to unseen datasets. For datasets that had mitosis centroid annotations, which included 200 TUPAC16, 202 ICPR12 and 524 ICPR14, a mean F1-score of 0.854 was found indicating high mitosis detection accuracy.
RESUMO
PURPOSE: to evaluate the current rate of pulmonary embolism (PE) in our medico-surgical intensive care unit (ICU), to identify risk factors, and to determine the outcome of PE in ICU. METHODS: We performed a prospective cohort study of consecutive patients requiring intensive care admission during a one-year period. We included, in this prospective study, all the patients with confirmed PE admitted in ICU with more than 18 years of age, and expected to stay in ICU for more than 48 hours. Only the patients who had a clinical suspicion (unexplained hypoxemia and/or shock) for PE underwent diagnostic studies. RESULTS: During the study period, 842 patients were admitted in our ICU. One hundred and two patients were excluded. The diagnosis of PE was confirmed in 75 patients (10.1%). In our study, all patients (100%) had received some forms of pharmaceutical prophylaxis (PP) during ICU stay. The median time from ICU admission to diagnosis of PE was 6 days. The diagnosis of PE was made by spiral CT in 74 patients (98.7%), and by echocardiography in 1 case (1.3%). The mean ICU stay was 26.3 ± 26.5 days (median: 20 days). During their ICU stay, 73 patients (97.3%) developed one, or more, organ failure. Respiratory failure was the most observed (97.3%). Moreover, 38 patients (50.6%) developed nosocomial infections and 29 (38.6%) died. The multivariate analysis showed that the risk factors associated with mortality were the presence of shock the day of PE diagnosis and the presence of right ventricular dilatation on echocardiography. CONCLUSION: Our findings confirm that subjects in the ICU are at high risk of PE, due to a high number of risk-factors. PE was associated with higher ICU mortality and a significantly higher ICU LOS. Our results invite to revise the preventive strategies of deep venous thrombosis and PE in patients requiring ICU admission.
