Deep learning is revolutionizing various scientific fields, with medical applications at the forefront. One key focus is automating image segmentation, a process crucial in many clinical services. However, medical images are often ambiguous and challenging even for experts. To address this, reliable models need to quantify their uncertainty, allowing physicians to understand the model’s confidence in its segmentation. This paper explores how the performance and uncertainty of a model are influenced by the number of annotations per input sample. We examine the effects of both single and multiple manual annotations on various deep learning architectures. To tackle this question, we employ three widely recognized deep learning architectures and evaluate them across four publicly available datasets. Furthermore, we explore the effects of dropout rates on Monte Carlo models by examining uncertainty models with dropout rates of 20%, 40%, 60%, and 80%. Subsequently, we evaluate the models using various measurement metrics. The findings reveal that the influence of multiple annotations varies significantly depending on the datasets. Additionally, we observe that the dropout rate has minimal or no impact on the model’s performance unless there is a substantial loss of training data, primarily evident in the 80% dropout rate scenario.

This case concentrates on the persistent left superior vena cava (PLSVC), a rare vascular anomaly which contributes to central venous catheter (CVC) misplacement. A 72-year-old woman with renal insufficiency presented to the hospital with recurrent bleeding from her permanent CVC device placed in the right common jugular vein. An initial attempt to replace the device was unsuccessful, necessitating the placement of a secondary catheter in the left jugular vein. Shortly after the procedure, the patient developed swelling of the face and neck. Further diagnostic imaging, including a chest radiograph and computed tomography (CT), revealed CVC misplacement in the PLSVC and coronary sinus, thrombosis of the common jugular vein, and a posterior mediastinal hematoma. Conservative therapy of the mediastinal hematoma was implemented and proved effective in this case. A temporary CVC was inserted into the left femoral vein. Two months later, the catheter underwent further dysfunction and a decision was made to place a long-term permanent CVC via the right femoral vein. The patient is currently awaiting an arteriovenous fistula for dialysis use. This case emphasizes the importance of radiological techniques for CVC procedural placement, as well as the detection of congenital abnormalities. Providers regularly placing CVCs should have an in-depth knowledge of the possible complications and potential anatomical variations, especially as seen in high-risk patients.

The role of induction chemotherapy (iCHT) in locally advanced head and neck squamous cell carcinoma (LA-HNSCC) is still to be established due to high toxicity and variable response rates. The aim of this retrospective study is to use NMR-based serum metabolomics to predict the response rates to iCHT from the pretreatment samples. The studied group consisted of 46 LA-HNSCC patients treated with iCHT. The response to the treatment was evaluated by the clinical, fiberoptic, and radiological examinations made before and after iCHT. The proton nuclear magnetic resonance (1H NMR) serum spectra of the samples collected before iCHT were acquired with a 400 MHz spectrometer and were analyzed using multivariate and univariate statistical methods. A significant multivariate model was obtained only for the male patients. The treatment-responsive men with >75% primary tumor regression after iCHT showed pretreatment elevated levels of isoleucine, alanine, glycine, tyrosine, N-acetylcysteine, and the lipid compounds, as well as decreased levels of acetate, glutamate, formate, and ketone bodies compared to those who did not respond (regression of the primary tumor <75%). The results indicate that the nutritional status, capacity of the immune system, and the efficiency of metabolism related to protein synthesis may be prognostic factors for the response to induction chemotherapy in male HNSCC patients. However, larger studies are required that would validate the findings and could contribute to the development of more personalized treatment protocols for HNSCC patients.

Background: Cardiothoracic ratio (CTR) and transverse cardiac diameter (TCD) are parameters that are used to assess cardiac size on chest radiographs (CXRs). We aimed to investigate the performance and efficiency of artificial intelligence (AI) in screening for cardiomegaly on CXRs. Methods: The U-net architecture was designed for lung and heart segmentation. The CTR and TCD were then calculated using these labels and a mathematical algorithm. For the training set, we retrospectively included 65 randomly selected patients who underwent CXRs, while for the testing set, we chose 50 patients who underwent cardiac magnetic resonance (CMR) imaging and had available CXRs in the medical documentation. Results: Using U-net for the training set, the Dice coefficient for the lung was 0.984 ± 0.003 (min. 0.977), while for the heart it was 0.983 ± 0.004 (min. 0.972). For the testing set, the Dice coefficient for the lung was 0.970 ± 0.012 (min. 0.926), while for the heart it was 0.950 ± 0.021 (min. 0.871). The mean CTR and TCD measurements were slightly greater when calculated from either manual or automated segmentation than when manually read. Receiver operating characteristic analyses showed that both the CTR and TCD measurements calculated from either manual or automated segmentation, or when manually read, were good predictors of cardiomegaly diagnosed in CMR. However, McNemar tests have shown that diagnoses made with TCD, rather than CTR, were more consistent with CMR diagnoses. According to a different definition of cardiomegaly based on CMR imaging, accuracy for CTR measurements ranged from 62.0 to 74.0% for automatic segmentation (for TCD it ranged from 64.0 to 72.0%). Conclusion: The use of AI may optimize the screening process for cardiomegaly on CXRs. Future studies should focus on improving the accuracy of AI algorithms and on assessing the usefulness both of CTR and TCD measurements in screening for cardiomegaly.