Post-mortem computer tomography in ten cases of death while diving: a retrospective evaluation

Machine Learning and Radiomics Analysis for Tumor Budding Prediction in Colorectal Liver Metastases Magnetic Resonance Imaging Assessment

PURPOSE

We aimed to assess the efficacy of machine learning and radiomics analysis using magnetic resonance imaging (MRI) with a hepatospecific contrast agent, in a pre-surgical setting, to predict tumor budding in liver metastases.

METHODS

Patients with MRI in a pre-surgical setting were retrospectively enrolled. Manual segmentation was made by means 3D Slicer image computing, and 851 radiomics features were extracted as median values using the PyRadiomics Python package. Balancing was performed and inter- and intraclass correlation coefficients were calculated to assess the between observer and within observer reproducibility of all radiomics extracted features. A Wilcoxon-Mann-Whitney nonparametric test and receiver operating characteristics (ROC) analysis were carried out. Balancing and feature selection procedures were performed. Linear and non-logistic regression models (LRM and NLRM) and different machine learning-based classifiers including decision tree (DT), k-nearest neighbor (KNN) and support vector machine (SVM) were considered.

RESULTS

The internal training set included 49 patients and 119 liver metastases. The validation cohort consisted of a total of 28 single lesion patients. The best single predictor to classify tumor budding was original_glcm_Idn obtained in the T1-W VIBE sequence arterial phase with an accuracy of 84%; wavelet_LLH_firstorder_10Percentile was obtained in the T1-W VIBE sequence portal phase with an accuracy of 92%; wavelet_HHL_glcm_MaximumProbability was obtained in the T1-W VIBE sequence hepatobiliary excretion phase with an accuracy of 88%; and wavelet_LLH_glcm_Imc1 was obtained in T2-W SPACE sequences with an accuracy of 88%. Considering the linear regression analysis, a statistically significant increase in accuracy to 96% was obtained using a linear weighted combination of 13 radiomic features extracted from the T1-W VIBE sequence arterial phase. Moreover, the best classifier was a KNN trained with the 13 radiomic features extracted from the arterial phase of the T1-W VIBE sequence, obtaining an accuracy of 95% and an AUC of 0.96. The validation set reached an accuracy of 94%, a sensitivity of 86% and a specificity of 95%.

CONCLUSIONS

Machine learning and radiomics analysis are promising tools in predicting tumor budding. Considering the linear regression analysis, there was a statistically significant increase in accuracy to 96% using a weighted linear combination of 13 radiomics features extracted from the arterial phase compared to a single radiomics feature.

Autoptic Findings in Cases of Sudden Death Due to Kawasaki Disease

Kawasaki disease (KD) is the second-most-common childhood vasculitis, and its etiology is still unknown today. Even though the acute illness is usually self-limiting, sometimes, it can generate complications, such as coronary artery aneurysms (CAA), acute myocardial infarction (AMI), heart failure, or arrhythmias, and can rarely cause sudden or unexpected deaths. We present a review of the literature, which collects autoptic and histopathological data relating to many of the cases of these deaths. On the basis of the titles and abstracts, we selected 54 scientific publications for a total of 117 cases. Among them, as expected, the majority of the deaths were due to AMI (41.03%), arrhythmia (8.55%), acute coronary syndrome (8.55%), and CAA rupture (11.97%), involving mostly 20-year-olds or younger individuls (69.23%). This is not surprising since the CAs are the most involved arteries. Gross autoptic and histopathological findings are reported in the paper. Our work revealed that, when compared with the incidence of KD, only a few cases suffered from sudden death, underwent an autoptic examination, and were then described in the literature. We suggest that researchers should perform autopsies to gain a better understanding of the molecular pathways involved in KD so as to propose further innovative therapeutic protocols or implement more appropriate prevention schemes.

Black box of diving accidents: Contribution of forensic underwater experts to three fatal cases

Diving is a popular activity, largely practiced worldwide. Diving fatalities are not rare events, with drowning being the most common cause of death, followed by cardiac-related natural causes, immersion pulmonary edema and arterial gas embolism. In such cases, positive signs of drowning are not specific, depending also on the time of submersion of corpses. Moreover, drowning can be the terminal event. Over the years, measures to perform appropriate post-mortem examination in cases of diving fatalities were suggested, including the execution of post-mortem CT-scan, the use of a decompression chamber and the adoption of specific autoptic techniques. Although a multidisciplinary approach in forensic investigations concerning diving fatalities is discussed, poor cases focus on how the analysis of diving computer records and equipment can contribute to determining the cause of death. The present study shows how the cooperation between a forensic underwater expert and a forensic pathologist played a crucial role in interpreting radiological findings, guiding the autopsy and confirming/denying circumstantial data emerging from the investigations. Technical analysis of dive computer records and diving equipment is a fundamental step in the definition of the cause of death in diving fatalities. All diving computer data, not only those related to maximum depth and ascent's profile, should be considered in detail, and the immersion graph carefully studied by both the forensic pathologist and the forensic underwater experts. The diving technical data can often play a crucial role in explaining any legal issue related to the circumstances of death, possibly leading the prosecutor to further investigation.

Radiation Recall Pneumonitis: The Open Challenge in Differential Diagnosis of Pneumonia Induced by Oncological Treatments

The treatment of primary and secondary lung neoplasms now sees the fundamental role of radiotherapy, associated with surgery and systemic therapies. The improvement in survival outcomes has also increased attention to the quality of life, treatment compliance and the management of side effects. The role of imaging is not only limited to recognizing the efficacy of treatment but also to identifying, as soon as possible, the uncommon effects, especially when more treatments, such as chemotherapy, immunotherapy and radiotherapy, are associated. Radiation recall pneumonitis is an uncommon treatment complication that should be correctly characterized, and it is essential to recognize the mechanisms of radiation recall pneumonitis pathogenesis and diagnostic features in order to promptly identify them and adopt the best therapeutic strategy, with the shortest possible withdrawal of the current oncological drug. In this setting, artificial intelligence could have a critical role, although a larger patient data set is required.