Maximilian Schulze‐Hagen

Recent advances in computer vision have shown promising results in image generation. Diffusion probabilistic models have generated realistic images from textual input, as demonstrated by DALL-E 2, Imagen, and Stable Diffusion. However, their use in medicine, where imaging data typically comprises three-dimensional volumes, has not been systematically evaluated. Synthetic images may play a crucial role in privacy-preserving artificial intelligence and can also be used to augment small datasets. We show that diffusion probabilistic models can synthesize high-quality medical data for magnetic resonance imaging (MRI) and computed tomography (CT). For quantitative evaluation, two radiologists rated the quality of the synthesized images regarding "realistic image appearance", "anatomical correctness", and "consistency between slices". Furthermore, we demonstrate that synthetic images can be used in self-supervised pre-training and improve the performance of breast segmentation models when data is scarce (Dice scores, 0.91 [without synthetic data], 0.95 [with synthetic data]).

Artificial intelligence (AI) has a multitude of applications in cancer research and oncology. However, the training of AI systems is impeded by the limited availability of large datasets due to data protection requirements and other regulatory obstacles. Federated and swarm learning represent possible solutions to this problem by collaboratively training AI models while avoiding data transfer. However, in these decentralized methods, weight updates are still transferred to the aggregation server for merging the models. This leaves the possibility for a breach of data privacy, for example by model inversion or membership inference attacks by untrusted servers. Somewhat-homomorphically-encrypted federated learning (SHEFL) is a solution to this problem because only encrypted weights are transferred, and model updates are performed in the encrypted space. Here, we demonstrate the first successful implementation of SHEFL in a range of clinically relevant tasks in cancer image analysis on multicentric datasets in radiology and histopathology. We show that SHEFL enables the training of AI models which outperform locally trained models and perform on par with models which are centrally trained. In the future, SHEFL can enable multiple institutions to co-train AI models without forsaking data governance and without ever transmitting any decryptable data to untrusted servers.

Objectives: Parameters of patients’ body composition have been suggested as prognostic markers in several clinical conditions including cancer and liver transplantation, but only limited data on its value in critical illness exist to date. In this study, we aimed at evaluating a potential prognostic value of the skeletal muscle mass and skeletal muscle myosteatosis of critically ill patients at admission to the ICU. Design: Exploratory observational cohort study. Setting: An urban, academic medical institution. Patients: One-hundred fifty-five patients treated for critical illness on a medical ICU. Interventions: None. Measurements and Main Results: We used routine CT scans to assess the patients’ individual body composition. The skeletal muscle index as a surrogate for sarcopenia was defined as the total skeletal muscle area at the level of the third lumbar vertebra on axial CT scan, normalized for the patient’s height. Myosteatosis was evaluated by assessing the mean skeletal muscle attenuation measured in Hounsfield unit at the same sectional plane. The skeletal muscle index and mean skeletal muscle attenuation at admission to the ICU were significantly higher in patients with long-term survival (180-day or 1-year mortality), while both parameters were comparable between short-term survivors and nonsurvivors (ICU mortality or 30-d mortality). Patients with a skeletal muscle index or mean skeletal muscle attenuation below our established ideal cutoff values (74.95 mm 2 /cm and 29 Hounsfield unit) showed a significantly reduced overall survival. These findings were confirmed in univariate and multivariate Cox regression analyses. Furthermore, myosteatosis significantly correlated with the time of mechanical ventilation, the duration of hospital stay, and the presence of sepsis. Conclusions: Our data suggest that sarcopenia and myosteatosis represent important prognostic factors in critically ill patients that can be easily obtained from routine CT scans. Both parameters at admission to the ICU yield important information on the patients’ long-term outcome and might be used for early clinical decision-making in these patients.

Background and Aims: While transarterial chemoembolization (TACE) represents a standard of therapy for intermediate-stage hepatocellular carcinoma (HCC) and is also routinely performed in patients with liver metastases, it is still debated which patients represent the ideal candidates for TACE therapy in terms of overall survival. Sarcopenia, the degenerative loss of skeletal muscle mass and strength, has been associated with an adverse outcome for various malignancies, but its role in the context of TACE has largely remained unknown. Here, we evaluated the role of sarcopenia on the outcome of patients undergoing TACE for primary and secondary liver cancer. METHODS: The patients' psoas muscle size was measured on axial computed tomography (CT) scans and normalized for the patients' height squared. This value was referred to as the psoas muscle index (PMI). The PMI was correlated with clinical and laboratory markers. RESULTS: had a significantly impaired long-term outcome with a median overall survival of 491 days compared to 1291 days for patients with a high PMI. This finding was confirmed by uni- and multivariate Cox-regression analyses. Moreover, a progressive rapid decline in muscle mass after TACE was a predictor for an unfavorable prognosis. CONCLUSION: Our data suggest that sarcopenia represents a previously unrecognized prognostic factor for patients undergoing TACE therapy which might yield important information on the patients' post-interventional outcome and should therefore be implemented into clinical stratification algorithms.

Abstract Objectives The aim of this study was to compare success, technical complexity, and complication rates of percutaneous transhepatic biliary drainage (PTBD) in patients with dilated vs. nondilated bile ducts. Methods In a retrospective analysis, we evaluated all consecutive PTBD performed in our department over a period of 5 years. Technical success, technical data (side, fluoroscopy time, radiation dose, amount of contrast media, use of disposable equipment), procedure-related complications and peri-interventional mortality were compared for patients with dilated vs. non-dilated bile ducts. Independent t test and χ 2 test were used to evaluate the statistical significance. Results A total of 253 procedures were performed on 187 patients, of whom 101/253 had dilated bile ducts and 152/253 not. In total, 243/253 procedures were successful. PTBD was significantly more often successful in patients with dilated vs. nondilated bile ducts (150/153 vs. 93/101; p 0.02). Overall complication rate (13%) did not differ significantly between patients with dilated vs. nondilated bile ducts. Procedures in patients with normal, nondilated bile ducts were associated with a significantly higher rate of post-interventional bleeding (5/101 vs. 0/152). Mean fluoroscopy time (42:36 ± 35:39 h vs. 30:28 ± 25:10 h; p 0.002) and amount of contrast media (66 ± 40 ml vs. 52 ± 24 ml; p 0.07) or use of disposables were significantly higher in patients with nondilated ducts. A significantly lower fluoroscopy time and amount of contrast medium were used in left hepatic PTBD. Conclusion Despite the higher technical complexity, PTBD with nondilated bile ducts was associated with similar overall complication rates but higher bleeding complications compared with PTBD with dilated bile ducts. Key Points • PTBD was associated with similar overall complication rates in patients with dilated vs. nondilated bile ducts. • Although overall complication rates were low, PTBD in patients with nondilated bile ducts was associated with a higher incidence of post-interventional bleeding. • PTBD in patients with nondilated bile ducts is technically more complex.