Predicting death or recurrence of portal hypertension symptoms after TIPS procedures.

BACKGROUND: Accurate prediction of portal hypertension recurrence after transjugular intrahepatic portosystemic shunt (TIPS) placement will improve clinical decision-making. PURPOSE: To evaluate if perioperative variables could predict disease-free survival (DFS) in cirrhotic patients with portal hypertension (PHT) treated with TIPS. MATERIALS AND METHODS: We recruited 206 cirrhotic patients with PHT treated with TIPS, randomly assigned to training (n = 138) and validation (n = 68) sets. We recorded 7 epidemiological, 4 clinical, and 9 radiological variables. TIPS-distal end positioning (TIPS-DEP) measured the distance between the distal end of the stent and the hepatocaval junction on contrast-enhanced CT scans. In the training set, the signature was defined as the random forest for survival algorithm achieving the lowest error rate for the prediction of DFS which was landmarked 4 weeks after the TIPS procedure. In the training set, a simple to use scoring system was derived from variables selected by the signature. The primary endpoint was to assess if TIPS-DEP was associated with DFS. The secondary endpoint was to validate the scoring system in the validation set. RESULTS: Overall, patients with TIPS-DEP ≥ 6 mm (n = 49) had a median DFS of 24.5 months vs. 72.8 months otherwise (n = 157, p = 0.004). In the training set, the scoring system was calculated by adding age ≥ 60 years old, Child-Pugh B or C, and TIPS-DEP ≥ 6 mm (1 point each) since the signature showed high DFS probability at 6.5 months post-landmark in patients that did not meet these criteria: 86%, 80%, and 78%, respectively. The hazard ratio [95 CI] between patients determined to be low-risk (< 2 points) and high-risk (≥ 2 points) was 2.30 [1.35-3.93] (p = 0.002) in the training set and 2.01 [0.94-4.32] (p = 0.072) in the validation set. CONCLUSION: TIPS-DEP is an actionable radiological biomarker which can be combined with age and Child-Pugh score to predict death or PHT symptom recurrence after TIPS procedure. KEY POINTS: • TIPS-DEP measurement was the third most important but only actionable variable for predicting DFS. • TIPS-DEP < 6 mm was associated with a DFS probability of 78% at 6.5 months post-landmark. • A simple scoring system calculated using age, Child-Pugh score, and TIPS-DEP predicted DFS after TIPS.

Investigating of the role of CT scan for cancer patients during the first wave of COVID-19 pandemic.

Introduction: Amidst this current COVID-19 pandemic, we undertook this systematic review to determine the role of medical imaging, with a special emphasis on computed tomography (CT), on guiding the care and management of oncologic patients. Material and Methods: Study selection focused on articles from 01/02/2020 to 04/23/2020. After removal of irrelevant articles, all systematic or non-systematic reviews, comments, correspondence, editorials, guidelines and meta-analysis and case reports with less than 5 patients were also excluded. Full-text articles of eligible publications were reviewed to select all imaging-based publications, and the existence or not of an oncologic population was reported for each publication. Two independent reviewers collected the following information: ( 1) General publication data; (2) Study design characteristics; (3) Demographic, clinical and pathological variables with percentage of cancer patients if available; (4) Imaging performances. The sensitivity and specificity of chest CT (C-CT) were pooled separately using a random-effects model. The positive predictive value (PPV) and negative predictive value (NPV) of C-CT as a test was estimated for a wide range of disease prevalence rates. Results: A total of 106 publications were fully reviewed. Among them, 96 were identified to have extractable data for a two-by-two contingency table for CT performance. At the end, 53 studies (including 6 that used two different populations) were included in diagnosis accuracy analysis (N = 59). We identified 53 studies totaling 11,352 patients for whom the sensitivity (95CI) was 0.886 (0.880; 0.894), while specificity remained low: in 93% of cases (55/59), specificity was ≤ 0.5. Among all the 106 reviewed studies, only 7 studies included oncologic patients and were included in the final analysis for C-CT performances. The percentage of patients with cancer in these studies was 0.3% (34/11352 patients), lower than the global prevalence of cancer. Among all these studies, only 1 (0.9%, 1/106) reported performance specifically in a cohort of cancer patients, but it however only reported true positives. Discussion: There is a concerning lack of COVID-19 studies involving oncologic patients, showing there is a real need for further investigation and evaluation of the performance of the different medical imaging modalities in this specific patient population.

Toward Generalizability in the Deployment of Artificial Intelligence in Radiology: Role of Computation Stress Testing to Overcome Underspecification.

The clinical deployment of artificial intelligence (AI) applications in medical imaging is perhaps the greatest challenge facing radiology in the next decade. One of the main obstacles to the incorporation of automated AI-based decision-making tools in medicine is the failure of models to generalize when deployed across institutions with heterogeneous populations and imaging protocols. The most well-understood pitfall in developing these AI models is overfitting, which has, in part, been overcome by optimizing training protocols. However, overfitting is not the only obstacle to the success and generalizability of AI. Underspecification is also a serious impediment that requires conceptual understanding and correction. It is well known that a single AI pipeline, with prescribed training and testing sets, can produce several models with various levels of generalizability. Underspecification defines the inability of the pipeline to identify whether these models have embedded the structure of the underlying system by using a test set independent of, but distributed identically, to the training set. An underspecified pipeline is unable to assess the degree to which the models will be generalizable. Stress testing is a known tool in AI that can limit underspecification and, importantly, assure broad generalizability of AI models. However, the application of stress tests is new in radiologic applications. This report describes the concept of underspecification from a radiologist perspective, discusses stress testing as a specific strategy to overcome underspecification, and explains how stress tests could be designed in radiology-by modifying medical images or stratifying testing datasets. In the upcoming years, stress tests should become in radiology the standard that crash tests have become in the automotive industry. Keywords: Computer Applications-General, Informatics, Computer-aided Diagnosis © RSNA, 2021.

Updated Trends in Imaging Practices for Pancreatic Neuroendocrine Tumors (PNETs): A Systematic Review and Meta-Analysis to Pave the Way for Standardization in the New Era of Big Data and Artificial Intelligence.

PURPOSE: Medical imaging plays a central and decisive role in guiding the management of patients with pancreatic neuroendocrine tumors (PNETs). Our aim was to synthesize all recent literature of PNETs, enabling a comparison of all imaging practices. METHODS: based on a systematic review and meta-analysis approach, we collected; using MEDLINE, EMBASE, and Cochrane Library databases; all recent imaging-based studies, published from December 2014 to December 2019. Study quality assessment was performed by QUADAS-2 and MINORS tools. RESULTS: 161 studies consisting of 19852 patients were included. There were 63 'imaging' studies evaluating the accuracy of medical imaging, and 98 'clinical' studies using medical imaging as a tool for response assessment. A wide heterogeneity of practices was demonstrated: imaging modalities were: CT (57.1%, n=92), MR (42.9%, n=69), PET/CT (13.3%, n=31), and SPECT/CT (9.3%, n=15). International imaging guidelines were mentioned in 2.5% (n=4/161) of studies. In clinical studies, imaging protocol was not mentioned in 30.6% (n=30/98) of cases and only mentioned imaging modality without further information in 63.3% (n=62/98), as compared to imaging studies (1.6% (n=1/63) of (p<0.001)). QUADAS-2 and MINORS tools deciphered existing biases in the current literature. CONCLUSION: We provide an overview of the updated current trends in use of medical imaging for diagnosis and response assessment in PNETs. The most commonly used imaging modalities are anatomical (CT and MRI), followed by PET/CT and SPECT/CT. Therefore, standardization and homogenization of PNETs imaging practices is needed to aggregate data and leverage a big data approach for Artificial Intelligence purposes.