Lung ultrasound as a tool for monitoring the interstitial changes in recently hospitalised patients with COVID-19 pneumonia - The COVIDLUS study.

Background Long-term respiratory effects can occur after COVID-19 pneumonia (CP). The COVID Lung Ultrasound Study (COVIDLUS) aimed to investigate the utility of serial lung ultrasound (LUS) to track functional and physiological recovery after hospitalisation in patients with CP. Methods Between April 2021 and April 2022, 21 patients were recruited at discharge (D0). LUS was performed on D0, day 41 (D41) and day 83 (D83). CT Thorax was performed on D83. Lymphocyte count, Ferritin, Lactate Dehydrogenase, Troponin, CRP, and D-dimers were measured at D0, D41 and D83. 6 minute walking test (6MWT) was performed on D83 and quality of life questionnaires and spirometry completed on D41 and D83. Results 19 subjects completed the study (10 males [52%]; mean age: 52 years [range:37-74]). 1 patient died. LUS scores were significantly higher at D0, compared to D41 and D83 (Mean score:10.9 [D0]/2.8 [D41]/1.5 [D83]; p < 0.0001). LUS scores correlated poorly with CT at D83 (Pearson r2 = 0.28). Mean lymphocyte counts were lower at D0 but increased at D41 and D83. Mean serum Ferritin was significantly lower at D41 and D83, as compared to D0. The mean 6MWT distance was 385 m (130-540 m). Quality of life measures did not differ at D41 and D83. Lung function increased between D41 and D83 with mean increase in FEV1 and FVC of 160 ml and 190 ml respectively. Conclusion LUS can monitor the early recovery of lung interstitial changes from CP. The utility of LUS to predict development of subsequent lung fibrosis post-COVID deserves further study.

A Role for Tumor Volume Assessment in Resectable Esophageal Cancer.

BACKGROUND: Esophageal cancer has a poor prognosis, and many patients undergoing surgery have a low chance of cure. Imaging studies suggest that tumor volume is prognostic. The study aimed to evaluate pathological tumor volume (PTV) as a prognostic variable in esophageal cancer. METHODS: This single-center cohort study included 283 patients who underwent esophageal cancer resections between 2000 and 2012. PTVs were obtained from pathological measurements using a validated volume formula. The prognostic value of PTV was analyzed using multivariable regression models, adjusting for age, tumor grade, tumor (T) stage, nodal stage, lymphovascular invasion, resection margin, resection type, and chemotherapy response, which provided hazard ratios (HRs) with 95 % confidence intervals (CIs). Primary outcomes were time to death and time to recurrence. Secondary outcomes were margin involvement and lymph node positivity. Correlation analysis was performed between imaging and PTVs. RESULTS: On unadjusted analysis, increasing PTV was associated with worse overall mortality (HR 2.30, 95 % CI 1.41-3.73) and disease recurrence (HR 1.87, 95 % CI 1.14-3.07). Adjusted analysis demonstrated worse overall mortality with increasing PTV but reached significance in only one subgroup (HR 1.70, 95 % CI 1.09-2.38). PTV was an independent predictor of margin involvement (OR 2.28, 95 % CI 1.02-5.13) and lymph node-positive status (OR 2.77, 95 % CI 1.23-6.28). Correlation analyses demonstrated significant positive correlation between computed tomography (CT) software and formula tumor volumes (r = 0.927, p < 0.0001), CT and positron emission tomography tumor volumes (r = 0.547, p < 0.0001), and CT and PTVs (r = 0.310, p < 0.001). CONCLUSIONS: Tumor volume may predict survival, margin status, and lymph node positivity after surgery for esophageal cancer.

Evaluation of state-of-the-art segmentation algorithms for left ventricle infarct from late Gadolinium enhancement MR images.

Studies have demonstrated the feasibility of late Gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) imaging for guiding the management of patients with sequelae to myocardial infarction, such as ventricular tachycardia and heart failure. Clinical implementation of these developments necessitates a reproducible and reliable segmentation of the infarcted regions. It is challenging to compare new algorithms for infarct segmentation in the left ventricle (LV) with existing algorithms. Benchmarking datasets with evaluation strategies are much needed to facilitate comparison. This manuscript presents a benchmarking evaluation framework for future algorithms that segment infarct from LGE CMR of the LV. The image database consists of 30 LGE CMR images of both humans and pigs that were acquired from two separate imaging centres. A consensus ground truth was obtained for all data using maximum likelihood estimation. Six widely-used fixed-thresholding methods and five recently developed algorithms are tested on the benchmarking framework. Results demonstrate that the algorithms have better overlap with the consensus ground truth than most of the n-SD fixed-thresholding methods, with the exception of the Full-Width-at-Half-Maximum (FWHM) fixed-thresholding method. Some of the pitfalls of fixed thresholding methods are demonstrated in this work. The benchmarking evaluation framework, which is a contribution of this work, can be used to test and benchmark future algorithms that detect and quantify infarct in LGE CMR images of the LV. The datasets, ground truth and evaluation code have been made publicly available through the website: https://www.cardiacatlas.org/web/guest/challenges.

Predicting Response to Neoadjuvant Chemotherapy with PET Imaging Using Convolutional Neural Networks.

Imaging of cancer with 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) has become a standard component of diagnosis and staging in oncology, and is becoming more important as a quantitative monitor of individual response to therapy. In this article we investigate the challenging problem of predicting a patient's response to neoadjuvant chemotherapy from a single 18F-FDG PET scan taken prior to treatment. We take a "radiomics" approach whereby a large amount of quantitative features is automatically extracted from pretherapy PET images in order to build a comprehensive quantification of the tumor phenotype. While the dominant methodology relies on hand-crafted texture features, we explore the potential of automatically learning low- to high-level features directly from PET scans. We report on a study that compares the performance of two competing radiomics strategies: an approach based on state-of-the-art statistical classifiers using over 100 quantitative imaging descriptors, including texture features as well as standardized uptake values, and a convolutional neural network, 3S-CNN, trained directly from PET scans by taking sets of adjacent intra-tumor slices. Our experimental results, based on a sample of 107 patients with esophageal cancer, provide initial evidence that convolutional neural networks have the potential to extract PET imaging representations that are highly predictive of response to therapy. On this dataset, 3S-CNN achieves an average 80.7% sensitivity and 81.6% specificity in predicting non-responders, and outperforms other competing predictive models.