Deep learning assists detection of esophageal cancer and precursor lesions in a prospective, randomized controlled study.

Endoscopy is the primary modality for detecting asymptomatic esophageal squamous cell carcinoma (ESCC) and precancerous lesions. Improving detection rate remains challenging. We developed a system based on deep convolutional neural networks (CNNs) for detecting esophageal cancer and precancerous lesions [high-risk esophageal lesions (HrELs)] and validated its efficacy in improving HrEL detection rate in clinical practice (trial registration ChiCTR2100044126 at www.chictr.org.cn). Between April 2021 and March 2022, 3117 patients ≥50 years old were consecutively recruited from Taizhou Hospital, Zhejiang Province, and randomly assigned 1:1 to an experimental group (CNN-assisted endoscopy) or a control group (unassisted endoscopy) based on block randomization. The primary endpoint was the HrEL detection rate. In the intention-to-treat population, the HrEL detection rate [28 of 1556 (1.8%)] was significantly higher in the experimental group than in the control group [14 of 1561 (0.9%), P = 0.029], and the experimental group detection rate was twice that of the control group. Similar findings were observed between the experimental and control groups [28 of 1524 (1.9%) versus 13 of 1534 (0.9%), respectively; P = 0.021]. The system's sensitivity, specificity, and accuracy for detecting HrELs were 89.7, 98.5, and 98.2%, respectively. No adverse events occurred. The proposed system thus improved HrEL detection rate during endoscopy and was safe. Deep learning assistance may enhance early diagnosis and treatment of esophageal cancer and may become a useful tool for esophageal cancer screening.

Prediction of Re-positivity for Coronavirus Nucleic Acid Among COVID-19 Patients in the Recovery Phase.

Background and Objectives: Although the pathogenesis and treatment of coronavirus disease 2019 (COVID-19) have been gradually revealed, the risk for re-emergence of coronavirus nucleic acids in recovered patients remains poorly understood. Hence, this study evaluated the risk predictors associated with re-positivity for virus nucleic acid. Methods: Between February 1 and March 20, 2020, we retrospectively reviewed the clinical epidemiological data of 129 COVID-19 patients who were treated at Zhongxiang People's Hospital of Hubei Province in China. Subsequently, a risk prediction model for the re-positivity of virus nucleic acid was developed, and a receiver operating characteristic (ROC) curve was drawn for further validation. Results: In this study, the rate of re-positivity for virus nucleic acid was 17.8% (23/129) where all re-positivity cases were asymptomatic. The median time interval from discharge to nucleic acid re-positivity to discharge after being cured again was 11.5 days (range: 7-23 days). Multivariate logistic regression analysis showed that leukocytopenia [odds ratio (OR) 7.316, 95% confidence interval (CI) 2.319-23.080, p = 0.001], prealbumin < 150 mg/L (OR 4.199, 95% CI 1.461-12.071, p = 0.008), and hyperpyrexia (body temperature >39°C, OR 4.643, 95% CI 1.426-15.117, p = 0.011) were independent risk factors associated with re-positivity. The area under the ROC curve was 0.815 (95% CI, 0.729-0.902). Conclusion: COVID-19 patients with leukocytopenia, low prealbumin level, and hyperpyrexia are more likely to test positive for virus nucleic acid after discharge. Timely and effective treatment and appropriate extension of hospital stays and quarantine periods may be feasible strategies for managing such patients.

Progress of esophageal stricture prevention after endoscopic submucosal dissection by regenerative medicine and tissue engineering.

Endoscopic submucosal dissection (ESD) has been widely accepted as an effective treatment for early esophageal cancer. However, post-ESD esophageal stricture remains a thorny issue. We herein review many strategies for preventing post-ESD esophageal stricture, as well as discuss their strengths and weaknesses. These strategies include pharmacological prophylaxis, esophageal stent and tissue engineering and regenerative medicine treatment. In this review, we summarize these studies and discuss the underlying progress and future directions of tissue engineering and regenerative medicine treatment.

Flat field concave holographic grating with broad spectral region and moderately high resolution.

In order to deal with the conflicts between broad spectral region and high resolution in compact spectrometers based on a flat field concave holographic grating and line array CCD, we present a simple and practical method to design a flat field concave holographic grating that is capable of imaging a broad spectral region at a moderately high resolution. First, we discuss the principle of realizing a broad spectral region and moderately high resolution. Second, we provide the practical method to realize our ideas, in which Namioka grating theory, a genetic algorithm, and ZEMAX are used to reach this purpose. Finally, a near-normal-incidence example modeled in ZEMAX is shown to verify our ideas. The results show that our work probably has a general applicability in compact spectrometers with a broad spectral region and moderately high resolution.

Dwell time algorithm in ion beam figuring.

To get a high-precision optical surface, the deconvolved process of dwell time was transferred to a matrix equation in which the damped factor and the extra removal amount were introduced to expand the freedom of solution. A path weight factor and a surface error weight factor were used to take the scanning path and the initial surface error into account. Combined with the Gerchberg bandlimited extrapolation algorithm for initial surface error map extension, a high-precision final surface could be obtained within a factual aperture. Two surface error maps were calculated to rms = 0.1 nm from rms = 130.23 nm and to rms = 0.08 nm from rms = 82.74 nm. The simulations show that a perfect dwell time solution could be obtained by the revised matrix equation and initial surface error map extension with the help of the least squares QR (LSQR) algorithm.