AI-based medical e-diagnosis for fast and automatic ventricular volume measurement in patients with normal pressure hydrocephalus.

Based on CT and MRI images acquired from normal pressure hydrocephalus (NPH) patients, using machine learning methods, we aim to establish a multimodal and high-performance automatic ventricle segmentation method to achieve an efficient and accurate automatic measurement of the ventricular volume. First, we extract the brain CT and MRI images of 143 definite NPH patients. Second, we manually label the ventricular volume (VV) and intracranial volume (ICV). Then, we use the machine learning method to extract features and establish automatic ventricle segmentation model. Finally, we verify the reliability of the model and achieved automatic measurement of VV and ICV. In CT images, the Dice similarity coefficient (DSC), intraclass correlation coefficient (ICC), Pearson correlation, and Bland-Altman analysis of the automatic and manual segmentation result of the VV were 0.95, 0.99, 0.99, and 4.2 ± 2.6, respectively. The results of ICV were 0.96, 0.99, 0.99, and 6.0 ± 3.8, respectively. The whole process takes 3.4 ± 0.3 s. In MRI images, the DSC, ICC, Pearson correlation, and Bland-Altman analysis of the automatic and manual segmentation result of the VV were 0.94, 0.99, 0.99, and 2.0 ± 0.6, respectively. The results of ICV were 0.93, 0.99, 0.99, and 7.9 ± 3.8, respectively. The whole process took 1.9 ± 0.1 s. We have established a multimodal and high-performance automatic ventricle segmentation method to achieve efficient and accurate automatic measurement of the ventricular volume of NPH patients. This can help clinicians quickly and accurately understand the situation of NPH patient's ventricles.

Unbox the black-box for the medical explainable AI via multi-modal and multi-centre data fusion: A mini-review, two showcases and beyond.

Explainable Artificial Intelligence (XAI) is an emerging research topic of machine learning aimed at unboxing how AI systems' black-box choices are made. This research field inspects the measures and models involved in decision-making and seeks solutions to explain them explicitly. Many of the machine learning algorithms cannot manifest how and why a decision has been cast. This is particularly true of the most popular deep neural network approaches currently in use. Consequently, our confidence in AI systems can be hindered by the lack of explainability in these black-box models. The XAI becomes more and more crucial for deep learning powered applications, especially for medical and healthcare studies, although in general these deep neural networks can return an arresting dividend in performance. The insufficient explainability and transparency in most existing AI systems can be one of the major reasons that successful implementation and integration of AI tools into routine clinical practice are uncommon. In this study, we first surveyed the current progress of XAI and in particular its advances in healthcare applications. We then introduced our solutions for XAI leveraging multi-modal and multi-centre data fusion, and subsequently validated in two showcases following real clinical scenarios. Comprehensive quantitative and qualitative analyses can prove the efficacy of our proposed XAI solutions, from which we can envisage successful applications in a broader range of clinical questions.

Robust weakly supervised learning for COVID-19 recognition using multi-center CT images.

The world is currently experiencing an ongoing pandemic of an infectious disease named coronavirus disease 2019 (i.e., COVID-19), which is caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Computed Tomography (CT) plays an important role in assessing the severity of the infection and can also be used to identify those symptomatic and asymptomatic COVID-19 carriers. With a surge of the cumulative number of COVID-19 patients, radiologists are increasingly stressed to examine the CT scans manually. Therefore, an automated 3D CT scan recognition tool is highly in demand since the manual analysis is time-consuming for radiologists and their fatigue can cause possible misjudgment. However, due to various technical specifications of CT scanners located in different hospitals, the appearance of CT images can be significantly different leading to the failure of many automated image recognition approaches. The multi-domain shift problem for the multi-center and multi-scanner studies is therefore nontrivial that is also crucial for a dependable recognition and critical for reproducible and objective diagnosis and prognosis. In this paper, we proposed a COVID-19 CT scan recognition model namely coronavirus information fusion and diagnosis network (CIFD-Net) that can efficiently handle the multi-domain shift problem via a new robust weakly supervised learning paradigm. Our model can resolve the problem of different appearance in CT scan images reliably and efficiently while attaining higher accuracy compared to other state-of-the-art methods.

Can Clinical Symptoms and Laboratory Results Predict CT Abnormality? Initial Findings Using Novel Machine Learning Techniques in Children With COVID-19 Infections.

The rapid spread of coronavirus 2019 disease (COVID-19) has manifested a global public health crisis, and chest CT has been proven to be a powerful tool for screening, triage, evaluation and prognosis in COVID-19 patients. However, CT is not only costly but also associated with an increased incidence of cancer, in particular for children. This study will question whether clinical symptoms and laboratory results can predict the CT outcomes for the pediatric patients with positive RT-PCR testing results in order to determine the necessity of CT for such a vulnerable group. Clinical data were collected from 244 consecutive pediatric patients (16 years of age and under) treated at Wuhan Children's Hospital with positive RT-PCR testing, and the chest CT were performed within 3 days of clinical data collection, from January 21 to March 8, 2020. This study was approved by the local ethics committee of Wuhan Children's Hospital. Advanced decision tree based machine learning models were developed for the prediction of CT outcomes. Results have shown that age, lymphocyte, neutrophils, ferritin and C-reactive protein are the most related clinical indicators for predicting CT outcomes for pediatric patients with positive RT-PCR testing. Our decision support system has managed to achieve an AUC of 0.84 with 0.82 accuracy and 0.84 sensitivity for predicting CT outcomes. Our model can effectively predict CT outcomes, and our findings have indicated that the use of CT should be reconsidered for pediatric patients, as it may not be indispensable.

Systematic and Comprehensive Automated Ventricle Segmentation on Ventricle Images of the Elderly Patients: A Retrospective Study.

Background and Objective: Ventricle volume is closely related to hydrocephalus, brain atrophy, Alzheimer's, Parkinson's syndrome, and other diseases. To accurately measure the volume of the ventricles for elderly patients, we use deep learning to establish a systematic and comprehensive automated ventricle segmentation framework. Methods: The study participation included 20 normal elderly people, 20 patients with cerebral atrophy, 64 patients with normal pressure hydrocephalus, and 51 patients with acquired hydrocephalus. Second, get their imaging data through the picture archiving and communication systems (PACS) system. Then use ITK software to manually label participants' ventricular structures. Finally, extract imaging features through machine learning. Results: This automated ventricle segmentation method can be applied not only to CT and MRI images but also to images with different scan slice thicknesses. More importantly, it produces excellent segmentation results (Dice > 0.9). Conclusion: This automated ventricle segmentation method has wide applicability and clinical practicability. It can help clinicians find early disease, diagnose disease, understand the patient's disease progression, and evaluate the patient's treatment effect.

Study on the Method of Voigt Profiles Two Wings Fitting Non-Uniform Flow Field Absorbance.

In the field of the absorption spectrum, especially for direct tunable diode laser absorption spectroscopy (dTDLAS) technology, the integrated area of the absorption spectrum is needed to be measured accurately for calculating the temperature and the component concentration of the flow field. Doing single optical path absorption spectroscopic measurement in the non-uniform flow field, spectral lineshape broadening is varied with the flow changes, in previous research reports, researchers mainly use single Voigt or Lorentz profile to fit absorbance curve or use directly integral to obtain the integrated area of the absorption spectrum. There are some shortcomings in these methods, resulting in certain error between the fitting result and the actual area, which is not conducive to the accurate measurement of flow field parameters. Firstly, the error is analyzed theoretically, and then, we adopt the simulation method to obtain the error size of the method. Finally, we proposed the Voigt wings fitting absorbance method to reduce the fitting error. The operation of Voigt wings fitting method is to Select the wings of the spectral line, and then use Voigt profile fitting, The difference between the two wings was used the numerical integral method to calculate area, the integrated area is sum of Voigt profile fitting area and numerical integral area. We have used water vapor as the target gas, with eight absorption lines which have different low-level states energy from HITRAN 2012 database being selected-, building two kinds of non-uniform flow field model base on the flat flame furnace, and through the method of segmentation to equivalent processing the no uniformity of flow field. Using Voigt profile fitting method, numerical integral method and Voigt profile wings fitting method to obtain the integral area of models, the error size is obtained by comparing with the theoretical value. As the result of contrast, the fitting error of Voigt profile fitting method is large and related to the different absorption line, the error of numerical integral method is biggest but it is nothing to do with absorption line, the fitting error of Voigt profile wings fitting method is least and stable. By force of contrast, we determined the appropriate method to obtain integral area in the different non-uniform flow field, which is beneficial to obtain accurate integrated area and flow field parameters.

Alignment error analysis of the snapshot imaging polarimeter.

A snapshot imaging polarimeter (SIP) system is able to reconstruct two-dimensional spatial polarization information through a single interferogram. In this system, the alignment errors of the half-wave plate (HWP) and the analyzer have a predominant impact on the accuracies of reconstructed complete Stokes parameters. A theoretical model for analyzing the alignment errors in the SIP system is presented in this paper. Based on this model, the accuracy of the reconstructed Stokes parameters has been evaluated by using different incident states of polarization. An optimum thickness of the Savart plate for alleviating the perturbation introduced by the alignment error of the HWP is found by using the condition number of the system measurement matrix as an objective function in a minimization procedure. The result shows that when the thickness of a Savart plate is 23 mm, corresponding to the condition number 2.06, the precision of the SIP system can reach to 0.21% at 1° alignment tolerance of the HWP.

Design principles for single standing nanowire solar cells: going beyond the planar efficiency limits.

Semiconductor nanowires (NWs) have long been used in photovoltaic applications but restricted to approaching the fundamental efficiency limits of the planar devices with less material. However, recent researches on standing NWs have started to reveal their potential of surpassing these limits when their unique optical property is utilized in novel manners. Here, we present a theoretical guideline for maximizing the conversion efficiency of a single standing NW cell based on a detailed study of its optical absorption mechanism. Under normal incidence, a standing NW behaves as a dielectric resonator antenna, and its optical cross-section shows its maximum when the lowest hybrid mode (HE11δ) is excited along with the presence of a back-reflector. The promotion of the cell efficiency beyond the planar limits is attributed to two effects: the built-in concentration caused by the enlarged optical cross-section, and the shifting of the absorption front resulted from the excited mode profile. By choosing an optimal NW radius to support the HE11δ mode within the main absorption spectrum, we demonstrate a relative conversion-efficiency enhancement of 33% above the planar cell limit on the exemplary a-Si solar cells. This work has provided a new basis for designing and analyzing standing NW based solar cells.

Enhanced carrier extraction of a-Si/c-Si solar cells by nanopillar-induced optical modulation.

We demonstrate improved short-wavelength internal quantum efficiency (IQE) of a-Si/c-Si heterojunction (HJ) solar cells with a surface nanopillar (NP) array via simulation. The gain in IQE is attributed to the light-field modulation caused by the cavity resonance inside the NPs, in which the light energy is effectively localized within the c-Si bulk rather than the a-Si layer. The average IQE in the short-wavelength range (330-450 nm) is enhanced from 43.94% to 62.88% by the optimal NP array, with a maximum IQE of 80.98% at λ = 400 nm. The resulting current gain is over 38.25% compared to a planar HJ cell in this wavelength range, showing a well suppressed recombination-induced current loss. This light-management scheme may also find applications in other types of cells.

Dispersion measurement of tapered air-silica microstructure fiber by white-light interferometry.

Dispersion properties of novel, tapered, air-silica microstructure fibers are measured between 1.3 and 1.65 microm by white-light interferometry. Dispersion values (beta2) of -181 and -152 ps2/km were obtained for 2.2- and 3-microm core sizes, respectively, at lambda = 1.55 microm.