Preoperative comprehensive malignancy risk estimation for thyroid nodules: Development and verification of a network-based prediction model

BackgroundIn order to avoid excessive treatment of thyroid nodules in the clinic, it is necessary to find a simple and practical analysis method to comprehensively and accurately reflect benign or malignant thyroid nodules. This study aimed to construct and validate a comprehensive and reliable network-based predictive model using a variety of imaging and laboratory criteria for thyroid nodules to stratify the risk of malignancy prior to surgery.MethodsWe retrospectively analyzed data from patients who underwent surgical treatment for thyroid nodules at the Thyroid and Breast Diagnosis and Treatment Center of Weifang Hospital of Traditional Chinese Medicine between January 2018 and December 2020. Binary logical regression analysis was performed to predict whether nodules were malignant or benign. The developmental dataset included 457 patients (January 2018–December 2020). The validation set included separate data points (n = 225, January 2018–December 2020).ResultsIn this study, criteria that showed significant predictive value for malignant nodules included TI-RADS: 4b (p = 0.065); Bethesda IV, Bethesda V, Bethesda VI (P < 0.0001); BRAF^V600E mutation (P < 0.0001); Calcitonin>5 pg/ml (p = 0.0037); and FNA-Tg>30 ng/ml (p = 0.0003). A 10-grade risk scoring system was developed. The risk of malignancy risk ranged from 2.06% to 100% and was positively associated with increasing risk grade. The areas under the receiver-operating characteristic curve of the development and validation sets were 0.972 and 0.946, respectively.ConclusionA simple, comprehensive and reliable web-based predictive model was designed using a variety of imaging and laboratory criteria to stratify thyroid nodules by probability of malignancy.     

Assessment of a novel deep learning-based marker-less motion capture system for gait study

BackgroundMarker-less systems based on digital video cameras and deep learning for gait analysis could have a deep impact in clinical routine. A recently developed system has shown promising results in terms of joint center position but has not been yet evaluated in terms of gait outcomes.Research questionHow does this novel marker-less system compare to a marker-based reference system in terms of clinically relevant gait parameters?MethodsThe deep learning method behind the developed marker-less system was trained on a dedicated dataset consisting of forty-one asymptomatic and pathological subjects each performing ten walking trials. The system could estimate the three-dimensional position of seventeen joint centers or keypoints (e.g., neck, shoulders, hip, knee, and ankles). We evaluated the marker-less system against a marker-based system in terms of differences in joint position (Euclidean distance), detection of gait events (e.g., heel strike and toe-off), spatiotemporal parameters (e.g., step length, time), kinematic parameters (e.g., hip and knee extension-flexion), and inter-trial reliability for kinematic parameters.ResultsThe marker-less system was able to estimate the three-dimensional position of joint centers with a mean difference of 13.1 mm (SD = 10.2 mm). 99% of the estimated gait events were estimated within 10 ms of the corresponding reference values. Estimated spatiotemporal parameters showed zero bias. The mean and standard deviation of the differences of the estimated kinematic parameters varied by parameter (for example, the mean and standard deviation for knee extension flexion angle were −3.0° and 2.7°). Inter-trial reliability of the measured parameters was similar to that of the marker-based references.SignificanceThe developed marker-less system can measure the spatiotemporal parameters within the range of the minimum detectable changes obtained using the marker-based reference system. Moreover, except for hip extension flexion, the system showed promising results in terms of several kinematic parameters.     

A new analytical cut-off point for determining 18 years of age using MRI on medial clavicular epiphysis

Evaluation of the ossification of the medial clavicular epiphysis plays a key role in forensic age estimation. The purpose of the present study was to assess a new numerical cut-off at the age of 18 years, taking into consideration Magnetic Resonance (MR) images of the medial clavicular epiphysis. We analyzed 163 MR scans of Italian subjects aged between 14 and 25 years. Using the data obtained we calculated two ratios: REM-1 (ratio between the length of the whole epiphysis and the length of the metaphysis) and REM-2 (ratio between the length of epiphyseal-metaphyseal fusion and the length of the metaphysis). In 68 out of 163 cases it was not possible to measure REM-2. The reproducibility was demonstrated using the Intraclass Correlation Coefficient (ICC) (Cronbach’s alpha > 0.80). REM-1 and REM-2 were compared in each category of age (adult and minor) by the Wilcoxon signed-rank test. The cut-off points for measurements of REM-1 and REM-2 were determined by logistic regression. For REM-1, the cut-off scores were 0.83 for all individuals (accuracy = 94.77%) and males (accuracy = 96.05%), and 0.86 for females (accuracy = 92.30%). For REM-2, the cut-off values were 0.40 for all individuals and males (accuracy = 100.00%), and 0.41 for females (accuracy = 100.00%). Finally, receiver operating characteristic (ROC) curves for age classification based on REM-1 and REM-2 were constructed, showing that REM-2 had the highest discriminative power. Thus, a new cut-off model for predicting the age of majority has been introduced, conducting a quantitative analysis thanks to the use of a high-resolution imaging tool.     

Sacral morphometrics for sex estimation of dead cases in Central Thailand

Sex estimation by various forensic anthropology approaches is a crucial factor for identification of human skeletal remains. However, inexpensive, uncomplicated and reliable methods are still required, especially in a remote crime scene and a high crime incidence area. Here, we examined 13 sacral parameters from 78 independent skeletons derived from deceases found in Central Thailand (male, n = 46; female, n = 32) using simple standard anthropometric techniques for sex allocation. Discriminant analysis exhibited that anterior-posterior diameter of S₁ vertebra corpus (APS) is the most accurate sacral parameter for sex determination in our study with 82.1% of correct discrimination rate. The accuracy could be improved up to 97.4% when additional three sacral variables including the length of sacrum measured from the medial anterior-superior sacral promontory to the medial anterior-inferior S₅ vertebra (ASL), alar index (ALI), and the maximum anterior breadth of sacrum measured across sacral alar (ABS) were computed together with APS. These encourage the use of sacral morphometrics for sex assessment of human sacrum remains in Central Thailand. However, further investigation with broadening sacral morphometric data across the country might provide a promising sex determination equation from a sacral skeleton for Thai population.     

Age estimation from ossification of sternum and true ribs using 3D post-mortem CT images in a Japanese population

Age estimation of cadavers from post-mortem “chest plate” using conventional radiography, which involves radiographic assessment of ossification around the sternum and rib ends, has been evaluated without fruitful results. This study examined the value of images of the chest plate obtained by three-dimensional post-mortem CT for estimation of age at time of death in a Japanese population. Five chest plate ossification scores were evaluated in 320 subjects, including ossification of the first costal cartilage (OF), ossification of the second to seventh costal cartilages at the rib (OR) and sternal (OS) ends, fusion of the manubriosternal joint (FM), and fusion of the xiphisternal joint (FX). OS was found to have the highest correlation with age while FM had no significant correlation. The best composite score for age estimation was the summative score for both sides of the OS and the right side of the OF and FX, for which the coefficient of determination (R²) and the standard error of estimation (SEE) were 0.608 and 12.44 years, respectively, for men and 0.590 and 14.65 years for women. The accuracy of the model was tested in a further 26 male and 24 female subjects, and the accuracy rate within the first SEE was 57.69% and 70.83%, respectively. This rapid and non-invasive method of age estimation in the chest plate area is superior to conventional methods and could be useful for estimation of age at time of death in the Japanese population.     

Neurophysiology of motor skill learning in chronic stroke

ObjectiveMotor learning is relevant in chronic stroke for acquiring compensatory strategies to motor control deficits. However, the neurophysiological mechanisms underlying motor skill acquisition with the paretic upper limb have received little systematic investigation. The aim of this study was to assess the modulation of corticomotor excitability and intracortical inhibition within ipsilesional primary motor cortex (M1) during motor skill learning.MethodsTen people at the chronic stage after stroke and twelve healthy controls trained on a sequential visuomotor isometric wrist extension task. Skill was quantified before, immediately after, 24 hours and 7 days post-training. Transcranial magnetic stimulation was used to examine corticomotor excitability and short- and long-interval intracortical inhibition (SICI and LICI) pre- and post-training.ResultsThe patient group exhibited successful skill acquisition and retention, although absolute skill level was lower compared with controls. In contrast to controls, patients’ ipsilesional corticomotor excitability was not modulated during skill acquisition, which may be attributed to excessive ipsilesional LICI relative to controls. SICI decreased after training for both patient and control groups.ConclusionsOur findings indicate distinct inhibitory networks within M1 that may be relevant for motor learning after stroke.SignificanceThese findings have potential clinical relevance for neurorehabilitation adjuvants aimed at augmenting the recovery of motor function.     

Interval Estimations for Variance Components: A Review and Implementations

Abstract

Statistical methods are often required in the pharmaceutical industry to ensure that precision parameters of interest are kept within acceptable levels. For example, analytical methods are validated to determine whether the method is fit for its intended purpose before used in a decision process. In these cases, analytical methods are evaluated on the basis of precision parameters of interest. An analysis of variance approach based on a random effect model is commonly used for describing precision parameters. Bounds on precision parameters are needed to quantify the uncertainty present in the data. Several frequentist, fiducial, and Bayesian approaches have been used for constructing bounds on precision parameters. However, in some cases, there is no clear guidance on how to effectively apply some of these methods. Other recently proposed (and less known) methods have been reported to have very good statistical properties and are rarely, or to the best of our knowledge, never used in precision studies. We survey and compare methods for constructing confidence intervals of linear combinations of variance components in a balanced random effect model. We consider not only widely used traditional methods, but also those that are less well known. We further elucidate potential issues and provide general guidance on the construction of the confidence intervals. Finally, the methods are demonstrated with two examples from applications with similar kinds of data.     

Demographic and social context of deaths during the 1854 cholera outbreak in Soho,London: a reappraisal of Dr John Snow's investigation

Deaths from cholera in Soho, London (late July to end of September 1854) exposed the epidemiology of the disease and demonstrated applied geospatial analysis by highlighting the shortest path principle followed by local residents when they obtained drinking water from a contaminated pump. The present investigation explores if households and individuals with different demographic and socio-economic characteristics were more or less likely to obtain their water from the pump and succumb to the disease. It combines information from the 1851 Population Census and topographic databases with the digital deaths and water pump data to reveal the risk of exposure and the mortality rate were greater for certain occupations, age groups and people living at high residential density irrespective of proximity to the contaminated water pump.