Traditional Fitting Research Articles

Using rare event algorithms to understand the statistics and dynamics of extreme heatwave seasons in South Asia

Abstract Computing the return times of extreme events and assessing the impact of climate change on such return times is fundamental to extreme event attribution studies. However, the rarity of such events in the observational record makes this task a challenging one, even more so for ‘record-shattering’ events that have not been previously observed at all. While climate models could be used to simulate such extremely rare events, such an approach entails a huge computational cost: gathering robust statistics for events with return time of centuries would require a few thousand years of simulation. In this study, we use an innovative tool, rare event algorithm, that allows us to sample numerous extremely rare events at a much lower cost than direct simulations. We employ the algorithm to sample extreme heatwave seasons, corresponding to large anomalies of the seasonal average temperature, in a heatwave hotspot of South Asia using the global climate model Plasim. We show that the algorithm estimates the return levels of extremely rare events with much greater precision than traditional statistical fits. It also enables the computation of various composite statistics, whose accuracy is demonstrated through comparison with a very long control run. In particular, our results reveal that extreme heatwave seasons are associated with an anticyclonic anomaly embedded within a large-scale hemispheric quasi-stationary wave-pattern. Additionally, the algorithm accurately represents the intensity-duration-frequency statistics of sub-seasonal heatwaves, offering insights into both seasonal and sub-seasonal aspects of extreme heatwave seasons. This innovative approach could be used in extreme event attribution studies to better constrain the changes in an event’s probability and intensity with global warming, particularly for events with return times spanning centuries or millennia.

A GPU Algorithm for Solving the Positions of New Pulsars

Timing newly discovered pulsars requires gradually building up a timing model that connects observations taken days to months apart. This sometimes can be challenging when our initial knowledge of the pulsar’s position is arcminutes off from its true position. Such a position error leads to significant arrival time shifts as a result of the Earth’s orbital motion. Traditional down-hill fitting timing algorithms become ineffective when our model predicts the wrong pulse rotations for our next observation. For some pulsars whose model prediction is not too far off, the correct rotation number could be found by trial-and-error methods. For the remaining challenging pulsars, a more generalized method is called for. This paper proposes a GPU-based algorithm that could exhaustively search a large area of trail positions for probable timing solutions. This could help find phase-connected timing solutions for new pulsars using brute force.

Separating the Spectral Counterparts in NGC 1275/Perseus Cluster in X-Rays

We present a model-independent method for separating the spectral counterparts of the active galactic nucleus (AGN) NGC 1275 from the surrounding emission of the Perseus cluster, as observed by Suzaku/XIS cameras. The Perseus cluster emission extends to higher energies than typically observed in AGN environments, reaching up to 9–10 keV. This necessitates precise separation of AGN and cluster spectra. To circumvent the degeneracy arising from numerous spectral fitting parameters, including elemental abundances, thermal and Compton emissions from the nucleus, and spectral parameters of the jet synchrotron self-Compton/inverse Compton emissions, we avoid traditional spectral fitting methods. Instead, we leverage spatial resolution and employ a double background subtraction approach. We apply this procedure to the complete set of Suzaku/XIS observational data for NGC 1275, resulting in cleaned spectra and a light curve of the AGN emission in this system. To demonstrate the applicability of our method, we also utilize the available XMM-Newton/EPIC data.

Denoising Medium Resolution Stellar Spectra With Neural Networks

ABSTRACTWe trained denoiser autoencoding neural networks on medium resolution simulated optical spectra of late‐type stars to demonstrate that the reconstruction of the original flux is possible at a typical relative error of a fraction of a percent down to a typical signal‐to‐noise ratio of per pixel. We show that relatively simple networks are capable of learning the characteristics of stellar spectra while still flexible enough to adapt to different values of extinction and fluxing imperfections that modifies the overall shape of the continuum, as well as to different values of Doppler shift. Denoised spectra can be used to find initial values for traditional stellar template fitting algorithms and—since evaluation of pretrained neural networks is significantly faster than traditional template fitting—denoiser networks can be useful when a fast analysis of the noisy spectrum is necessary, for example during observations, between individual exposures.

Evaluating possible glacial modification in the Martian tropics near Gale crater using a new U-shaped valley metric

Growing evidence supports the case for glacial processes within the equatorial regions of Mars. However, few examples of modeling or numerical analysis exist to support this hypothesis. Here we apply an automated method on valleys near Gale crater to extract morphometric data and analyze formative processes responsible for their current expression. The V-index utilized is a new robust method that is an alternative to traditional valley parabolic curve fits. This approach more easily characterizes irregular valley cross sections and thus aids in distinguishing between glacial and nonglacial forms. We applied this method along with standard power law and quadratic curve fits on valley cross sections along a 2300 km extent of the Martian tropics near Gale crater. Both V-index and power law b fit values suggest that the majority of the valleys are U-shaped and possibly created through glacial erosive processes. Further, V-index and power law b values plotted against elevation show a positive trend with higher V-index (glaciated) values associated with higher elevation. This suggests that an equilibrium-line altitude (ELA) may have existed at the time of valley formation. The timing of the formation of these U-shaped valleys is poorly defined. However, our initial crater statistics conducted within the study area suggest a young ∼1.05 (±0.25) Ga relative ages of the valley floors. Equatorial glaciers that produced these and other related glacial forms may be as old as the Hesperian to Amazonian transition (∼3 Ga) but more intriguingly may be associated with Amazonian equatorward migration of the Martian cryosphere during more recent periods of high obliquity oscillations.

A dual-template molecularly imprinted electrochemical sensor assisted with BO-XGBoost algorithm for simultaneous determination of dopamine and acetaminophen

The simultaneous detection of dopamine (DA) and acetaminophen (AP) is crucial for diagnosing and treating related mental disorders. However, accurately determining DA and AP in biological samples remains challenging due to the interference from other biomolecules, such as ascorbic acid (AA), uric acid (UA), epinephrine (EP), etc. Here, we present a dual-template molecularly imprinted electrochemical sensor for the selective recognition of DA and AP. Reduced graphene(rGO)-gold nanoparticles(AuNPs) composite were utilized to enhance the effective area and increase the electron transfer rate of glassy carbon electrode(GCE), thus the electrochemical response signals were amplified. This sensor combined the advantages of nanocomposites and molecularly imprinted polymer (MIP) to achieve highly sensitive and selective detection of DA and AP. However, decreases in the current response of some analytes and variations in relationships of the peak current versus analyte concentration will occur due to adsorption competition on active sites in multianalyte mixtures, rendering the detection range constrained and traditional linear fitting methods inaccurate for predicting analyte concentrations. Therefore, a concentration prediction model based on XGBoost was developed for the sensor to achieve the reliable multi-component determination of DA and AP within the identical measurement range as individual detection. In this model, nine characteristic parameters were extracted from the differential pulse voltammetry (DPV) response curve and Bayesian optimization (BO) was adopted for automatic hyperparameter search, thereby the prediction errors were reduced and the generalization ability was improved. Experiments indicated that the MIP/AuNPs/rGO/GCE exhibited a wide detection range of 2–240 μM and 3–240 μM for DA and AP, with detection limits down to 0.26 μM and 0.33 μM, respectively. In addition, the intelligent MIP/AuNPs/rGO/GCE sensor based on BO-XGBoost model provides more accurate prediction results for untrained concentration combinations obtained from fetal bovine serum samples, indicating that the proposed detection scheme holds potential in future clinical diagnosis.

Tunnel full-space deformation measurement based on improved downsampling and registration approach by point clouds

To fully leverage the volume measurement advantage of three-dimensional laser scanning technology in tunnel deformation monitoring, a full-space deformation analysis is proposed. A Box Improved Moving Average (BIMA) method featured by the adaptive window size downsampling is proposed which aims to achieve varying levels of downsampling in different densities regions. the combination of Maximal Cliques (MAC) and Iterative Closest Point (ICP) is utilized to achieve accurate registration. Point normal information is utilized to determine the correspondence to calculate deformation which shows an average accuracy of 7 % higher than the ICP by a convergence measurement test. The deformation convergence accuracy can go beyond 76 % when the point cloud reserved rate is approximately 10 %. Comparison between the method proposed in this study and the traditional point cloud cross sectional fitting method shows that the discrepancy between these two methods is approximately within 5 mm, with a minimum value of only 0.15 mm.

Control model of community elderly recreational exercise assistive robot based on improved dense trajectory algorithm

As the number of elderly population in the community grows, more efficient and precise recreation and exercise aids are needed to safeguard their quality of life. The study proposes a control model based on an improved dense trajectory algorithm to enhance the recognition and response capabilities of recreation and exercise assistance robots. The main method of the model is to improve the dense trajectory algorithm to enhance its recognition speed and accuracy for complex and small movements. Specifically, the study deeply refined the control process of a health and exercise assisted robot, and combined action capture to construct a health and exercise assisted robot model. The control model of the health and exercise assisted robot was optimized using an improved dense algorithm. The improved dense trajectory algorithm has feature embedding and attention mechanism, which can supplement the input data of the model, thus enabling more accurate action recognition. The results show that among the five samples, the recreation effectiveness score of the experimental group averaged 8.9, which was significantly higher than that of the control group, which was 7.3. The recognition accuracy has been improved by 2.7 % and 3.9 %, respectively, effectively suppressing the influence of camera motion. After using the improved dense trajectory algorithm, the fitness of the health training assistant robot reached 96.25 % under the same processing time, which is 8.93 % higher than the traditional model's fitness of 87.32 %. In summary, the control model of a community elderly health exercise assistance robot based on improved dense trajectory algorithm has achieved more accurate and faster recognition and response to the actions of the elderly, providing a more efficient technical means for health exercise and improving the health effect of the elderly.

Hyperspectral imaging with deep learning for quantification of tissue hemoglobin, melanin, and scattering.

Hyperspectral cameras capture spectral information at each pixel in an image. Acquired spectra can be analyzed to estimate quantities of absorbing and scattering components, but the use of traditional fitting algorithms over megapixel images can be computationally intensive. Deep learning algorithms can be trained to rapidly analyze spectral data and can potentially process hyperspectral camera data in real time. A hyperspectral camera was used to capture wide-field reflectance images of in vivo human tissue at 205 wavelength bands from 420 to 830nm. The optical properties of oxyhemoglobin, deoxyhemoglobin, melanin, and scattering were used with multi-layer Monte Carlo models to generate simulated diffuse reflectance spectra for 24,000 random combinations of physiologically relevant tissue components. These spectra were then used to train an artificial neural network (ANN) to predict tissue component concentrations from an input reflectance spectrum. The ANN achieved low root mean square errors in a test set of 6000 independent simulated diffuse reflectance spectra while calculating concentration values more than 4000× faster than a conventional iterative least squares approach. In vivo finger occlusion and gingival abrasion studies demonstrate the ability of this approach to rapidly generate high-resolution images of tissue component concentrations from a hyperspectral dataset acquired from human subjects.

Modified Beer-Lambert algorithm to measure pulsatile blood flow, critical closing pressure, and intracranial hypertension.

We introduce a frequency-domain modified Beer-Lambert algorithm for diffuse correlation spectroscopy to non-invasively measure flow pulsatility and thus critical closing pressure (CrCP). Using the same optical measurements, CrCP was obtained with the new algorithm and with traditional nonlinear diffusion fitting. Results were compared to invasive determination of intracranial pressure (ICP) in piglets (n = 18). The new algorithm better predicted ICP elevations; the area under curve (AUC) from logistic regression analysis was 0.85 for ICP ≥ 20 mmHg. The corresponding AUC for traditional analysis was 0.60. Improved diagnostic performance likely results from better filtering of extra-cerebral tissue contamination and measurement noise.

A novel temperature drift compensation method based on LSTM for NMR sensor

Nuclear Magnetic Resonance (NMR) sensors have become one of the best choices for angular velocity sensors in future Inertial Navigation Systems (INS) due to their small size and high precision. The performance of NMR sensors can be affected by temperature changes, resulting in temperature drift. Therefore, temperature compensation is essential. Accurate temperature compensation is challenging due to the lack of direct temperature observations in the core sensitive area (cell) and the complex temperature drift model. Considering the time correlation of temperatures inside and outside the cell, a new temperature drift compensation method based on the Long-Short Term Memory (LSTM) network is proposed. The memory characteristics of the LSTM enable it to fully utilize current and previous data to learn complex models. The main contributions of this study are as follows: (1) the mechanism of temperature drift in NMR sensors was analyzed; (2) a multi-time scale input–output model for temperature changes, temperature change rates, and ambient temperature versus sensor output bias was established; (3) a temperature drift model identification method based on LSTM was designed. Experimental results show that, compared to traditional polynomial fitting and memory-free methods, the proposed method improves temperature compensation accuracy by 26.0% to 47.0%. This method effectively reduces the adverse impact of temperature changes on the NMR sensor.

Comparative Electromyographic Analysis in Leg Press of Traditional Fitness Equipment, Traditional Outdoor Fitness Equipment, and a New Model of Outdoor Fitness Equipment in Trained Young Men

This study compares the electromyographic activity (EMG) of different muscle groups (rectus femoris, vastus lateralis, biceps femoris, tibialis anterior, and gastrocnemius) of the lower limbs when performing a traditional seated leg press (SLP) with a classic piece of outdoor fitness equipment (OFE-SLP), and with a new OFE leg press that allows the user to adjust the intensity of the exercise by means of a selectorized system (BIOFIT-LP). It was found that the EMG of the OFE-SLP was significantly lower than that of the SLP, but similar activations to those of the SLP were achieved with the BIOFIT-LP. In conclusion, the inclusion of a system to be able to change intensity of the exercise in OFE achieves an EMG activity similar to traditional machinery in trained young men.

Learning from Protein Engineering by Deconvolution of Multi-Mutational Variants.

This review analyzes a development in biochemistry, enzymology and biotechnology that originally came as a surprise. Following the establishment of directed evolution of stereoselective enzymes in organic chemistry, the concept of partial or complete deconvolution of selective multi-mutational variants was introduced. Early deconvolution experiments of stereoselective variants led to the finding that mutations can interact cooperatively or antagonistically with one another, not just additively. During the past decade, this phenomenon was shown to be general. In some studies, molecular dynamics (MD) and quantum mechanics/molecular mechanics (QM/MM) computations were performed in order to shed light on the origin of non-additivity at all stages of an evolutionary upward climb. Data of complete deconvolution can be used to construct unique multi-dimensional rugged fitness pathway landscapes, which provide mechanistic insights different from traditional fitness landscapes. Along a related line, biochemists have long tested the result of introducing two point mutations in an enzyme for mechanistic reasons, followed by a comparison of the respective double mutant in so-called double mutant cycles, which originally showed only additive effects, but more recently also uncovered cooperative and antagonistic non-additive effects. We conclude with suggestions for future work, and call for a unified overall picture of non-additivity and epistasis.

Lernen aus Protein‐Engineering durch Dekonvolution von Multi‐Mutationalen Enzymvarianten**

AbstractDiese Übersicht analysiert eine Entwicklung in der Biochemie, Enzymologie und Biotechnologie, die ursprünglich als Überraschung betrachtet wurde. Nach der Etablierung der gerichteten Evolution stereoselektiver Enzyme in der organischen Chemie wurde das Konzept der partiellen oder vollständigen Dekonvolution selektiver multi‐mutationaler Varianten eingeführt. Frühe Dekonvolutionsexperimente von stereoselektiven Enzymvarianten führten zu der Erkenntnis, dass Mutationen kooperativ oder antagonistisch miteinander interagieren können, nicht nur additiv. Im letzten Jahrzehnt wurde dieses Phänomen als allgemeingültig nachgewiesen. In einigen Studien wurden molekulardynamische (MD) und quantenmechanische/ molekülmechanische (QM/MM) Berechnungen durchgeführt, um Licht auf den Ursprung der Nicht‐Additivität in allen Phasen eines evolutionären Aufstiegs zu werfen. Daten der vollständigen Dekonvolution können verwendet werden, um einzigartige mehrdimensionale, zerklüftete („rugged“) Fitness‐Pfad‐Landschaften experimentell zu konstruieren. Sie bieten mechanistische Einblicke, die sich von traditionellen Fitnesslandschaften erheblich unterscheiden. In einem verwandten Bereich haben Biochemiker lange das Ergebnis der Einführung von zwei Punktmutationen in einem Enzym aus mechanistischen Gründen getestet, gefolgt von einem Vergleich mit der entsprechenden Doppelmutante in sogenannten Doppelmutanten‐Zyklen, die ursprünglich nur additive Effekte zeigten. In jüngerer Zeit wurden auch hier kooperative und antagonistische nicht‐additive Effekte aufgedeckt. Wir schließen mit Vorschlägen für zukünftige Arbeiten und fordern ein einheitliches Gesamtbild von Nicht‐Additivität und Epistasis.

Performance of hybrid gain formula versus traditional fitting formulas in hearing aid fitting in tinnitus patients with hearing loss.

Hearing aid fitting can be challenging when tinnitus accompanies hearing loss, as speech intelligibility and quality of life are affected by both hearing loss and tinnitus perception. However, studies focusing on the optimal hearing aid fitting for this group are scarce. Here, we aim to investigate the performance of alternative hearing aid fitting scenarios in improving hearing aid benefit and managing tinnitus. Sixty-six participants were included in the study and randomly divided into three groups based on the fitting formula: NAL-NL2, DSL pediatric and hybrid gain fitting procedure (covering NAL-NL2 for low frequencies and DSL pediatric formulas for high frequencies). Hearing aid benefit was evaluated using the Abbreviated Profile of Hearing Aid Benefit (APHAB) questionnaire and speech perception in noise (SPIN). To evaluate tinnitus perception, psychoacoustic characteristics of tinnitus were determined, and the Tinnitus Handicap Inventory (THI) was gathered. The NAL-NL2 fitting procedure showed better results in hearing aid benefit and SPIN compared to the DSL pediatric procedure. In the DSL pediatric procedure, better results were obtained in tinnitus management compared to NAL-NL2. There was no difference between the hybrid fitting procedure and DSL pediatric in tinnitus management. The hybrid fitting procedure also did not differ from NAL-NL2 in SPIN and hearing aid benefit. Here, we propose a hybrid gain fitting procedure that can be a better alternative to boost hearing aid performance and tinnitus management in clinical practice.

Effectiveness of colonoscopy, immune fecal occult blood testing, and risk-graded screening strategies in colorectal cancer screening.

Colorectal cancer (CRC) is one of the most common malignant tumors, and early screening is crucial to improving the survival rate of patients. The combination of colonoscopy and immune fecal occult blood detection has garnered significant attention as a novel method for CRC screening. Colonoscopy and fecal occult blood tests, when combined, can improve screening accuracy and early detection rates, thereby facilitating early intervention and treatment. However, certain risks and costs accompany it, making the establishment of a risk classification model crucial for accurate classification and management of screened subjects. To evaluate the feasibility and effectiveness of colonoscopy, immune fecal occult blood test (FIT), and risk-graded screening strategies in CRC screening. Based on the randomized controlled trial of CRC screening in the population conducted by our hospital May 2020 to May 2023, participants who met the requirements were randomly assigned to a colonoscopy group, an FIT group, or a graded screening group at a ratio of 1:2:2 (after risk assessment, the high-risk group received colonoscopy, the low-risk group received an FIT test, and the FIT-positive group received colonoscopy). The three groups received CRC screening with different protocols, among which the colonoscopy group only received baseline screening, and the FIT group and the graded screening group received annual follow-up screening based on baseline screening. The primary outcome was the detection rate of advanced tumors, including CRC and advanced adenoma. The population participation rate, advanced tumor detection rate, and colonoscopy load of the three screening programs were compared. A total of 19373 subjects who met the inclusion and exclusion criteria were enrolled, including 8082 males (41.7%) and 11291 females (58.3%). The mean age was 60.05 ± 6.5 years. Among them, 3883 patients were enrolled in the colonoscopy group, 7793 in the FIT group, and 7697 in the graded screening group. Two rounds of follow-up screening were completed in the FIT group and the graded screening group. The graded screening group (89.2%) and the colonoscopy group (42.3%) had the lowest overall screening participation rates, while the FIT group had the highest (99.3%). The results of the intentional analysis showed that the detection rate of advanced tumors in the colonoscopy group was greater than that of the FIT group [2.76% vs 2.17%, odds ratio (OR) = 1.30, 95% confidence interval (CI): 1.01-1.65, P = 0.037]. There was no significant difference in the detection rate of advanced tumors between the colonoscopy group and the graded screening group (2.76% vs 2.35%, OR = 1.9, 95%CI: 0.93-1.51, P = 0.156), as well as between the graded screening group and the FIT group (2.35% vs 2.17%, OR = 1.09%, 95%CI: 0.88-1.34, P = 0.440). The number of colonoscopy examinations required for each patient with advanced tumors was used as an index to evaluate the colonoscopy load during population screening. The graded screening group had the highest colonoscopy load (15.4 times), followed by the colonoscopy group (10.2 times), and the FIT group had the lowest (7.8 times). A hierarchical screening strategy based on CRC risk assessment is feasible for screening for CRC in the population. It can be used as an effective supplement to traditional colonoscopy and FIT screening programs.

Analysis of Rutherford backscattering spectra with CNN-GRU mixture density network

Ion Beam Analysis (IBA) utilizing MeV ion beams provides valuable insights into surface elemental composition across the entire periodic table. While ion beam measurements have advanced towards high throughput for mapping applications, data analysis has lagged behind due to the challenges posed by large volumes of data and multiple detectors providing diverse analytical information. Traditional physics-based fitting algorithms for these spectra can be time-consuming and prone to local minima traps, often taking days or weeks to complete. This study presents an approach employing a Mixture Density Network (MDN) to model the posterior distribution of Elemental Depth Profiles (EDP) from input spectra. Our MDN architecture includes an encoder module (EM), leveraging a Convolutional Neural Network-Gated Recurrent Unit (CNN-GRU), and a Mixture Density Head (MDH) employing a Multi-Layer Perceptron (MLP). Validation across three datasets with varying complexities demonstrates that for simple and intermediate cases, the MDN performs comparably to the conventional automatic fitting method (Autofit). However, for more complex datasets, Autofit still outperforms the MDN. Additionally, our integrated approach, combining MDN with the automatic fit method, significantly enhances accuracy while still reducing computational time, offering a promising avenue for improved analysis in IBA.

Emerging Technologies and Open-Source Platforms for Remote Physical Exercise: Innovations and Opportunities for Healthy Population-A Narrative Review.

The emergence of tele-exercise as a response to the impact of technology on physical activity has opened up new possibilities for promoting physical health. By integrating innovative technologies and open-source platforms, tele-exercise encourages people to stay active. In our latest analysis, we delved into the scientific literature surrounding the use of tele-exercise technologies in training healthy individuals. After conducting an extensive search on the PubMed database using the keywords "tele-exercise" and "physical activity" (from 2020 to 2023), we identified 44 clinical trials that were applicable to tele-exercise, but less than 10% of them were aimed at healthy individuals, precisely 9.09% (four out of forty-four studies analyzed). Our review highlights the potential of tele-exercise to help maintain physical fitness and psychological well-being, especially when traditional fitness facilities are not an option. We also underscore the importance of interoperability, standardization, and the incorporation of biomechanics, exercise physiology, and neuroscience into the development of tele-exercise platforms. Nevertheless, despite these promising benefits, research has shown that there is still a significant gap in the knowledge concerning the definition and evaluation of training parameters for healthy individuals. As a result, we call for further research to establish evidence-based practices for tele-exercise in the healthy population.

Predicting Structural Temperature of Ancient City Walls: A Case Study Using Ambient Temperature and Deep Learning

ABSTRACT Accurate prediction of the structural temperature field is required for the performance assessment of engineering structures and for maintaining structural safety. However, there are complex time-lag effects between ambient temperature and structural temperature. It is difficult to accurately describe and predict the structural temperature using traditional linear and nonlinear fitting methods. Hence, this study uses the advantages of gated recurrent unit (GRU) neural networks in dealing with complex nonlinear mapping issues, further considering solar radiation and developing a structural temperature prediction framework using ambient temperature and the deep learning model. The framework is applied in structural temperature prediction for an ancient city wall in China. The solar altitude angle at the location of the city wall is calculated, and the input and output structures of the deep learning neural network model are optimized. In the long-term prediction of the structural temperature of the city wall, the root mean square error (RMSE) of the GRU model is less than 0.6°C, which solves the problem of low accuracy caused by time-lag effects.

Effect of modified Baduanjin exercise on nutritional status in patients with nasopharyngeal carcinoma: a randomized controlled trial.

As a traditional Chinese fitness technique, Baduanjin is a low- to medium-intensity aerobic exercise that has a common regulatory effect on both body and mind and is also an important means of disease prevention and treatment. However, the role of Baduanjin in improving patients' nutritional status and promoting tumor recovery remains to be confirmed. This study aims to investigate the effectiveness of the modified Baduanjin exercise on the nutritional status of patients with nasopharyngeal carcinoma. This is a randomized controlled trial. The participants were recruited from patients in the Radiotherapy Department of the First Affiliated Hospital of Guangxi Medical University in China. A total of 121 patients with nasopharyngeal carcinoma were randomly divided into the control group and the Baduanjin group. Finally, 106 patients completed the study (53 cases each in the control group and the Baduanjin group) with the intervention time from the beginning to the end of radiotherapy. The control group received conventional care (health education and regular conventional exercise), and the Baduanjin exercise group received health education and regularly improved Baduanjin exercise, with the intervention time from the beginning to the end of the radiotherapy. Patient-generated subjective global assessment (PG-SGA) was evaluated before, during (15 times), and at the end of radiotherapy as the main evaluation index to compare nutritional status between the two groups. From August 2022 to December 2022, 121 patients with nasopharyngeal carcinoma were randomly divided into the control group and the Baduanjin group. During the intervention, 15 patients withdrew from the study, leading to 53 of 59 patients in the control group and 53 of 62 patients in the Baduanjin group. After the intervention, the PG-SGA score, radioactive oral mucositis, andoropharyngeal pain score were lower (P < 0.05), whereas anorexia scores,the levels of hemoglobin, albumin, prealbumin, and total protein were higher than those in the control group (P < 0.05). Modified Baduanjin exercise can improve the nutritional status of patients with nasopharyngeal carcinoma and deserves further clinical application. This study was registered in the Chinese Clinical Trial Registry under the registration number ChiCTR2200064519, registered on August 27, 2022. The public research topic is the construction and intervention research based on Internet + nasopharyngeal carcinoma.