Geophysics最新文献

{"title":"TRAIL C1595T Variant Critically Alters the Level of sTRAIL in Terms of Histopathological Parameters in Colorectal Cancer.","authors":"Cem Horozoglu, Asli Yildiz, Dilara Sonmez, Seyda Demirkol, Yemliha Yildiz, Soykan Arikan, Ilhan Yaylim","doi":"10.1007/s12291-023-01146-z","DOIUrl":"10.1007/s12291-023-01146-z","url":null,"abstract":"<p><p>TRAIL, a member of the TNF family, is expressed in tumor and tumor surrounding tissue in many solid organ cancers. While the induction of tumor-specific apoptosis in correlation with cytokine stimulation may cause anti-tumoral effects, the pro-tumorigenic effects of its expression by tumor surrounding tissue members have been reported in the literature. In our study, it was aimed to evaluate the effect of the gene variant of TRAIL on soluble levels in patients with colorectal cancer (CRC) on the molecular pathological axis. TRAIL C1595 gene variant PCR-RFLP and sTRAIL levels were determined by ELISA in age and sex adjusted CRC and control groups. It was determined that CT carriage was high in patients without perineural invasion and sTRAIL levels were approximately 2.72 times lower than CC in patients with CT in this group (<i>p</i> < 0.05). Similarly, sTRAIL level was found to be high in patients with CC genotype in CRC without lymph node metastas. It was determined that CT carriage was high in patients without perineural invasion and sTRAIL levels were approximately 2.49 times lower than CC in patients with CT in this group.is (<i>p</i> < 0.05). We think that TRAIL C1595T in CRC can change sTRAIL levels in the clinicopathological axis in advanced stages such as metastasis and invasion, but both are not independent risk factors.</p>","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"56 1","pages":"593-599"},"PeriodicalIF":1.5,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11436522/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88006711","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The Effect of Height on Adverse Short-Term Outcomes After Lower-Extremity Bypass Surgery in Patients with Diabetes Mellitus.","authors":"Kushkaran Kaur, Rhonda S Cornell, Lawrence Oresanya, Andrew J Meyr","doi":"10.7547/21-258","DOIUrl":"10.7547/21-258","url":null,"abstract":"<p><strong>Background: </strong>We evaluated adverse short-term outcomes after open lower-extremity bypass surgery in patients with diabetes mellitus with a comparison performed based on patient height.</p><p><strong>Methods: </strong>The American College of Surgeons National Surgical Quality Improvement Program database was analyzed to select patients with Current Procedural Terminology codes 35533, 35540, 35556, 35558, 35565, 35566, 35570, and 35571 and with the diagnosis of diabetes mellitus. This resulted in 83 patients 60 inches or less in height, 1,084 between 60 and 72 inches, and 211 patients 72 inches and taller.</p><p><strong>Results: </strong>No differences were observed among groups with respect to the development of a superficial surgical site infection (9.6% versus 6.4% versus 5.7%; P = .458), deep incisional infection (1.2% versus 1.4% versus 2.8%; P = .289), sepsis (2.4% versus 2.0% versus 2.8%; P = .751), unplanned reoperation (19.3% versus 15.6% versus 21.8%; P = .071), or unplanned hospital readmission (19.3% versus 14.8% versus 17.1%; P = .573). A significant difference was observed among groups in the development of a wound disruption (4.8% versus 1.3% versus 4.7%; P = .001). A multivariate regression analysis was performed of the wound disruption outcome with the variables of age, sex, race, ethnicity, height, weight, current smoker, and open wound/wound infection. Race (P = .025) and weight (P = .003) were found to be independently associated with wound disruption, but height was not (P = .701).</p><p><strong>Conclusions: </strong>The results of this investigation demonstrate no significant differences in short-term adverse outcomes after lower-extremity bypass surgery based on patient height.</p>","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"27 1","pages":""},"PeriodicalIF":0.5,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87944565","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Stress-dependent reflection and transmission of elastic waves under confining, uniaxial, and pure shear prestresses","authors":"Haidi Yang, Li-Yun Fu, Tobias M. Müller, Bo-Ye Fu","doi":"10.1190/geo2023-0620.1","DOIUrl":"https://doi.org/10.1190/geo2023-0620.1","url":null,"abstract":"Insights into the reflection and transmission (R/T) of waves at a prestressed interface are important in geophysical applications, such as evaluating the angle-dependent elastic properties for monitoring geopressure and tectonic stress using sonic logging data or seismic data. Although many studies deal with wave propagation in prestressed media, the angle-dependent R/T of waves at an interface subject to different prestress loading modes remains largely unaddressed. We address this issue by applying the theory of acoustoelasticity with third-order acoustoelastic constants to study the R/T coefficients at the interface between two prestressed media. Stress-induced elastic deformations are assumed to be locally homogeneous without boundary dislocations caused by stress concentration so that the static boundary conditions can be applied. We consider three typical prestress modes (confining, uniaxial, and pure shear), each of which takes into account the incidence of upgoing and downgoing P and S waves. The Knott equations under different types of prestresses are derived, followed by the estimation of angle-dependent R/T coefficients. The energy conservation at the interface and the acoustoelastic finite-difference simulation of predicted P and S modes verify the correctness of the angle-dependent R/T coefficients under confining prestress. Comparisons with the elastic case (prestress <span><math display=\"inline\" overflow=\"scroll\"><mrow><mi>σ</mi><mo>=</mo><mn>0</mn><mtext> </mtext><mi>MPa</mi></mrow></math></span>) indicate the important influence of prestresses on the energy distribution of reflected and transmitted waves, including stress-dependent critical angles, converted waves, and R/T energy ratios. Such acoustoelastic effects mainly occur around/after the critical angle. For small-angle incidence, prestresses mainly affect the gradient of R/T coefficients. The type and magnitude of prestress are closely related to the angle-dependent R/T coefficients and must be considered for amplitude-variation-with-offset analysis in prestressed media.","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"44 1","pages":""},"PeriodicalIF":3.3,"publicationDate":"2024-05-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140839753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DeepNRMS: Unsupervised deep learning for noise-robust CO2 monitoring in time-lapse seismic images","authors":"Min Jun Park, Julio Frigerio, Bob Clapp, Biondo Biondi","doi":"10.1190/geo2023-0608.1","DOIUrl":"https://doi.org/10.1190/geo2023-0608.1","url":null,"abstract":"Monitoring stored <span><math display=\"inline\" overflow=\"scroll\"><mrow><msub><mi>CO</mi><mn>2</mn></msub></mrow></math></span> in carbon capture and storage projects is crucial for ensuring safety and effectiveness. We introduce DeepNRMS, a novel noise-robust method that effectively handles time-lapse noise in seismic images. The DeepNRMS leverages unsupervised deep learning to acquire knowledge of time-lapse noise characteristics from preinjection surveys. By using this learned knowledge, our approach accurately discerns <span><math display=\"inline\" overflow=\"scroll\"><mrow><msub><mi>CO</mi><mn>2</mn></msub></mrow></math></span>-induced subtle signals from the high-amplitude time-lapse noise, ensuring fidelity in monitoring while reducing costs by enabling sparse acquisition. We evaluate our method using synthetic data and field data acquired in the Aquistore project. In the synthetic experiments, we simulate time-lapse noise by incorporating random near-surface effects in the elastic properties of the subsurface model. We train our neural networks exclusively on preinjection seismic images and subsequently predict <span><math display=\"inline\" overflow=\"scroll\"><mrow><msub><mi>CO</mi><mn>2</mn></msub></mrow></math></span> locations from postinjection seismic images. In the field data analysis from Aquistore, the images from preinjection surveys are used to train the neural networks with the characteristics of time-lapse noise, followed by identifying <span><math display=\"inline\" overflow=\"scroll\"><mrow><msub><mi>CO</mi><mn>2</mn></msub></mrow></math></span> plumes within two postinjection surveys. The outcomes demonstrate the improved accuracy achieved by DeepNRMS, effectively addressing the strong time-lapse noise.","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"32 1","pages":""},"PeriodicalIF":3.3,"publicationDate":"2024-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140842528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Improvement of quality of life after 2-month exoskeleton training in patients with chronic spinal cord injury.","authors":"Ilse J W van Nes, Rosanne B van Dijsseldonk, Frank H M van Herpen, Hennie Rijken, Alexander C H Geurts, Noël L W Keijsers","doi":"10.1080/10790268.2022.2052502","DOIUrl":"10.1080/10790268.2022.2052502","url":null,"abstract":"<p><strong>Objective: </strong>To examine changes in quality of life (QoL) after an eight-week period of robotic exoskeleton training in a homogeneous group of patients with chronic complete spinal cord injury (SCI).</p><p><strong>Design: </strong>Prospective single-group pre-post study.</p><p><strong>Setting: </strong>Rehabilitation center.</p><p><strong>Participants: </strong>Patients with a chronic (>6 months) motor complete SCI (T1-L1).</p><p><strong>Intervention: </strong>Twenty-four training sessions with the ReWalk exoskeleton over an eight-week period.</p><p><strong>Main outcome measure: </strong>QoL, assessed with the sum score of the Short Form-36 with Walk Wheel modification (SF-36ww). Secondary outcome measures were the eight SF-36ww subdomains, satisfaction with bladder and bowel management, lower extremity joint passive range of motion (pROM), and lower extremity spasticity.</p><p><strong>Results: </strong>Twenty-one participants completed the training. QoL significantly improved after the training period (average SF-36 sum score 621 ± 90) compared to baseline (571 ± 133) (<i>t</i>(20)=-2.5, <i>P</i>=.02). Improvements were seen on the SF-36ww subdomains for pain (<i>P</i>=.003), social functioning (<i>P</i>=.03), mental health (<i>P</i>=.02), and general health perception (<i>P</i>=.01). Satisfaction with bladder management (range 1-5) improved from median 3 at baseline to 4 after exoskeleton training (<i>P</i>=0.01). No changes in satisfaction with bowel management (<i>P</i>=.11), pROM (hip-extension (<i>P</i>=.49), knee-extension (<i>P</i>=.36), ankle dorsiflexion (<i>P</i>=.69)), or spasticity (<i>P</i>=.94) were found.</p><p><strong>Conclusion: </strong>Even in patients with chronic motor complete SCI and a relatively high level of QoL at baseline, a short-term exoskeleton training improved their QoL, pain and satisfaction with bladder management; findings that warrant further controlled studies in this specific SCI population.</p>","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"37 1","pages":"354-360"},"PeriodicalIF":1.7,"publicationDate":"2024-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11044750/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81146949","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Free-surface multiple attenuation and seismic deghosting for blended data using convolutional neural networks","authors":"Mert S. R. Kiraz, Roel Snieder, Jon Sheiman","doi":"10.1190/geo2023-0417.1","DOIUrl":"https://doi.org/10.1190/geo2023-0417.1","url":null,"abstract":"Simultaneous source acquisition has become common over the past few decades for marine seismic surveys because of the increased efficiency of seismic acquisition by limiting the time, reducing the cost, and having less environmental impact than conventional single-source (or unblended) acquisition surveys. For simultaneous source acquisition, seismic sources at different locations are fired with time delays, and the recorded data are referred to as the blended data. The air-water interface (or free surface) creates strong multiples and ghost reflections for blended seismic data. The multiples and/or ghost reflections caused by a source in the blended data overlap with the primary reflections of another source, thus creating a strong interference between the primary and multiple events of different sources. We develop a convolutional neural network (CNN) method to attenuate free-surface multiples and remove ghost reflections simultaneously from the blended seismic data. The CNN-based solution that we develop operates on single traces and is not sensitive to the missing near-offset traces, missing traces, and irregular/sparse acquisition parameters (e.g., for ocean-bottom node acquisition and time-lapse monitoring studies). We illustrate the efficacy of our free-surface multiple attenuation and seismic deghosting method by presenting synthetic and field data applications. The numerical experiments demonstrate that our CNN-based approach for simultaneously attenuating free-surface multiples and removing ghost reflections can be applied to the blended data without the deblending step. Although the interference of primaries and multiples from different shots in the blended data makes free-surface multiple attenuation harder than the unblended data, we determine that our CNN-based method effectively attenuates free-surface multiples in the blended synthetic and field data even when the delay time for the blending is different in the training data than in the data to which the CNN is applied.","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"29 1","pages":""},"PeriodicalIF":3.3,"publicationDate":"2024-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140567700","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The effects of pore structure on wave dispersion and attenuation due to squirt flow: A dynamic stress-strain simulation on a simple digital pore-crack model","authors":"Zhifang Yang, Hong Cao, Luanxiao Zhao, Xinfei Yan, Yirong Wang, Wei Zhu","doi":"10.1190/geo2023-0521.1","DOIUrl":"https://doi.org/10.1190/geo2023-0521.1","url":null,"abstract":"Squirt flow, a phenomenon typically observed in porous cracked rocks, occurs due to the contrasting compressibility between the pores and cracks, leading to the pore pressure diffusion and dissipation of wave energy. Understanding the influence of pore structure on wave dispersion and attenuation signatures due to squirt flow is essential for interpreting seismic and sonic logging data in various fields of earth and energy sciences, such as hydrocarbon exploration, geothermal energy exploitation, and CO₂ sequestration. In this study, we construct a simple digital pore-crack model with varying pore structures and use a dynamic stress-strain simulation approach to characterize wave dispersion and attenuation signatures due to squirt flow. Numerical simulation suggests that, in addition to the commonly considered parameters such as porosity, crack density, and crack aspect ratio, additional pore structure parameters, such as pore size, pore aspect ratio, crack orientation, crack length, and crack width, significantly affect the wave dispersion and attenuation signatures. Increasing the pore size leads to the pronounced enhancement of attenuation magnitude and a decrease in characteristic frequency. We demonstrate that variations in crack length have a pronounced impact on attenuation magnitude, whereas crack width is decisive in controlling the characteristic frequency. Furthermore, it is found that the saturation paths (the gas filling the pore first or gas filling the crack first) considerably affect the velocity-saturation and attenuation-saturation relationship, suggesting that the coupling effects of pore structure and fluid distribution complicate the fluid pressure diffusion and wave attenuation behaviors. The presented results offer insights for deciphering pore structure parameters using attenuation- or dissipation-related seismic attributes.","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"595 1","pages":""},"PeriodicalIF":3.3,"publicationDate":"2024-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140567621","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Q-compensated image-domain least-squares reverse time migration through preconditioned point-spread functions","authors":"Wei Zhang, Jinghuai Gao, Ying Shi, Xuan Ke, Zhen Li, Tao Yang, Wenbo Sun","doi":"10.1190/geo2023-0333.1","DOIUrl":"https://doi.org/10.1190/geo2023-0333.1","url":null,"abstract":"Image-domain least-squares reverse time migration (IDLSRTM) through point-spread functions (PSFs) has been proven to be a feasible approach to improve the spatial resolution and amplitude fidelity of reflection images recovered by reverse time migration (RTM). However, it usually ignores the earth’s <i>Q</i>-effects, which may lead to an unfocused reflection image with an undesired spatial resolution. In this paper, we develop a <i>Q</i>-compensated IDLSRTM approach (denoted as <i>Q</i>-IDLSRTM) through PSFs, in which we use the viscoacoustic wave equation based on the generalized standard linear solid model to simulate inherent subsurface attenuation and the linear inversion to compensate for the amplitude attenuation. The PSFs are estimated by a round of modeling-migration computation and spatial interpolation on the fly. There are two key points in the developed <i>Q</i>-IDLSRTM approach. The first is that we must apply the deblurring filter as a preconditioner to compensate for the attenuation of image amplitude of PSFs and RTM in a viscoacoustic medium, before the iterative solution. The preconditioned PSFs and RTM images can help us to construct a less ill-posed image-domain inverse problem that can produce an improved image quality and a faster convergence rate, compared with the conventional <i>Q</i>-IDLSRTM approach without the deblurring filter. The second key point is that we can impose the L1-norm constraint and total variation regularization on the reflection image to stabilize the solution of the ill-posed inverse problem. Several 2D and 3D experiments verify that the developed approach can achieve better imaging quality in terms of amplitude fidelity and spatial resolution relative to the conventional <i>Q</i>-IDLSRTM and acoustic IDLSRTM approaches.","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"32 1","pages":""},"PeriodicalIF":3.3,"publicationDate":"2024-03-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140315525","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Distributed wide-field electromagnetic method for coal mining goaf detection in a complex urban environment: A case study in Jinan, China","authors":"Heng Zhang, Yang Yang, Hu Li, Pingfan Luo, Changyu Zhou, Yuzhen Zhu, Yonghui Peng, Huaifeng Sun","doi":"10.1190/geo2023-0248.1","DOIUrl":"https://doi.org/10.1190/geo2023-0248.1","url":null,"abstract":"Urban expansion has resulted in many proposed projects located over coal mining zones, which has highlighted the importance of detecting the spatial scope and water abundance of goafs before the commencement of construction work. Although electromagnetic (EM) methods have proven effective for goaf detection, their applications in intensely noisy and urbanized environments remain limited. To address this challenge, we evaluate an investigation of a coal mining goaf in Jinan, China, using the distributed wide-field EM (DWFEM) method. A third-order <span><math display=\"inline\" overflow=\"scroll\"><mrow><msup><mrow><mn>2</mn></mrow><mrow><mi>n</mi></mrow></msup></mrow></math></span> sequence pseudorandom signal with 39 survey frequencies is transmitted to achieve long-time data acquisition at each station. Unlike the controlled-source audio-frequency magnetotellurics method, the DWFEM records only the electric field <span><math display=\"inline\" overflow=\"scroll\"><mrow><msub><mi>E</mi><mi>x</mi></msub></mrow></math></span> component. The synthetic model tests and field data demonstrate the consistency of the <span><math display=\"inline\" overflow=\"scroll\"><mrow><msub><mi>E</mi><mi>x</mi></msub></mrow></math></span> apparent resistivity and the Cagniard resistivity in the far field. The long-time acquisition and 1C recording greatly improve data quality and exploration efficiency. We also use an all-angle resistivity calculation formula and an electrode layout method parallel to the wire source to obtain electrical connections in different directions. The DWFEM inversion results are obtained using the 1D Gauss-Newton iterative method under a plane-wave assumption. By interpolating the data from different measurement stations, we image resistivity depth profiles and obtain 3D subsurface electrical data for the subsurface from 0 to 1000 m. We interpret the obtained profiles with geologic and mining information, revealing two significant water-enriched goaf areas. Validation is performed using seismic data and drill cores. The results significantly enhance our understanding of the characteristics of the coal mine under our project and highlight the applicability of the DWFEM for detecting goafs in complex urban environments.","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"22 1","pages":""},"PeriodicalIF":3.3,"publicationDate":"2024-03-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140203942","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The effects of data leakage on connectome-based machine learning models.","authors":"Matthew Rosenblatt, Link Tejavibulya, Rongtao Jiang, Stephanie Noble, Dustin Scheinost","doi":"10.1101/2023.06.09.544383","DOIUrl":"10.1101/2023.06.09.544383","url":null,"abstract":"<p><p>Predictive modeling has now become a central technique in neuroimaging to identify complex brain-behavior relationships and test their generalizability to unseen data. However, data leakage, which unintentionally breaches the separation between data used to train and test the model, undermines the validity of predictive models. Previous literature suggests that leakage is generally pervasive in machine learning, but few studies have empirically evaluated the effects of leakage in neuroimaging data. Although leakage is always an incorrect practice, understanding the effects of leakage on neuroimaging predictive models provides insight into the extent to which leakage may affect the literature. Here, we investigated the effects of leakage on machine learning models in two common neuroimaging modalities, functional and structural connectomes. Using over 400 different pipelines spanning four large datasets and three phenotypes, we evaluated five forms of leakage fitting into three broad categories: feature selection, covariate correction, and lack of independence between subjects. As expected, leakage via feature selection and repeated subjects drastically inflated prediction performance. Notably, other forms of leakage had only minor effects (e.g., leaky site correction) or even decreased prediction performance (e.g., leaky covariate regression). In some cases, leakage affected not only prediction performance, but also model coefficients, and thus neurobiological interpretations. Finally, we found that predictive models using small datasets were more sensitive to leakage. Overall, our results illustrate the variable effects of leakage on prediction pipelines and underscore the importance of avoiding data leakage to improve the validity and reproducibility of predictive modeling.</p>","PeriodicalId":55102,"journal":{"name":"Geophysics","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-12-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10793416/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90754611","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}