Machine learning of fMRI network connectivity predicts repetitive TMS for smoking cessation
Poster No:
62
Submission Type:
Abstract Submission
Authors:
Xingbao Li1, Kevin Caulfield2
Institutions:
1Medical University of South Carolina, Charleston, SC, 2Medical University of South Carolina, Unknown/Not reported, SC
First Author:
Co-Author:
Introduction:
Functional magnetic resonance imaging (fMRI) can be used to identify therapeutic targets and evaluate the effect of repetitive transcranial magnetic stimulation (rTMS) in neural networks in tobacco use disorder (TUD). Using machine learning methods, we investigated whether large network connectivity can be used to predict the rTMS effect for smoking cessation.
Methods:
Smoking cue exposure task-fMRI(T-fMRI) and resting-state fMRI (RS-fMRI) scans were acquired before and after the 10 sessions of active or sham rTMS (10 Hz, 3000 pulses per session) over the left dorsal lateral prefrontal cortex (DLPFC) in 42 treatment-seeking smokers (≥10 cigarettes per day). Five large networks, including default model networks (DMN), central executive network (CEN), dorsal attention network (DAN), salience network (SN), and reward network (RN) were compared between prior before and after 10 sessions of rTMS, and between active and sham rTMS. Neural network and regression analysis were carried out using average connectivity of the large networks.
·Five large networks rendered on the brain
Results:
Network analyses indicated a higher salience connectivity in T-fMRI and a lower salience connectivity in RS-fMRI, which predicts a better outcome of TMS treatment for smoking cessation. It also indicated a lower executive control connectivity in T-fMRI and a higher reward connectivity in RS-fMRI predicted a better outcome.
Correlation analysis showed that the reduction of CPD was positively correlated to the pre-treatment average SN ( r = 0.663, p 0.001). In contrast, it was negatively tendency correlated to CEN ( r = -0.375, p = 0.122). (Figure 2. A). With a response or non-response to rTMS treatment as dependence variables, Logistic analysis showed that the average SN was into the mode (Wald =3.69)
Linear Regression Analysis: We used the percent change of CPD as a dependent variable and averaged DMN, CEN, DAN, SN, and RN as independent variables. Adjusted R Square = 0.614, R = 0.846, R Square 0.715. Significance F change = 0.002. We found that the averaged SN can positively predict the change of CPD. (beta = 0.77, t = 4.477, p = 0.001). In contrast, we found that the averaged CEN negatively predicted the change of CPD (beta = -0.411, t = - 2.821, p = 0.014).
Sham group: No significant correlation was found between the reduction of CPD and the averaged network (p > 0.05).
We performed Linear Regression Analysis with sham group data. We did not find any network that can predict the change of CPD.
Resting-state fMRI
Resting-state fMRI analysis showed that both RN ( -0.566, p = 0.009) and SN (-0.506, p =0.023) would predict the response of TMS for smoking cessation. However, they showed a different direction from the task-fMRI results. Both RN and SN negatively predict the reduction of CPD. (Figure 2. B)
Neural Network Modeling
Response rate was the dependent variable, while averaged SN was the independent variable. However, the resting-state SN average predicted the Response of TMS with 100%. (Figure 3.A)
Task-fMRI averaged SN predicted 100% correction and 0% incorrection. With Task-fMRI averaged data, SN predicted 100% correction to quit or non-quit cases.
With resting-state data, averaged SN predicted 100% correction and 20% incorrection. (Figure2. B)
Correlation analysis showed that the reduction of CPD was positively correlated to the pre-treatment average SN ( r = 0.663, p 0.001). In contrast, it was negatively tendency correlated to CEN ( r = -0.375, p = 0.122). (Figure 2. A). With a response or non-response to rTMS treatment as dependence variables, Logistic analysis showed that the average SN was into the mode (Wald =3.69)
Linear Regression Analysis: We used the percent change of CPD as a dependent variable and averaged DMN, CEN, DAN, SN, and RN as independent variables. Adjusted R Square = 0.614, R = 0.846, R Square 0.715. Significance F change = 0.002. We found that the averaged SN can positively predict the change of CPD. (beta = 0.77, t = 4.477, p = 0.001). In contrast, we found that the averaged CEN negatively predicted the change of CPD (beta = -0.411, t = - 2.821, p = 0.014).
Sham group: No significant correlation was found between the reduction of CPD and the averaged network (p > 0.05).
We performed Linear Regression Analysis with sham group data. We did not find any network that can predict the change of CPD.
Resting-state fMRI
Resting-state fMRI analysis showed that both RN ( -0.566, p = 0.009) and SN (-0.506, p =0.023) would predict the response of TMS for smoking cessation. However, they showed a different direction from the task-fMRI results. Both RN and SN negatively predict the reduction of CPD. (Figure 2. B)
Neural Network Modeling
Response rate was the dependent variable, while averaged SN was the independent variable. However, the resting-state SN average predicted the Response of TMS with 100%. (Figure 3.A)
Task-fMRI averaged SN predicted 100% correction and 0% incorrection. With Task-fMRI averaged data, SN predicted 100% correction to quit or non-quit cases.
With resting-state data, averaged SN predicted 100% correction and 20% incorrection. (Figure2. B)
·Reduced CPD was associated with changes in the averaged salience network in both task fMRI and resting-state fMRI, with reverse direction.
Conclusions:
In summary, using fMRI, we provide preliminary evidence for a mechanistic basis for the apparent anti-smoking effect of rTMS applied to L-DLPFC in smokers. Our results support the hypothesis that an aberrant neural network connections is critical for craving and relapse in nicotine dependence [40] and that rTMS treatment can remediate the aberrant FC that can result in smoking cessation. Interestingly, the neural network and regression analysis showed that SN is crucial in predicting the TMS treatment effect. Notably, the dACC is a key node in the in-salience network and could be an alternative rTMS site for TUDs using deep TMS.
Brain Stimulation:
TMS 1
Emotion, Motivation and Social Neuroscience:
Reward and Punishment
Modeling and Analysis Methods:
fMRI Connectivity and Network Modeling 2
Novel Imaging Acquisition Methods:
BOLD fMRI
Keywords:
Addictions
Machine Learning
Therapy
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Are you Internal Review Board (IRB) certified? Please note: Failure to have IRB, if applicable will lead to automatic rejection of abstract.
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
Li X, Hartwell KJ, Owens M, Lematty T, Borckardt JJ, Hanlon CA, Brady KT, George MS. Repetitive transcranial magnetic stimulation of the dorsolateral prefrontal cortex reduces nicotine cue craving. Biol Psychiatry. 2013;73:714-720.
Li X, Sahlem GL, Badran BW, McTeague LM, Hanlon CA, Hartwell KJ, Henderson S, George MS. Transcranial magnetic stimulation of the dorsal lateral prefrontal cortex inhibits medial orbitofrontal activity in smokers. Am J Addict. 2017;26:788-794.
Li X, Caulfield KA, Hartwell KJ, Henderson S, Brady KT, George MS. Reduced executive and reward connectivity is associated with smoking cessation response to repetitive transcranial magnetic stimulation: A double-blind, randomized, sham-controlled trial. Brain Imaging Behav. 2023.
Li X, George MS, Zangen A. Brain stimulation therapeutics. Addiction Neuroscience. 2023;6.