Two Subtypes of ME/CFS Identified by DTI-ALPS Index Using Machine Learning
Poster No:
496
Submission Type:
Abstract Submission
Authors:
Zack Shan1, Qiang Yu1, Anya Bonner1, Richard Kwiatek1, Peter Del Fante1, Vince Calhoun2, Grant Bateman3
Institutions:
1University of the Sunshine Coast, Birtinya, Queensland, 2GSU/GATech/Emory, Atlanta, GA, 3John Hunter Hospital, Newcastle, New South Wales
First Author:
Co-Author(s):
Introduction:
Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is characterised by unexplained, persistent fatigue that does not improve with rest. Beyond this core feature, ME/CFS exhibits considerable variability in clinical presentation, disease onset, biological abnormalities, and biomarker profiles. This suggests that ME/CFS may represent a heterogeneous condition involving distinct underlying disease processes.
The persistent fatigue and cognitive dysfunction ("brain fog") in ME/CFS may result from impaired clearance of neurotoxic waste products. The glymphatic system, responsible for waste clearance via perivascular spaces (PVS), is most active during sleep, which is often disrupted in ME/CFS patients. Thus, glymphatic dysfunction may serve as a neurobiological pathway in ME/CFS.
This study employed Diffusion Tensor Imaging Analysis along the Perivascular Space (DTI-ALPS) to investigate PVS function in ME/CFS. Subtypes were identified using a semi-supervised machine learning approach, Heterogeneity through Discriminative Analysis (HYDRA).
The persistent fatigue and cognitive dysfunction ("brain fog") in ME/CFS may result from impaired clearance of neurotoxic waste products. The glymphatic system, responsible for waste clearance via perivascular spaces (PVS), is most active during sleep, which is often disrupted in ME/CFS patients. Thus, glymphatic dysfunction may serve as a neurobiological pathway in ME/CFS.
This study employed Diffusion Tensor Imaging Analysis along the Perivascular Space (DTI-ALPS) to investigate PVS function in ME/CFS. Subtypes were identified using a semi-supervised machine learning approach, Heterogeneity through Discriminative Analysis (HYDRA).
Methods:
This prospective study used data collected for the ongoing study of ME/CFS1, approved by the University of the Sunshine Coast Ethic Committee (A191288) and registered with The Australian New Zealand Clinical Trials Registry (ACTRN12622001095752). A total of 120 participants, including 75 ME/CFS patients (mean age: 42.28 ± 12.52 years; 62 women) and 45 healthy controls (HCs; mean age: 40.67 ± 8.32 years; 37 women), were included.
Detailed DTI sequence parameters and analysis procedures are available in our protocol1 and previous publications2. The DTI-ALPS indices for the bilateral hemispheres were calculated using an established method3.
HYDRA4 was applied to the DTI-ALPS indices to identify ME/CFS subtypes, with age, sex, and BMI included as covariates. Clustering consistency across resolutions (1–10 clusters) was assessed using the Adjusted Rand Index (ARI). Clinical measures, including disease severity, disease onsets, the Hospital Anxiety and Depression Scale (HADS), Pittsburgh Sleep Quality Index (PSQI), and 36-Item Short Form Health Survey (SF-36), were compared between subtypes to determine clinical relevance.
Detailed DTI sequence parameters and analysis procedures are available in our protocol1 and previous publications2. The DTI-ALPS indices for the bilateral hemispheres were calculated using an established method3.
HYDRA4 was applied to the DTI-ALPS indices to identify ME/CFS subtypes, with age, sex, and BMI included as covariates. Clustering consistency across resolutions (1–10 clusters) was assessed using the Adjusted Rand Index (ARI). Clinical measures, including disease severity, disease onsets, the Hospital Anxiety and Depression Scale (HADS), Pittsburgh Sleep Quality Index (PSQI), and 36-Item Short Form Health Survey (SF-36), were compared between subtypes to determine clinical relevance.
Results:
Two distinct subtypes of ME/CFS were identified, with the highest clustering reproducibility at ARI = 0.84 (Fig. 1). Subtype 1 included 29 patients (mean age: 37.86 ± 12.07 years; 25 women), while Subtype 2 included 46 patients (mean age: 45.07 ± 11.98 years; 37 women).
Subtype 2 exhibited significantly higher HADS depressive symptoms compared to both HCs (FDR-q < 0.05) and Subtype 1 (FDR-q < 0.05) (Fig. 2a). Additionally, Subtype 2 demonstrated significantly lower SF-36 Mental Component Summary (MCS) scores compared to HCs (FDR-q < 0.05) and Subtype 1 (FDR-q < 0.05) (Fig. 2b). No significant differences were observed between subtypes for HADS anxiety, PSQI, SF-36 Physical Component Summary (PCS), disease severity, or disease onset.
Subtype 2 exhibited significantly higher HADS depressive symptoms compared to both HCs (FDR-q < 0.05) and Subtype 1 (FDR-q < 0.05) (Fig. 2a). Additionally, Subtype 2 demonstrated significantly lower SF-36 Mental Component Summary (MCS) scores compared to HCs (FDR-q < 0.05) and Subtype 1 (FDR-q < 0.05) (Fig. 2b). No significant differences were observed between subtypes for HADS anxiety, PSQI, SF-36 Physical Component Summary (PCS), disease severity, or disease onset.
Conclusions:
This study identified two distinct ME/CFS subtypes, suggesting the presence of two glymphatic axes contributing to the condition. Subtype 2 was associated with more severe depressive symptoms and poorer mental health scores compared to both Subtype 1 and healthy controls.
Future research will further characterise these subtypes by examining their relationships with brain structure, function, and clinical features. These findings may pave the way for precision clinical care that acknowledges biological heterogeneity in ME/CFS diagnosis, prognosis, and treatment. Importantly, this approach leverages widely accessible clinical brain imaging techniques to enhance diagnostic accuracy and therapeutic strategies.
Future research will further characterise these subtypes by examining their relationships with brain structure, function, and clinical features. These findings may pave the way for precision clinical care that acknowledges biological heterogeneity in ME/CFS diagnosis, prognosis, and treatment. Importantly, this approach leverages widely accessible clinical brain imaging techniques to enhance diagnostic accuracy and therapeutic strategies.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia) 1
Modeling and Analysis Methods:
Classification and Predictive Modeling 2
Keywords:
DISORDERS
Machine Learning
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
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?
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.
2. Yu Q, Kwiatek RA, Del Fante P, et al. Opposite white matter abnormalities in post-infectious vs. gradual onset chronic fatigue syndrome revealed by diffusion MRI. medRxiv 2024: 2024.08.04.24311483.
3. Zhang W, Zhou Y, Wang J, et al. Glymphatic clearance function in patients with cerebral small vessel disease. Neuroimage 2021; 238: 118257.
4. Varol E, Sotiras A, Davatzikos C. HYDRA: Revealing heterogeneity of imaging and genetic patterns through a multiple max-margin discriminative analysis framework. Neuroimage 2017; 145(Pt B): 346-64.