Brain SPECT as an Imaging Biomarker for Evaluating Effects of Novel Treatments in Psychiatry–A Case Series

The difficulties of evaluating patients with complex neuropsychiatric conditions and prescribing appropriate treatments are well known. Imaging complements clinical assessments and allows a clinician to narrow the differential diagnosis by facilitating accurate and efficient evaluation. This is particularly relevant to neuropsychiatric conditions that are often diagnosed using a trial-and error process of exclusion. Single Photon Emission Computed Tomography (SPECT) is a functional brain imaging procedure that allows practitioners to measure the functional changes of gray matter structures based on regional cerebral blood flow (rCBF). The accurate diagnosis and treatment selection in psychiatry is challenging due to complex cases and frequent comorbidities. However, such complex neuropsychiatric conditions are increasingly benefitting from new treatment approaches, in addition to established medications. Among these are combination transcranial magnetic stimulation with ketamine infusions (CTK), hyperbaric oxygen therapy (HBOT) and perispinal administration of etanercept (PSE). This article provides readers with six case study examples that demonstrate how brain SPECT imaging can be used, both as a diagnostic tool, and as a potential biomarker for monitoring and evaluating novel treatments for patients with complex neuropsychiatric conditions. Six patients were assessed in our clinic and baseline brain SPECT imagesTourettes and a long history of alcohol were visually compared with SPECT images collected after periods of treatment with CTK or HBOT followed by PSE. This retrospective review demonstrates the clinical utility of these novel treatments and describes how SPECT imaging can complement standard diagnostic assessments. A novel display technique for SPECT images is described and we argue that SPECT imaging can be used for monitoring biomarker for clinical change.

Diagnostic Evaluation of 99mTc-TRODAT-1 SPECT in Parkinsonism

Background Parkinsonism as a group of movement disorders, exhibit similar clinical presentation. Therefore, clinically differentiating these diseases is difficult. We investigated the diagnostic value of 99m Tc-TRODAT-1 SPECT in this setting. Due to the fact that this modality has some limitations in imaging small organs like the sub-regions of basal ganglia, we also evaluated the use of anatomical MR imaging along with functional SPECT imaging in parkinsonism. Methods This follow-up diagnostic test evaluation study was performed with 40 patients with the clinical presentation of parkinsonism, and 10 healthy subjects as controls. After administration of the radiopharmaceutical, SPECT images were acquired, then co-registered on MRI. Uptake values were evaluated in basal ganglia semi-quantitatively. Results In this study, 99mTc-TRODAT-1 SPECT was able to differentiate essential tremor and healthy subjects from progressive supranuclear palsy (PSP) and Parkinson’s disease (PD) with a sensitivity of 76.47% and specificity of 100% at a cut-off of 0.53; however, findings were not significant in differentiation of PD from PSP (p ˃0.05), and the results were similar in SPECT and co-registered MRI/SPECT images. In evaluation of the uptake pattern in basal ganglia, the lateralization of decreased uptake was only seen in PD; and in PSP, the dysfunction was bilateral in all patients. Conclusion 99mTc-TRODAT-1 SPECT is sensitive and specific in diagnosing basal ganglia dysfunction; however, 99mTc-TRODAT-1 SPECT alone or co-registration on MRI are not adequate in differentiation of the etiologies of basal ganglia dysfunction.

Correlation of Dopaminergic Denervation and the Progression of Autonomic Dysfunctions in Different Clinical Subtypes of Parkinson’s Disease

Background. Autonomic dysfunctions occur in the early stage of Parkinson’s disease (PD) and impact the quality of life during the progression of the disease. In this study, we evaluated the serial progression of autonomic dysfunctions between different subtypes of a prospective PD cohort. Materials and Methods. From the Parkinson’s Progression Markers Initiative (PPMI) database, 325 PD patients (age: 61.2 ± 9.7, M : F = 215 : 110) were enrolled. Patients were subgrouped into tremor-dominant (TD), indeterminate, and postural instability and gait disorder (PIGD) subtypes. The progression of autonomic dysfunctions and dopaminergic denervation from I-123 FP-CIT SPECT images of each group were analyzed and compared at baseline, 12 months, 24 months, and 48 months of follow-up periods. Results. The SCOPA-AUT score of the indeterminate subtype was significantly higher than that of the TD subtype ( P < 0.05 ) at baseline and was significantly higher than that of both TD and PIGD subtypes ( P < 0.05 ) at 48 months. The indeterminate subtype had the most significant correlation between the aggravation of dopaminergic denervation in I-123 FP-CIT SPECT images and the increase of SCOPA-AUT scores during 48 months of follow-up (r = 0.56, P < 0.01 ). Conclusions. Autonomic dysfunctions were most severe in the indeterminate subtype throughout the 48 months of the follow-up period, with a significant correlation with dopaminergic denervation. We suggest a positive relationship between dopaminergic denervation and autonomic dysfunctions of the indeterminate subtype, beginning from the early stage of PD.

Data-driven identification of diagnostically useful extrastriatal signal in dopamine transporter SPECT using explainable AI

This study used explainable artificial intelligence for data-driven identification of extrastriatal brain regions that can contribute to the interpretation of dopamine transporter SPECT with 123I-FP-CIT in parkinsonian syndromes. A total of 1306 123I-FP-CIT-SPECT were included retrospectively. Binary classification as ‘reduced’ or ‘normal’ striatal 123I-FP-CIT uptake by an experienced reader served as standard-of-truth. A custom-made 3-dimensional convolutional neural network (CNN) was trained for classification of the SPECT images with 1006 randomly selected images in three different settings: “full image”, “striatum only” (3-dimensional region covering the striata cropped from the full image), “without striatum” (full image with striatal region removed). The remaining 300 SPECT images were used to test the CNN classification performance. Layer-wise relevance propagation (LRP) was used for voxelwise quantification of the relevance for the CNN-based classification in this test set. Overall accuracy of CNN-based classification was 97.0%, 95.7%, and 69.3% in the “full image”, “striatum only”, and “without striatum” setting. Prominent contributions in the LRP-based relevance maps beyond the striatal signal were detected in insula, amygdala, ventromedial prefrontal cortex, thalamus, anterior temporal cortex, superior frontal lobe, and pons, suggesting that 123I-FP-CIT uptake in these brain regions provides clinically useful information for the differentiation of neurodegenerative and non-neurodegenerative parkinsonian syndromes.

The Feasibility of Differentiating Lewy Body Dementia and Alzheimer’s Disease by Deep Learning Using ECD SPECT Images

The correct differential diagnosis of dementia has an important impact on patient treatment and follow-up care strategies. Tc-99m-ECD SPECT imaging, which is low cost and accessible in general clinics, is used to identify the two common types of dementia, Alzheimer’s disease (AD) and Lewy body dementia (LBD). Two-stage transfer learning technology and reducing model complexity based on the ResNet-50 model were performed using the ImageNet data set and ADNI database. To improve training accuracy, the three-dimensional image was reorganized into three sets of two-dimensional images for data augmentation and ensemble learning, then the performance of various deep learning models for Tc-99m-ECD SPECT images to distinguish AD/normal cognition (NC), LBD/NC, and AD/LBD were investigated. In the AD/NC, LBD/NC, and AD/LBD tasks, the AUC values were around 0.94, 0.95, and 0.74, regardless of training models, with an accuracy of 90%, 87%, and 71%, and F1 scores of 89%, 86%, and 76% in the best cases. The use of transfer learning and a modified model resulted in better prediction results, increasing the accuracy by 32% for AD/NC. The proposed method is practical and could rapidly utilize a deep learning model to automatically extract image features based on a small number of SPECT brain perfusion images in general clinics to objectively distinguish AD and LBD.