Distribution pattern of iron deposition in the basal ganglia of different motor subtypes of Parkinson's disease.

OBJECTIVE: The quantitative susceptibility mapping (QSM) technique was used to analyze the distribution pattern of iron deposition in the basal ganglia region of patients with motor subtypes of Parkinson's disease (PD) and to explore the difference in iron content in the basal ganglia region of PD motor subtypes on the major motor symptomatic side. METHODS: The study included 76 patients with PD and 37 healthy controls (HC). Patients with PD were divided into two groups: postural instability/gait disorder (PIGD)(n = 48), and tremor dominance (TD)(n = 28). We classified patients with PD according to the side of the major motor symptoms as left PIGD (n = 23), left TD (n = 14), right PIGD (n = 25), and right TD (n = 14). All subjects underwent brain magnetic resonance scanning to obtain QSM and susceptibility values in the corresponding regions of interest (ROI). RESULTS: (1) Compared with the HC, the bilateral SN in the PD-PIGD and TD group showed greater susceptibility values. The susceptibility values in the left CN, bilateral PUT were also greater in the PD-PIGD group than the HC. (2) Compared with the TD, the left PUT susceptibility values were greater in the PIGD group, especially in patients whose major symptomatic side were on the right limb. (3) Correlation analysis showed that in the PD group, bilateral SN was positively correlated with the unified Parkinson's disease rating scale III part scores of the Movement Disorder Society (MDS-UPDRS III) and the Hoehn-Yahr stage. Bilateral dentate nucleus (DN) susceptibility values were significantly positively correlated with TD scores, and left PUT susceptibility values were positively correlated with PIGD scores. The left SN within the PIGD group was positively correlated with the PIGD score. CONCLUSION: There were different iron deposition patterns in the basal ganglia between the PD-PIGD and TD groups. There also seems to be a difference in iron deposition in PD motor subtypes on different major motor symptom sides.

Automatic Classification of Mass Shape and Margin on Mammography with Artificial Intelligence: Deep CNN Versus Radiomics.

The purpose of this study is to test the feasibility for deep CNN-based artificial intelligence methods for automatic classification of the mass margin and shape, while radiomic feature-based machine learning methods were also implemented in this study as baseline and for comparison study. In this retrospective study, 596 patients with breast mass that underwent mammography from 4 hospitals were enrolled from January 2012 to October 2019. Margin and shape of each mass were annotated according to BI-RADS by 2 experienced radiologists. Deep CNN-based AI was implemented for the classification task based on Resnet50. Balanced sampler and CBAM were also used to improve the performance of the Deep CNNs. As comparison, image texture features were extracted and then dimensionality reduction methods (such as PCA, ICA) and classical classifiers (such as SVM, DT, KNN) were used for classification task. Based on Python programming software, accuracy (ACC) was used to evaluate the performance of the model, and the model with the highest ACC value was selected. Deep CNN based on Resnet50 with balanced sampler and CBAM achieved the best performance for both margin and shape classification, with ACC of 0.838 and 0.874, respectively. For the radiomics-based machine learning, the best performance for margin was achieved as 0.676 by the combination of FA + RF, while the best performance for shape was 0.802 by the combination of PCA + MLP. The feasibility for automatic classification with coarse labeling of the mass shape and margin was testified with the deep CNN-based AI methods, while radiomic feature-based machine learning methods achieved inferior classification results.