Improving interobserver agreement and performance of deep learning models for segmenting acute ischemic stroke by combining DWI with optimized ADC thresholds

Artificial intelligence in ischemic stroke images: current applications and future directions

This paper reviews the current research progress in the application of Artificial Intelligence (AI) based on ischemic stroke imaging, analyzes the main challenges, and explores future research directions. This study emphasizes the application of AI in areas such as automatic segmentation of infarct areas, detection of large vessel occlusion, prediction of stroke outcomes, assessment of hemorrhagic transformation risk, forecasting of recurrent ischemic stroke risk, and automatic grading of collateral circulation. The research indicates that Machine Learning (ML) and Deep Learning (DL) technologies have tremendous potential for improving diagnostic accuracy, accelerating disease identification, and predicting disease progression and treatment responses. However, the clinical application of these technologies still faces challenges such as limitations in data volume, model interpretability, and the need for real-time monitoring and updating. Additionally, this paper discusses the prospects of applying large language models, such as the transformer architecture, in ischemic stroke imaging analysis, emphasizing the importance of establishing large public databases and the need for future research to focus on the interpretability of algorithms and the comprehensiveness of clinical decision support. Overall, AI has significant application value in the management of ischemic stroke; however, existing technological and practical challenges must be overcome to achieve its widespread application in clinical practice.

Long period changes of hippocampal diffusion kurtosis imaging and its correlation with cognitive dysfunction after incomplete cerebral ischemia-reperfusion in rats

This study aims to summarize the changes of functional diffusion kurtosis imaging (DKI) parameters in the bilateral hippocampal CA1 region of the hemorrhagic shock reperfusion (HSR) model of rats and their correlation with cognitive dysfunction. Adult male Sprague-Dawley rats (9-10 weeks of age, weighing 350-400 g) were randomized into the HSR group (n = 30) and the sham-operated group (Sham) (n = 30). Rats in the HSR group and the Sham group were subdivided into five time points (1, 2, 4, 8, and 12 weeks) for examination. Diffusion kurtosis imaging (DKI) was performed. Cognitive dysfunction was analyzed by the Morris Water Maze. The correlation between the DKI parameters and cognitive dysfunction was analyzed by the Spearman correlation. In the HSR group, the values of axial kurtosis (Ka), radial kurtosis (Kr), and mean kurtosis (MK) in the bilateral hippocampal CA1 of rats at 1, 2, 4, 8 and 12 weeks after the surgery were significantly higher. The rats in the HSR group had significantly longer escape latency than in the Sham group. The rats in the HSR group had significantly shorter time and shorter distance in target quadrant than those in the Sham group. The escape latency had positive correlation with MK, Ka, and Kr. The distance and the time in target quadrant had negative correlation with MK, Ka, and Kr. The parameters get from the DKI could accurately evaluate the abnormal blood perfusion and microstructure changes in hippocampal CA1 area of the incomplete cerebral ischemia reperfusion rats induced by HSR. MK, Ka, and Kr values could reflect the decreased learning and memory ability in HSR rat model.

MSA-YOLOv5: Multi-scale attention-based YOLOv5 for automatic detection of acute ischemic stroke from multi-modality MRI images

BACKGROUND AND OBJECTIVE

Acute ischemic stroke (AIS) is a common neurological disorder characterized by the sudden onset of cerebral ischemia, leading to functional impairments. Swift and precise detection of AIS lesions is crucial for stroke diagnosis and treatment but poses a significant challenge. This study aims to leverage multimodal fusion technology to combine complementary information from various modalities, thereby enhancing the detection performance of AIS target detection models.

METHODS

In this retrospective study of AIS, we collected data from 316 AIS patients and created a multi-modality magnetic resonance imaging (MRI) dataset. We propose a Multi-Scale Attention-based YOLOv5 (MSA-YOLOv5), targeting challenges such as small lesion size and blurred borders at low resolutions. Specifically, we augment YOLOv5 with a prediction head to detect objects at various scales. Next, we replace the original prediction head with a Multi-Scale Swin Transformer Prediction Head (MS-STPH), which reduces computational complexity to linear levels and enhances the ability to detect small lesions. We incorporate a Second-Order channel attention (SOCA) module to adaptively rescale channel features by employing second-order feature statistics for more discriminative representations. Finally, we further validate the effectiveness of our method using the ISLES 2022 dataset.

RESULTS

On our in-house AIS dataset, MSA-YOLOv5 achieves a 79.0% mAP0.5, substantially surpassing other single-stage models. Compared to two-stage models, it maintains a comparable performance level while significantly reducing the number of parameters and resolution. On the ISLES 2022 dataset, MSA-YOLOv5 attains an 80.0% mAP0.5, outperforming other network models by a considerable margin. MS-STPH and SOCA modules can significantly increase mAP0.5 by 2.7% and 1.9%, respectively. Visualization interpretability results show that the proposed MSA-YOLOv5 restricts high attention in the small regions of AIS lesions.

CONCLUSIONS

The proposed MSA-YOLOv5 is capable of automatically and effectively detecting acute ischemic stroke lesions in multimodal images, particularly for small lesions and artifacts. Our enhanced model reduces the number of parameters while improving detection accuracy. This model can potentially assist radiologists in providing more accurate diagnosis, and enable clinicians to develop better treatment plans.

The role of input imaging combination and ADC threshold on segmentation of acute ischemic stroke lesion using U-Net

BACKGROUND

To evaluate the effect of the weighting of input imaging combo and ADC threshold on the performance of the U-Net and to find an optimized input imaging combo and ADC threshold in segmenting acute ischemic stroke (AIS) lesion.

METHODS

This study retrospectively enrolled a total of 212 patients having AIS. Four combos, including ADC-ADC-ADC (AAA), DWI-ADC-ADC (DAA), DWI-DWI-ADC (DDA), and DWI-DWI-DWI (DDD), were used as input images, respectively. Three ADC thresholds including 0.6, 0.8 and 1.8 × 10-3 mm2/s were applied. Dice similarity coefficient (DSC) was used to evaluate the segmentation performance of U-Nets. Nonparametric Kruskal-Wallis test with Tukey-Kramer post-hoc tests were used for comparison. A p < .05 was considered statistically significant.

RESULTS

The DSC significantly varied among different combos of images and different ADC thresholds. Hybrid U-Nets outperformed uniform U-Nets at ADC thresholds of 0.6 × 10-3 mm2/s and 0.8 × 10-3 mm2/s (p < .001). The U-Net with imaging combo of DDD had segmentation performance similar to hybrid U-Nets at an ADC threshold of 1.8 × 10-3 mm2/s (p = .062 to 1). The U-Net using the imaging combo of DAA at the ADC threshold of 0.6 × 10-3 mm2/s achieved the highest DSC in the segmentation of AIS lesion.

CONCLUSIONS

The segmentation performance of U-Net for AIS varies among the input imaging combos and ADC thresholds. The U-Net is optimized by choosing the imaging combo of DAA at an ADC threshold of 0.6 × 10-3 mm2/s in segmentating AIS lesion with highest DSC.

KEY POINTS

• Segmentation performance of U-Net for AIS differs among input imaging combos. • Segmentation performance of U-Net for AIS differs among ADC thresholds. • U-Net is optimized using DAA with ADC = 0.6 × 10-3 mm2/s.

Correlation of mean apparent diffusion coefficient (ADC) and maximal standard uptake value (SUVmax) evaluated by diffusion-weighted MRI and 18F-FDG-PET/CT in children with Hodgkin lymphoma: a feasibility study

BACKGROUND

The objective was to analyse if magnetic resonance imaging (MRI) can act as a non-radiation exposure surrogate for (18)F-Fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) in children with histologically confirmed Hodgkin lymphoma (HL) before treatment. This was done by analysing a potential correlation between apparent diffusion coefficient (ADC) in MRI and the maximum standardized uptake value (SUVmax) in FDG-PET/CT.

PATIENTS AND METHODS

Seventeen patients (six female, eleven male, median age: 16 years, range: 12-20 years) with histologically confirmed HL were retrospectively analysed. The patients underwent both MRI and (18)F-FDG PET/CT before the start of treatment. (18)F-FDG PET/CT data and correlating ADC maps in MRI were collected. For each HL-lesion two readers independently evaluated the SUVmax and correlating meanADC.

RESULTS

The seventeen patients had a total of 72 evaluable lesions of HL and there was no significant difference in the number of lesions between male and female patients (median male: 15, range: 12-19 years, median female: 17 range: 12-18 years, p = 0.021). The mean duration between MRI and PET/CT was 5.9 ± 5.3 days. The inter-reader agreement as assessed by the intraclass correlation coefficient (ICC) was excellent (ICC = 0.98, 95% CI: 0.97-0.99). The correlated SUVmax and meanADC of all 17 patients (ROIs n = 72) showed a strong negative correlation of -0.75 (95% CI: -0.84, - -0.63, p = 0.001). Analysis revealed a difference in the correlations of the examination fields. The correlated SUVmax and meanADC showed a strong correlation at neck and thoracal examinations (neck: -0.83, 95% CI: -0.93, - -0.63, p < 0.0001, thoracal: -0.82, 95% CI: -0.91, - -0.64, p < 0.0001) and a fair correlation at abdominal examinations of -0.62 (95% CI: -0.83, - -0.28, p = 0.001).

CONCLUSIONS

SUVmax and meanADC showed a strong negative correlation in paediatric HL lesions. The assessment seemed robust according to inter-reader agreements. Our results suggest that ADC maps and meanADC have the potential to replace PET/CT in the analysis of disease activity in paediatric Hodgkin lymphoma patients. This may help reduce the number of PET/CT examinations and decrease radiation exposure to children.

Improving performance of deep learning models using 3.5D U-Net via majority voting for tooth segmentation on cone beam computed tomography

Deep learning allows automatic segmentation of teeth on cone beam computed tomography (CBCT). However, the segmentation performance of deep learning varies among different training strategies. Our aim was to propose a 3.5D U-Net to improve the performance of the U-Net in segmenting teeth on CBCT. This study retrospectively enrolled 24 patients who received CBCT. Five U-Nets, including 2Da U-Net, 2Dc U-Net, 2Ds U-Net, 2.5Da U-Net, 3D U-Net, were trained to segment the teeth. Four additional U-Nets, including 2.5Dv U-Net, 3.5Dv5 U-Net, 3.5Dv4 U-Net, and 3.5Dv3 U-Net, were obtained using majority voting. Mathematical morphology operations including erosion and dilation (E&D) were applied to remove diminutive noise speckles. Segmentation performance was evaluated by fourfold cross validation using Dice similarity coefficient (DSC), accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV). Kruskal-Wallis test with post hoc analysis using Bonferroni correction was used for group comparison. P < 0.05 was considered statistically significant. Performance of U-Nets significantly varies among different training strategies for teeth segmentation on CBCT (P < 0.05). The 3.5Dv5 U-Net and 2.5Dv U-Net showed DSC and PPV significantly higher than any of five originally trained U-Nets (all P < 0.05). E&D significantly improved the DSC, accuracy, specificity, and PPV (all P < 0.005). The 3.5Dv5 U-Net achieved highest DSC and accuracy among all U-Nets. The segmentation performance of the U-Net can be improved by majority voting and E&D. Overall speaking, the 3.5Dv5 U-Net achieved the best segmentation performance among all U-Nets.