Explainable AI to improve acceptance of convolutional neural networks for automatic classification of dopamine transporter SPECT in the diagnosis of clinically uncertain parkinsonian syndromes.

PURPOSE: Deep convolutional neural networks (CNN) provide high accuracy for automatic classification of dopamine transporter (DAT) SPECT images. However, CNN are inherently black-box in nature lacking any kind of explanation for their decisions. This limits their acceptance for clinical use. This study tested layer-wise relevance propagation (LRP) to explain CNN-based classification of DAT-SPECT in patients with clinically uncertain parkinsonian syndromes. METHODS: The study retrospectively included 1296 clinical DAT-SPECT with visual binary interpretation as "normal" or "reduced" by two experienced readers as standard-of-truth. A custom-made CNN was trained with 1008 randomly selected DAT-SPECT. The remaining 288 DAT-SPECT were used to assess classification performance of the CNN and to test LRP for explanation of the CNN-based classification. RESULTS: Overall accuracy, sensitivity, and specificity of the CNN were 95.8%, 92.8%, and 98.7%, respectively. LRP provided relevance maps that were easy to interpret in each individual DAT-SPECT. In particular, the putamen in the hemisphere most affected by nigrostriatal degeneration was the most relevant brain region for CNN-based classification in all reduced DAT-SPECT. Some misclassified DAT-SPECT showed an "inconsistent" relevance map more typical for the true class label. CONCLUSION: LRP is useful to provide explanation of CNN-based decisions in individual DAT-SPECT and, therefore, can be recommended to support CNN-based classification of DAT-SPECT in clinical routine. Total computation time of 3 s is compatible with busy clinical workflow. The utility of "inconsistent" relevance maps to identify misclassified cases requires further investigation.

Automated and robust organ segmentation for 3D-based internal dose calculation.

PURPOSE: In this work, we address image segmentation in the scope of dosimetry using deep learning and make three main contributions: (a) to extend and optimize the architecture of an existing convolutional neural network (CNN) in order to obtain a fast, robust and accurate computed tomography (CT)-based organ segmentation method for kidneys and livers; (b) to train the CNN with an inhomogeneous set of CT scans and validate the CNN for daily dosimetry; and (c) to evaluate dosimetry results obtained using automated organ segmentation in comparison with manual segmentation done by two independent experts. METHODS: We adapted a performant deep learning approach using CT-images to delineate organ boundaries with sufficiently high accuracy and adequate processing time. The segmented organs were consequently used as binary masks for further convolution with a point spread function to retrieve the activity values from quantitatively reconstructed SPECT images for "volumetric"/3D dosimetry. The resulting activities were used to perform dosimetry calculations with the kidneys as source organs. RESULTS: The computational expense of the algorithm was sufficient for clinical daily routine, required minimum pre-processing and performed with acceptable accuracy a Dice coefficient of [Formula: see text] for liver segmentation and of [Formula: see text] for kidney segmentation, respectively. In addition, kidney self-absorbed doses calculated using automated segmentation differed by [Formula: see text] from dosimetry performed by two medical physicists in 8 patients. CONCLUSION: The proposed approach may accelerate volumetric dosimetry of kidneys in molecular radiotherapy with 177Lu-labelled radiopharmaceuticals such as 177Lu-DOTATOC. However, even though a fully automated segmentation methodology based on CT images accelerates organ segmentation and performs with high accuracy, it does not remove the need for supervision and corrections by experts, mostly due to misalignments in the co-registration between SPECT and CT images. Trial registration EudraCT, 2016-001897-13. Registered 26.04.2016, www.clinicaltrialsregister.eu/ctr-search/search?query=2016-001897-13 .

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.