Propagating Uncertainty Across Cascaded Medical Imaging Tasks for Improved Deep Learning Inference.

Although deep networks have been shown to perform very well on a variety of medical imaging tasks, inference in the presence of pathology presents several challenges to common models. These challenges impede the integration of deep learning models into real clinical workflows, where the customary process of cascading deterministic outputs from a sequence of image-based inference steps (e.g. registration, segmentation) generally leads to an accumulation of errors that impacts the accuracy of downstream inference tasks. In this paper, we propose that by embedding uncertainty estimates across cascaded inference tasks, performance on the downstream inference tasks should be improved. We demonstrate the effectiveness of the proposed approach in three different clinical contexts: (i) We demonstrate that by propagating T2 weighted lesion segmentation results and their associated uncertainties, subsequent T2 lesion detection performance is improved when evaluated on a proprietary large-scale, multi-site, clinical trial dataset acquired from patients with Multiple Sclerosis. (ii) We show an improvement in brain tumour segmentation performance when the uncertainty map associated with a synthesised missing MR volume is provided as an additional input to a follow-up brain tumour segmentation network, when evaluated on the publicly available BraTS-2018 dataset. (iii) We show that by propagating uncertainties from a voxel-level hippocampus segmentation task, the subsequent regression of the Alzheimer's disease clinical score is improved.

All-optical approaches to studying psychiatric disease.

Improvements in all-optical means of monitoring and manipulating neural activity have generated new ways of studying psychiatric disease. The combination of calcium imaging techniques with optogenetics to concurrently record and manipulate neural activity has been used to create new disease models that link distinct circuit abnormalities to specific disease dimensions. These approaches represent a new path towards the development of more effective treatments, as they allow researchers to identify circuit manipulations that normalize pathological network activity. In this review we highlight the utility of all-optical approaches to generate new psychiatric disease models where the specific circuit abnormalities associated with disease symptomology can be assessed in vivo and in response to manipulations designed to normalize disease states. We then outline the principles underlying all-optical interrogations of neural circuits and discuss practical considerations for experimental design.