Object segmentation and structure localization are important steps in automated image analysis pipelines for microscopy images. We present a convolution neural network (CNN) based deep learning architecture for segmentation of objects in microscopy images. The proposed network can be used to segment cells, nuclei and glands in fluorescence microscopy and histology images after slight tuning of input parameters. The network trains at multiple resolutions of the input image, connects the intermediate layers for better localization and context and generates the output using multi-resolution deconvolution filters. The extra convolutional layers which bypass the max-pooling operation allow the network to train for variable input intensities and object size and make it robust to noisy data. We compare our results on publicly available data sets and show that the proposed network outperforms recent deep learning algorithms.

Purpose: Atu027 is a novel liposomal RNA interference therapeutic that includes a short-interfering RNA (siRNA), which silences expression of protein kinase N3 in the vascular endothelium. Atu027 has previously been shown to inhibit local tumor invasion as well as lymph node and pulmonary metastasis in mouse cancer models. This first-in-human study aimed to assess the safety, tolerability, and pharmacokinetics of Atu027 while evaluating therapeutic effects on both primary tumors and metastatic lesions. Patients and Methods: Thirty-four patients with advanced solid tumors received 10 escalating doses of Atu027 without premedication, as one single followed by eight intravenous infusions twice per week during a 28-day cycle. Response was monitored by computed tomography/magnetic resonance imaging at baseline, at the end of treatment (EoT), and at final follow-up (EoS), and was assessed according to RECIST. Results: Atu027 was well tolerated up to dose levels of 0.336 mg/kg; most adverse events (AEs) were low-grade toxicities (grade 1 or 2). No maximum tolerated dose was reached. Plasma levels of siRNA strands and lipids were dose proportional, peaking during 4-hour infusion. Disease stabilization was achieved in 41% of patients at EoT (n = 14 of 34 treated patients); eight patients had stable disease at EoS, and some experienced complete or partial regression of metastases. sFLT1 (soluble variant of vascular endothelial growth factor receptor-1) decreased from pretreatment levels in most patients after dose levels 04 to 10. Conclusion: Atu027 was safe in patients with advanced solid tumors, with 41% of patients having stable disease for at least 8 weeks. In view of these results, further clinical trials have been initiated, and sFLT1 will be investigated as a potential biomarker.