Keras R-CNN: library for cell detection in biological images using deep neural networks.

BACKGROUND: A common yet still manual task in basic biology research, high-throughput drug screening and digital pathology is identifying the number, location, and type of individual cells in images. Object detection methods can be useful for identifying individual cells as well as their phenotype in one step. State-of-the-art deep learning for object detection is poised to improve the accuracy and efficiency of biological image analysis. RESULTS: We created Keras R-CNN to bring leading computational research to the everyday practice of bioimage analysts. Keras R-CNN implements deep learning object detection techniques using Keras and Tensorflow ( https://github.com/broadinstitute/keras-rcnn ). We demonstrate the command line tool's simplified Application Programming Interface on two important biological problems, nucleus detection and malaria stage classification, and show its potential for identifying and classifying a large number of cells. For malaria stage classification, we compare results with expert human annotators and find comparable performance. CONCLUSIONS: Keras R-CNN is a Python package that performs automated cell identification for both brightfield and fluorescence images and can process large image sets. Both the package and image datasets are freely available on GitHub and the Broad Bioimage Benchmark Collection.

Monitoring the Efficacy of Chloroquine-Primaquine Therapy for Uncomplicated Plasmodium vivax Malaria in the Main Transmission Hot Spot of Brazil.

Emerging Plasmodium vivax resistance to chloroquine (CQ) may undermine malaria elimination efforts in South America. CQ-resistant P. vivax has been found in the major port city of Manaus but not in the main malaria hot spots across the Amazon Basin of Brazil, where CQ is routinely coadministered with primaquine (PQ) for radical cure of vivax malaria. Here we randomly assigned 204 uncomplicated vivax malaria patients from Juruá Valley, northwestern Brazil, to receive either sequential (arm 1) or concomitant (arm 2) CQ-PQ treatment. Because PQ may synergize the blood schizontocidal effect of CQ and mask low-level CQ resistance, we monitored CQ-only efficacy in arm 1 subjects, who had PQ administered only at the end of the 28-day follow-up. We found adequate clinical and parasitological responses in all subjects assigned to arm 2. However, 2.2% of arm 1 patients had microscopy-detected parasite recrudescences at day 28. When PCR-detected parasitemias at day 28 were considered, response rates decreased to 92.1% and 98.8% in arms 1 and 2, respectively. Therapeutic CQ levels were documented in 6 of 8 recurrences, consistent with true CQ resistance in vivo In contrast, ex vivo assays provided no evidence of CQ resistance in 49 local P. vivax isolates analyzed. CQ-PQ coadministration was not found to potentiate the antirelapse efficacy of PQ over 180 days of surveillance; however, we suggest that larger studies are needed to examine whether and how CQ-PQ interactions, e.g., CQ-mediated inhibition of PQ metabolism, modulate radical cure efficacy in different P. vivax-infected populations. (This study has been registered at ClinicalTrials.gov under identifier NCT02691910.).