Cluster division is a critical issue in fluorescence microscopy-based analytical cytology when preparation protocols do not provide appropriate separation of objects. Overlooking clustered nuclei and analyzing only isolated nuclei may dramatically increase analysis time or affect the statistical validation of the results. Automatic segmentation of clustered nuclei requires the implementation of specific image segmentation tools. Most algorithms are inspired by one of the two following strategies: 1) cluster division by the detection of internuclei gradients; or 2) division by definition of domains of influence (geometrical approach). Both strategies lead to completely different implementations, and usually algorithms based on a single view strategy fail to correctly segment most clustered nuclei, or perform well just for a specific type of sample. An algorithm based on morphological watersheds has been implemented and tested on the segmentation of microscopic nuclei clusters. This algorithm provides a tool that can be used for the implementation of both gradient- and domain-based algorithms, and, more importantly, for the implementation of mixed (gradient- and shape-based) algorithms. Using this algorithm, almost 90% of the test clusters were correctly segmented in peripheral blood and bone marrow preparations. The algorithm was valid for both types of samples, using the appropriate markers and transformations. Cytometry 28:289–297, 1997.

We present a study of multiple sclerosis segmentation algorithms conducted at the international MICCAI 2016 challenge. This challenge was operated using a new open-science computing infrastructure. This allowed for the automatic and independent evaluation of a large range of algorithms in a fair and completely automatic manner. This computing infrastructure was used to evaluate thirteen methods of MS lesions segmentation, exploring a broad range of state-of-theart algorithms, against a high-quality database of 53 MS cases coming from four centers following a common definition of the acquisition protocol. Each case was annotated manually by an unprecedented number of seven different experts. Results of the challenge highlighted that automatic algorithms, including the recent machine learning methods (random forests, deep learning, …), are still trailing human expertise on both detection and delineation criteria. In addition, we demonstrate that computing a statistically robust consensus of the algorithms performs closer to human expertise on one score (segmentation) although still trailing on detection scores.

This work describes a systematic evaluation of several autofocus functions used for analytical fluorescent image cytometry studies of counterstained nuclei. Focusing is the first step in the automatic fluorescence in situ hybridization analysis of cells. Thirteen functions have been evaluated using qualitative and quantitative procedures. For the last of these procedures a figure‐of‐merit (FOM) is defined and proposed. This new FOM takes into account five important features of the focusing function. Our results show that functions based on correlation measures have the best performance for this type of image.

Attenuation correction in hybrid PET/MR scanners is still a challenging task. This paper describes a methodology for synthesizing a pseudo-CT volume from a single T1 weighted volume, thus allowing us to create correct AC maps. Methods: We propose a fast pseudo-CT generation from a patient-specific MRI T1-weighted image using a group-wise patch-based approach and a MRI-CT atlas dictionary. For every voxel in the input MR image, we compute the similarity of the patch containing that voxel with the patches of all MR images in the database, that lie in a certain anatomical neighborhood. The pseudo-CT is obtained as a local weighted linear combination of the CT values of the corresponding patches. The algorithm was implemented in a Graphical Processing Unit. Results: We evaluated our method both qualitatively and quantitatively for PET/MR correction. The approach performed successfully in all cases considered. We compared the SUVs of the PET image obtained after attenuation correction using the patient specific CT volume, and using the corresponding computed pseudo-CT volume. The voxel by voxel correlation between SUV obtained with both methods was very high (R2=0.9980, p<0.0001), while the Bland-Altman test showed that the average of the differences was very low (0.0006±0.0594). A region of interest analysis was also performed. The correlation between mean SUV value and maximum SUV value of every region was very high (R2=0.9989, p<0.0001 and R2=0.9904, p<0.0001, respectively). Conclusion: Results indicate how our method can accurately approximate the patient-specific CT and may serve as a potential solution for accurate attenuation correction in hybrid PET/MR systems. The quality of the corrected PET using our pseudo-CT is comparable to having acquired a patient-specific CT, thus, improving the results obtained with the UTE-based attenuation correction maps currently used in the mMR scanner. The GPU implementation led to a substantial decrease in computational time making the approach suitable for real applications.