Try a new search

Format these results:

Searched for:



Total Results:


A Deep Learning Approach for Segmentation, Classification and Visualization of 3D High Frequency Ultrasound Images of Mouse Embryos

Qiu, Ziming; Xu, Tongda; Langerman, Jack; Das, William; Wang, Chuiyu; Nair, Nitin; Aristizabal, Orlando; Mamou, Jonathan; Turnbull, Daniel H; Ketterling, Jeffrey A; Wang, Yao
Segmentation and mutant classification of high-frequency ultrasound (HFU) mouse embryo brain ventricle (BV) and body images can provide valuable information for developmental biologists. However, manual segmentation and identification of BV and body requires substantial time and expertise. This paper proposes an accurate, efficient and explainable deep learning pipeline for automatic segmentation and classification of the BV and body. For segmentation, a two-stage framework is implemented. The first stage produces a low-resolution segmentation map, which is then used to crop a region of interest (ROI) around the target object and serve as the probability map of the auto-context input for the second-stage fine-resolution refinement network. The segmentation then becomes tractable on high-resolution 3D images without time-consuming sliding windows. The proposed segmentation method significantly reduces inference time (102.36 to 0.09 s/volume≈1000x faster) while maintaining high accuracy comparable to previous sliding-window approaches. Based on the BV and body segmentation map, a volumetric convolutional neural network (CNN) is trained to perform a mutant classification task. Through backpropagating the gradients of the predictions to the input BV and body segmentation map, the trained classifier is found to largely focus on the region where the Engrailed-1 (En1) mutation phenotype is known to manifest itself. This suggests that gradient backpropagation of deep learning classifiers may provide a powerful tool for automatically detecting unknown phenotypes associated with a known genetic mutation.
PMID: 33755564
ISSN: 1525-8955
CID: 4822582

CSF1R inhibition depletes tumor-associated macrophages and attenuates tumor progression in a mouse sonic Hedgehog-Medulloblastoma model

Tan, I-Li; Arifa, Raquel Duque Nascimento; Rallapalli, Harikrishna; Kana, Veronika; Lao, Zhimin; Sanghrajka, Reeti Mayur; Sumru Bayin, N; Tanne, Antoine; Wojcinski, Alexandre; Korshunov, Andrey; Bhardwaj, Nina; Merad, Miriam; Turnbull, Daniel H; Lafaille, Juan J; Joyner, Alexandra L
The immune microenvironment of tumors can play a critical role in promoting or inhibiting tumor progression depending on the context. We present evidence that tumor-associated macrophages/microglia (TAMs) can promote tumor progression in the sonic hedgehog subgroup of medulloblastoma (SHH-MB). By combining longitudinal manganese-enhanced magnetic resonance imaging (MEMRI) and immune profiling of a sporadic mouse model of SHH-MB, we found the density of TAMs is higher in the ~50% of tumors that progress to lethal disease. Furthermore, reducing regulatory T cells or eliminating B and T cells in Rag1 mutants does not alter SHH-MB tumor progression. As TAMs are a dominant immune component in tumors and are normally dependent on colony-stimulating factor 1 receptor (CSF1R), we treated mice with a CSF1R inhibitor, PLX5622. Significantly, PLX5622 reduces a subset of TAMs, prolongs mouse survival, and reduces the volume of most tumors within 4 weeks of treatment. Moreover, concomitant with a reduction in TAMs the percentage of infiltrating cytotoxic T cells is increased, indicating a change in the tumor environment. Our studies in an immunocompetent preclinical mouse model demonstrate TAMs can have a functional role in promoting SHH-MB progression. Thus, CSF1R inhibition could have therapeutic potential for a subset of SHH-MB patients.
PMID: 33159168
ISSN: 1476-5594
CID: 4664582

Scanner independent deep learning-based segmentation framework applied to mouse embryos

Chapter by: Aristizabal, Orlando; Turnbull, Daniel H.; Ketterling, Jeffrey A.; Wang, Yao; Qiu, Ziming; Xu, Tongda; Goldman, Hannah; Mamou, Jonathan
in: IEEE International Ultrasonics Symposium, IUS by
[S.l.] : IEEE Computer Society, 2020
pp. ?-?
ISBN: 9781728154480
CID: 4733912

Longitudinal MEMRI analysis of brain phenotypes in a mouse model of Niemann-Pick Type C disease

Rallapalli, Harikrishna; Darwin, Benjamin C; Toro-Montoya, Estefania; Lerch, Jason P; Turnbull, Daniel H
Niemann-Pick Type C (NPC) is a rare genetic disorder characterized by progressive cell death in various tissues, particularly in the cerebellar Purkinje cells, with no known cure. Mouse models for human NPC have been generated and characterized histologically, behaviorally, and using longitudinal magnetic resonance imaging (MRI). Previous imaging studies revealed significant brain volume differences between mutant and wild-type animals, but stopped short of making volumetric comparisons of the cerebellar sub-regions. In this study, we present longitudinal manganese-enhanced MRI (MEMRI) data from cohorts of wild-type, heterozygote carrier, and homozygote mutant NPC mice, as well as deformation-based morphometry (DBM) driven brain volume comparisons across genotypes, including the cerebellar cortex, white matter, and nuclei. We also present the first comparisons of MEMRI signal intensities, reflecting brain and cerebellum sub-regional Mn2+-uptake over time and across genotypes.
PMID: 32417449
ISSN: 1095-9572
CID: 4443622


Xu, Tongda; Qiu, Ziming; Das, William; Wang, Chuiyu; Langerman, Jack; Nair, Nitin; Aristizábal, Orlando; Mamou, Jonathan; Turnbull, Daniel H; Ketterling, Jeffrey A; Wang, Yao
The segmentation of the brain ventricle (BV) and body in embryonic mice high-frequency ultrasound (HFU) volumes can provide useful information for biological researchers. However, manual segmentation of the BV and body requires substantial time and expertise. This work proposes a novel deep learning based end-to-end auto-context refinement framework, consisting of two stages. The first stage produces a low resolution segmentation of the BV and body simultaneously. The resulting probability map for each object (BV or body) is then used to crop a region of interest (ROI) around the target object in both the original image and the probability map to provide context to the refinement segmentation network. Joint training of the two stages provides significant improvement in Dice Similarity Coefficient (DSC) over using only the first stage (0.818 to 0.906 for the BV, and 0.919 to 0.934 for the body). The proposed method significantly reduces the inference time (102.36 to 0.09 s/volume ≈1000x faster) while slightly improves the segmentation accuracy over the previous methods using slide-window approaches.
PMID: 33381278
ISSN: 1945-7928
CID: 4731902

MEMRI-based imaging pipeline for guiding preclinical studies in mouse models of sporadic medulloblastoma

Rallapalli, Harikrishna; Tan, I-Li; Volkova, Eugenia; Wojcinski, Alexandre; Darwin, Benjamin C; Lerch, Jason P; Joyner, Alexandra L; Turnbull, Daniel H
PURPOSE/OBJECTIVE:Genetically engineered mouse models of sporadic cancers are critical for studying tumor biology and for preclinical testing of therapeutics. We present an MRI-based pipeline designed to produce high resolution, quantitative information about tumor progression and response to novel therapies in mouse models of medulloblastoma (MB). METHODS:Sporadic MB was modeled in mice by inducing expression of an activated form of the Smoothened gene (aSmo) in a small number of cerebellar granule cell precursors. aSmo mice were imaged and analyzed at defined time-points using a 3D manganese-enhanced MRI-based pipeline optimized for high-throughput. RESULTS:A semi-automated segmentation protocol was established that estimates tumor volume in a time-frame compatible with a high-throughput pipeline. Both an empirical, volume-based classifier and a linear discriminant analysis-based classifier were tested to distinguish progressing from nonprogressing lesions at early stages of tumorigenesis. Tumor centroids measured at early stages revealed that there is a very specific location of the probable origin of the aSmo MB tumors. The efficacy of the manganese-enhanced MRI pipeline was demonstrated with a small-scale experimental drug trial designed to reduce the number of tumor associated macrophages and microglia. CONCLUSION/CONCLUSIONS:Our results revealed a high level of heterogeneity between tumors within and between aSmo MB models, indicating that meaningful studies of sporadic tumor progression and response to therapy could not be conducted without an imaging-based pipeline approach.
PMID: 31403226
ISSN: 1522-2594
CID: 4041832

Somatic mutations of chromatin regulator KMT2D in cerebellar precursors influences shhmedulloblastoma tumorigenesis [Meeting Abstract]

Sanghrajka, R; Tan, I -L; Wojcinski, A; Rallapalli, H; Turnbull, D; Ge, K; Joyner, A
Medulloblastoma (MB), the most common malignant pediatric brain tumor, is a classic example of dysregulation of developmental pathways leading to tumorogenesis. Despite advancements in multi-modal therapies, most patients suffer from long-term neurocognitive and neuroendocrine disabilities. The Sonic Hedgehog subgroup of MB (SHH-MB) accounts for ~30% of all cases and originates from ATOH1+ cerebellar granule cell precursors (GCPs). Experimental data in mice has shown that activating mutations in the SHH pathway induce tumors only in rare GCPs, suggesting that additional mutations and epigenetic changes are required to influence tumor progression. The KMT2D gene, encoding the histone-lysine N-methyltransferase 2D, is amongst the ten most frequently mutated genes in MB, with somatic mutations seen in ~15% of all SHH-MB patients. We developed sporadic mouse models of SHH-MB with a low penetrance to enable studies of secondary mutations (Tan, PNAS, 2018). Immunofluorescence staining for KMT2D on early-stage SHH-MB lesions, mid-stage and late-stage tumors revealed that a subset of lesions/tumors (16/98) do not express KMT2D and are negative for H3K4me3. Interestingly, P53 and KMT2D expression showed a positive correlation in ~94% of tumors/lesions and NeuN and KMT2D showed a positive correlation in ~92% of tumors/lesions. In order to determine the roles for KMT2D in GCP proliferation and differentiation, and uncover whether and how KMT2D promotes SHH-MB tumorigenesis, we are using genetic mouse-models whereby Kmt2d is heterozygously or homozygously deleted alone, or in conjunction with activation of the SHH pathway. Mice with SHH-MB tumors expressing SmoM2 and a loss of Kmt2d develop aggressive tumors at high penetrance, with metastatic leptomeningeal spread in the brain stem and spinal cord. Thus, loss of Kmt2d increases SHH-MB tumor progression and leads to malignancy. Ongoing studies are determining how the chromatin landscape and gene expression are changed when Kmt2d is deleted in GCPs
ISSN: 1523-5866
CID: 4388182

Automatic Mouse Embryo Brain Ventricle & Body Segmentation and Mutant Classification from Ultrasound Data Using Deep Learning

Chapter by: Qiu, Ziming; Nair, Nitin; Langerman, Jack; Aristizabal, Orlando; Mamou, Jonathan; Turnbull, Daniel H.; Ketterling, Jeffrey A.; Wang, Yao
in: IEEE International Ultrasonics Symposium, IUS by
[S.l.] : IEEE Computer, 2019
pp. 12-15
ISBN: 9781728145969
CID: 4332082

Cerebellar folding is initiated by mechanical constraints on a fluid-like layer without a cellular pre-pattern

Lawton, Andrew K; Engstrom, Tyler; Rohrbach, Daniel; Omura, Masaaki; Turnbull, Daniel H; Mamou, Jonathan; Zhang, Teng; Schwarz, J M; Joyner, Alexandra L
Models based in differential expansion of elastic material, axonal constraints, directed growth, or multi-phasic combinations have been proposed to explain brain folding. However, the cellular and physical processes present during folding have not been defined. We used the murine cerebellum to challenge folding models with in vivo data. We show that at folding initiation differential expansion is created by the outer layer of proliferating progenitors expanding faster than the core. However, the stiffness differential, compressive forces, and emergent thickness variations required by elastic material models are not present. We find that folding occurs without an obvious cellular pre-pattern, that the outer layer expansion is uniform and fluid-like, and that the cerebellum is under radial and circumferential constraints. Lastly, we find that a multi-phase model incorporating differential expansion of a fluid outer layer and radial and circumferential constraints approximates the in vivo shape evolution observed during initiation of cerebellar folding.
PMID: 30990415
ISSN: 2050-084x
CID: 3810482

Granule cell precursors in the lateral cerebellum are preferentially sensitive to elevated sonic hedgehog signaling and formation of medulloblastoma [Meeting Abstract]

Tan, I L; Wojcinski, A; Rallapalli, H; Lao, Z; Sanighrajka, R M; Stephen, D; Volkova, E; Korshunov, A; Remke, M; Taylor, M D; Turnbull, D H; Joyner, A L
Objective: Granule cell precursors (GCPs) are a sonic hedgehog (SHH)- dependent progenitor population in the developing cerebellum and the main cell of origin for the SHH subgroup of medulloblastoma (MB). Unlike other subgroups of MB, SHH-MBs occur preferentially in the lateral cerebellum (hemispheres) and have four main driver mutations. We studied whether the timing or type of mutation affects tumor location and identified factors influencing SHH-MB progression.
Method(s): We analyzed the association between type of mutation and tumor location in 38 SHH-MB patient samples. To generate sporadic mouse models of SHH-MB, inducible recombinases were used to express a constitutive activate SMO receptor (SmoM2) or delete Ptch1 in only scattered GCPs. Tumor location, expression profiles and GCP behaviors were analyzed in the models.
Result(s): Our analysis of patient data indicates that adult tumors with SMO mutations form more specifically in the hemispheres than those with PTCH1 mutations. Using sporadic mouse models, we found that regardless of the number of GCPs mutated, timing or type of mutation, tumors developed almost exclusively in the hemispheres with SmoM2-mutants showing a stronger specificity. We further uncovered that GCPs in the hemispheres are more susceptible to high level SHH signaling compared to GCPs in the medial cerebellum (vermis), as more mutant cells in the hemisphere remain undifferentiated and show increased tumorigenicity when transplanted. We also identified location-specific gene expression profiles, and found that deletion of the genes most highly expressed in the hemispheres or vermis showed opposing effects on GCP differentiation.
Conclusion(s): We found that GCPs respond differentially to two driver mutations and a subset of GCPs is more susceptible to high level of SHH signaling as well as tumors formation. We redefined themain cell of origin by showing that GCPs are heterogeneous with molecularly distinct populations based on their location
ISSN: 1473-4230
CID: 3703462