scholarly journals Reconstructing cell cycle and disease progression using deep learning

2017 ◽  
Vol 8 (1) ◽  
Author(s):  
Philipp Eulenberg ◽  
Niklas Köhler ◽  
Thomas Blasi ◽  
Andrew Filby ◽  
Anne E. Carpenter ◽  
...  
Keyword(s):  
Cell Cycle ◽  
Deep Learning ◽  
Disease Progression

scholarly journals Reconstructing cell cycle and disease progression using deep learning

2016 ◽  
Author(s):  
Philipp Eulenberg ◽  
Niklas Köhler ◽  
Thomas Blasi ◽  
Andrew Filby ◽  
Anne E. Carpenter ◽  
...  
Keyword(s):  
Cell Cycle ◽  
Deep Learning ◽  
Disease Progression ◽  
Image Data ◽  
Image Features ◽  
Jurkat Cells ◽  
Biological Processes ◽  
Deep Convolutional Neural Networks ◽  
Recent Approach ◽  
Fold Reduction

AbstractWe show that deep convolutional neural networks combined with non-linear dimension reduction enable reconstructing biological processes based on raw image data. We demonstrate this by recon-structing the cell cycle of Jurkat cells and disease progression in diabetic retinopathy. In further analysis of Jurkat cells, we detect and separate a subpopulation of dead cells in an unsupervised manner and, in classifying discrete cell cycle stages, we reach a 6-fold reduction in error rate compared to a recent approach based on boosting on image features. In contrast to previous methods, deep learning based predictions are fast enough for on-the-fly analysis in an imaging flow cytometer.

scholarly journals An Imbalanced Image Classification Method for the Cell Cycle Phase

Information ◽  
2021 ◽  
Vol 12 (6) ◽  
pp. 249
Author(s):  
Xin Jin ◽  
Yuanwen Zou ◽  
Zhongbing Huang
Keyword(s):  
Cell Cycle ◽  
Deep Learning ◽  
Image Classification ◽  
Classification Accuracy ◽  
Data Augmentation ◽  
Cycle Phase ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Cellular Life

The cell cycle is an important process in cellular life. In recent years, some image processing methods have been developed to determine the cell cycle stages of individual cells. However, in most of these methods, cells have to be segmented, and their features need to be extracted. During feature extraction, some important information may be lost, resulting in lower classification accuracy. Thus, we used a deep learning method to retain all cell features. In order to solve the problems surrounding insufficient numbers of original images and the imbalanced distribution of original images, we used the Wasserstein generative adversarial network-gradient penalty (WGAN-GP) for data augmentation. At the same time, a residual network (ResNet) was used for image classification. ResNet is one of the most used deep learning classification networks. The classification accuracy of cell cycle images was achieved more effectively with our method, reaching 83.88%. Compared with an accuracy of 79.40% in previous experiments, our accuracy increased by 4.48%. Another dataset was used to verify the effect of our model and, compared with the accuracy from previous results, our accuracy increased by 12.52%. The results showed that our new cell cycle image classification system based on WGAN-GP and ResNet is useful for the classification of imbalanced images. Moreover, our method could potentially solve the low classification accuracy in biomedical images caused by insufficient numbers of original images and the imbalanced distribution of original images.

scholarly journals A temporal visualization of chronic obstructive pulmonary disease progression using deep learning and unstructured clinical notes

2019 ◽  
Vol 19 (S8) ◽  
Author(s):  
Chunlei Tang ◽  
Joseph M. Plasek ◽  
Haohan Zhang ◽  
Min-Jeoung Kang ◽  
Haokai Sheng ◽  
...  
Keyword(s):  
Chronic Obstructive Pulmonary Disease ◽  
Deep Learning ◽  
Disease Progression ◽  
Pulmonary Disease ◽  
Time Lapse ◽  
Chronic Obstructive ◽  
Allocation Model ◽  
Obstructive Pulmonary Disease ◽  
Clinical Notes ◽  
Copd Patients

Abstract Background Chronic obstructive pulmonary disease (COPD) is a progressive lung disease that is classified into stages based on disease severity. We aimed to characterize the time to progression prior to death in patients with COPD and to generate a temporal visualization that describes signs and symptoms during different stages of COPD progression. Methods We present a two-step approach for visualizing COPD progression at the level of unstructured clinical notes. We included 15,500 COPD patients who both received care within Partners Healthcare’s network and died between 2011 and 2017. We first propose a four-layer deep learning model that utilizes a specially configured recurrent neural network to capture irregular time lapse segments. Using those irregular time lapse segments, we created a temporal visualization (the COPD atlas) to demonstrate COPD progression, which consisted of representative sentences at each time window prior to death based on a fraction of theme words produced by a latent Dirichlet allocation model. We evaluated our approach on an annotated corpus of COPD patients’ unstructured pulmonary, radiology, and cardiology notes. Results Experiments compared to the baselines showed that our proposed approach improved interpretability as well as the accuracy of estimating COPD progression. Conclusions Our experiments demonstrated that the proposed deep-learning approach to handling temporal variation in COPD progression is feasible and can be used to generate a graphical representation of disease progression using information extracted from clinical notes.

B Cell Receptor (BCR) Ligation Enhances Cell Survival in Progressive Chronic Lymphocytic Leukemia (CLL).

Blood ◽  
2004 ◽  
Vol 104 (11) ◽  
pp. 966-966
Author(s):  
Pierre-Antoine Deglesne ◽  
Natalie Chevallier ◽  
Remi Letestu ◽  
Celia Salanoubat ◽  
Laurence Sanhes ◽  
...  
Keyword(s):  
Cell Cycle ◽  
Cell Survival ◽  
Disease Progression ◽  
Mtt Assay ◽  
Specific Antibody ◽  
Progressive Disease ◽  
Lymphocytic Leukemia ◽  
Gene Profiling ◽  
Cell Cycle Protein ◽  
Promoting Effect

Abstract Recent gene profiling studies demonstrated that signalisation through the BCR plays a central role during CLL disease progression. However, incidence of BCR engagement on cell survival remains unclear as both increased and decreased apoptosis have been reported after BCR ligation by anti-mu antibodies. We investigated the effect of anti-mu exposure on fresh cells from 32 untreated CLL cases. Cell viability was assessed both by MTT assay and Annexin V/ IP labelling. Culture conditions were set after time and dose experiments. A 25% increase of viability, obtained by MTT assay after 72h of culture, was considered as significant. When a coated rabbit anti-human IgH mu specific antibody was used, it significantly prevented apoptosis in 20/32 cases (named responders) and never increased it. On the same CLL cases, soluble goat F(ab’)2 anti-human IgH mu specific antibody increased apoptosis in all but one case. However, coating of this latter antibody to the culture plate cancelled the pro-apoptotic effect and restored cell survival promoting effect in 13/14 cases tested. Unspecific effects were ruled out using control antibodies. Cytometry experiments and confocal microscopy demonstrated that uncoated antibody was rapidly internalized and degraded. In responders cases, enhanced survival was supported by increased expression of early cell cycle protein cyclin D2 and cdk4. There was no subsequent cell cycle progression as shown by IP staining and absence of pRB phosphorylation most likely related to the lack of p27 down regulation. There was a strong correlation between the extent of apoptosis prevention after anti-mu exposure and disease progression. Indeed, the 12 over 32 cases who were found to have no IgVH somatic hypermutations, happened to be all responders. Conversely among the 20 mutated cases, 12/20 were non responders and had a clinically stable disease. As for the 8 responders with somatic mutations, four had detectable levels of ZAP 70 protein and 2 others had a clinically progressive disease. In conclusion, coated anti-mu antibody is likely to mimic in vivo conditions of stimulation, suggesting that BCR ligation could play a major role in the pathophysiology of the progressive disease, giving a survival advantage to these CLL cells

A Novel Assay for Development of Chromosomal Lesions in Human CD34+ Cells Following Exposure to Ionizing Radiation Reveals Increased Susceptibility To Acquire Unstable Chromosomal Lesions in CML CD34+ Cells.

Blood ◽  
2007 ◽  
Vol 110 (11) ◽  
pp. 997-997
Author(s):  
Sujata Chakraborty ◽  
Arjun Sehgal ◽  
Stephen Forman ◽  
Smita Bhatia ◽  
Ravi Bhatia
Keyword(s):  
Cell Cycle ◽  
Dna Damage ◽  
Cell Division ◽  
Radiation Exposure ◽  
Disease Progression ◽  
Chromosome Painting ◽  
Chromosomal Abnormalities ◽  
Disease Evolution ◽  
Cell Divisions ◽  
Cd34 Cells

Abstract Acquisition of chromosomal lesions likely plays an important role in pathogenesis of primary and therapy-related leukemia, as well as in disease evolution in myeloproliferative and myelodysplastic disorders. However the process of acquisition of chromosomal lesions in hematopoietic stem cells (HSC) is not well understood. Development of persistent chromosomal abnormalities may involve several factors including acquisition of DNA lesions, repair of DNA damage, sensing of nonrepaired or misrepaired lesions and activation of cell cycle checkpoint and apoptotic pathways, and clonal growth advantage conferred by the lesions. In the current study we assessed the frequency, nature and kinetics of chromosomal lesions following exposure to genotoxic agents in normal human HSC and determined whether these were altered in CML, a prototypic HSC malignancy associated with genetic instability and acquisition of new chromosomal abnormalities during disease progression. CD34+ cells were selected from 4 normal donors and 4 newly diagnosed, untreated chronic phase CML patients. Cells were exposed to increasing doses of γ-radiation, cultured with growth factors and metaphase spreads assessed for development of chromosomal lesions. Chromosome painting was performed using chromosomes 1, 3, 5, 7, 11 and 21 probes, representing 32% of genomic DNA, with >100 metaphases scored per dose per time. Radiation exposure resulted in induction of chromosomal lesions in normal CD34+ cells in a dose-dependent manner. Chromosomal lesions were not seen in cells cultured without radiation exposure. The frequency of aberrant metaphases after 72 h culture (shown by cell cycle analysis to represent the first cell division for normal CD34+ cells) was 2.7% with 0.5 Gy, 4.8% with 1.0 Gy, 9.9% with 2.0 Gy and 23.5% with 4.0 Gy exposures. Chromosome aberrations observed at first division included both stable (translocations, insertions) and unstable (excess fragments, dicentrics) lesions. Subsequent results are shown for exposure to 2.0 Gy. The frequency of aberrant metaphases dropped to 5.3% at 144 h (3–4 cell divisions); only stable lesions persisted at this time. In contrast to normal CD34+ cells, first cell division in CML CD34+ cells was seen at 24 h, at which time 15.3% of the metaphases showed aberrations, whereas 11.2% of metaphases showed aberrations after 72 h (3–4 divisions) and 11.2% after 144 h (6–7 divisions). A striking difference between normal and CML cells was persistence of unstable aberrations after several cell divisions in the latter (42.9% unstable lesions present even after 6–7 divisions). These observations suggest impaired ability to sense and eliminate cells with chromosomal lesions or continued generation of such lesions after initial radiation exposure in CML cells. In conclusion we have developed a novel chromosome painting based assay for evaluation of acquisition of chromosomal lesions in primitive hematopoietic cells. We have demonstrated an inherent chromosomal instability that may contribute to clonal evolution and disease progression in CML CD34+ cells. This assay will provide a useful platform for: i) assessment of mechanisms underlying development of chromosomal lesions in response to DNA damage; and ii) assessing susceptibility to genotoxic agents, and allow improved understanding of pathogenesis and disease evolution in myeloid malignancies.

Chronic Lymphocytic Leukemia B Cells Variably Express Functional Activation-Induced Cytosine Deaminase Protein

Blood ◽  
2010 ◽  
Vol 116 (21) ◽  
pp. 378-378
Author(s):  
Piers EM Patten ◽  
Charles C Chu ◽  
Rajendra N Damle ◽  
Steven L. Allen ◽  
Jonathan E Kolitz ◽  
...  
Keyword(s):  
Cell Cycle ◽  
Cell Division ◽  
B Cells ◽  
Disease Progression ◽  
Peripheral Blood ◽  
Lymphoid Tissues ◽  
Class Switching ◽  
Variable Nature ◽  
Leukemic Clone ◽  
Number Of Cells

Abstract Abstract 378 During the course of CLL, ongoing genetic changes occur within the leukemic clone and such changes associate with disease progression. Activation-induced cytidine deaminase (AID), the enzyme required for IGV gene somatic hypermutation and isotype class switching in B cells, is a candidate enzyme for causing such changes. Depending upon the detection method, circulating CLL cells express mRNA for AID in 40 – 100% of patients, although at any point only a very small percentage of cells within the clone express this message. Because B lymphocytes must be in the cell cycle for AID-induced DNA changes to occur, we hypothesized that AID protein would be contained within recently divided CLL cells and that these cells would exhibit new IGV mutations and/or class switching. Using appropriate markers for recently divided cells, we found that AID mRNA is enriched in/limited to this subset. Furthermore, because dividing cells in CLL are principally found within bone marrow and secondary lymphoid tissues, we analyzed such cells in lymph nodes (LNs) for AID protein expression and activity. In 50% of LNs infiltrated with CLL (n=10), AID protein was detected in large cells expressing a CLL phenotype; these cells were predominantly in the cell cycle. Nevertheless, even in those cases where CLL cells expressed AID, most cycling cells were AID protein negative. FACS analysis of dispersed LN cells confirmed the presence of AID protein-expressing cells and such cells had the phenotype of recently divided cells. To demonstrate that AID protein was functionally competent, we co-cultured peripheral blood CLL cells with anti-CD40 mAb and IL-4 in the presence of irradiated CD32-transfected fibroblasts, a model that mimics the tissue microenvironment. In 16 patients, we showed that peripheral blood leukemic cells could express AID protein, although the degree of upregulation was highly variable between cases. Using the dye CFSE to track CD5+CD19+ cell division, we found that AID protein always occurred when multiply divided cells were present. Some cases showed immediate AID production prior to division, while others exhibited no or little expression until passing through several cell cycles. AID protein causes double strand breaks (DSBs) within DNA, for example in IG switch regions during class switch recombination. We therefore used confocal microscopy to detect the presence of phospho-histone H2A.X (pH2AX), which localizes to DSBs, in CFSE-labeled cells stimulated for 14 days by the conditions mentioned above. At least 10 × 60 magnification images from 3 cultures showing cell division and AID upregulation were obtained, and the fluorescent signals for CFSE and pH2AX quantified from greater than 250 CD23+ CLL cells in each case. A mean of 20.4% of cells (range 10.4 – 38.2%) showed increased fluorescence with anti-pH2AX compared to unstimulated cells. Moreover, stimulated cultures demonstrated increased anti-pH2AX signal in a significantly greater number of cells with diminished CFSE intensity, which are the most divided cells, as compared to less/undivided cells with higher CFSE intensity (p<0.0001 in all cases analyzed, Fisher's exact test). In addition, 20 cell aliquots of unstimulated CLL cells and stimulated CFSE-labeled CLL cells were sorted after culture, yielding pure populations of either undivided cells or cells that had undergone 5 – 6 divisions. While all sorted populations yielded unswitched mu IG transcripts (≥75% wells positive in all groups), switched gamma transcripts with the same V-D-J rearrangement as the leukemic clone were only obtained from divided cells (range 4–9% wells positive), and not present in either undivided or unstimulated cells (0% wells positive). Taken together, the presence of heightened numbers of DSBs in the most divided cells compared to no/minimally divided cells and evidence of IG class switching in the former, indicate that AID protein was functional in these activated CLL peripheral blood cultures. In all, these data demonstrate that in CLL functional AID protein predominates in cells that are dividing or have a recently divided phenotype, although cases vary in the number of cells expressing AID as well as the relative amounts of enzyme expression. Differential AID activity between discrete CLL cases may relate to the development of new DNA mutations leading to clonal evolution and the variable nature of disease progression seen in this disease. Disclosures: No relevant conflicts of interest to declare.

Sign in / Sign up

Export Citation Format

Share Document