scholarly journals Assessment of Ki-67 proliferation index with deep learning in DCIS (ductal carcinoma in situ)

2021 ◽  
Author(s):  
Lukasz Fulawka ◽  
Jakub Błaszczyk ◽  
Martin Tabakov ◽  
Agnieszka Halon
Keyword(s):  
Deep Learning ◽  
Surface Area ◽  
Ductal Carcinoma ◽  
Region Of Interest ◽  
Absolute Error ◽  
Visual Assessment ◽  
Total Surface Area ◽  
Proliferation Index ◽  
Ki 67 ◽  
The Impact

Abstract The proliferation index (PI) is crucial in histopathologic diagnostics, in particular tumors. It is calculated based on Ki-67 protein expression by immunohistochemistry. PI is routinely evaluated by a visual assessment of the sample by a pathologist. However, this approach is far from ideal due to its poor intra- and interobserver variability and time-consuming. These factors force the community to seek out more precise solutions. Virtual pathology as being increasingly popular in diagnostics, armed with artificial intelligence, may potentially address this issue. The proposed solution calculates the Ki-67 proliferation index by utilizing a deep learning model and fuzzy-set interpretations for hot-spots detection. The obtained region-of-interest is then used to segment relevant cells via classical methods of image processing. The index value is approximated by relating the total surface area occupied by immunopositive cells to the total surface area of relevant cells. The achieved results are compared to the manual calculation of the Ki-67 index made by a domain expert. To increase results reliability, we trained several models in a 3-fold manner and compared the impact of different hyper-parameters. Our best-proposed method estimates PI with 0,026 mean absolute error, which gives a significant advantage over the current state-of-the-art solution.

Estimating the Throwing Power of SS316 when Coupled with AA7075 Through Finite Element Modeling

CORROSION ◽  
10.5006/3438 ◽  
2020 ◽  
Vol 76 (5) ◽  
pp. 476-484
Author(s):  
R.S. Marshall ◽  
A. Goff ◽  
C. Sprinkle ◽  
A. Britos ◽  
R.G. Kelly
Keyword(s):  
Surface Area ◽  
Noble Metal ◽  
Galvanic Corrosion ◽  
Sol Gel ◽  
Total Surface Area ◽  
Electrochemical Kinetics ◽  
Throwing Power ◽  
Galvanic Current ◽  
Fastener Hole ◽  
The Impact

Galvanic corrosion is common in applications involving a fastener and panel assembly. Often, the fastener is made from a more noble metal and the panel is made from a less noble metal, selected for their respective mechanical properties. The ability for the more noble material to galvanically couple to the panel’s surface as a function of distance is referenced to as “throwing power,” and was the main subject of this research. In this work, SS316 and AA7075 were investigated as the fastener and panel material, respectively. A Ti-6Al-4V fastener and a sol-gel coated SS316 fastener were also considered to determine the impact of different materials on the galvanically driven throwing power. Along with different fastener materials, different fastener geometries were considered as well. Raised fasteners are generally used in tandem with washers, while countersunk fasteners are not in order to remain flush with the surface. The difference between these two geometries on the throwing power was investigated. It was determined that the SS316 washer was the largest contributor to the galvanic current in the raised fastener assembly, due to its large surface area. At distances of two inches away, the SS316 fastener and washer were able to double the natural corrosion rate of AA7075. A countersunk SS316 fastener, with the same total surface area as that of the raised fastener and washer assembly, was seen to lower the throwing power which forced a large amount of current down the fastener hole. Throughout all of the computational tests, the model relies on the generation of accurate electrochemical kinetics measured in solutions of appropriate composition.

Predictors of invasive breast cancer or DCIS recurrence in estrogen receptor positive (ER+) and estrogen receptor negative (ER-) ductal carcinoma in situ (DCIS) patients with and without associated invasive carcinoma (IC)

2009 ◽  
Vol 27 (15_suppl) ◽  
pp. e11523-e11523 ◽  
Author(s):  
J. Picarsic ◽  
A. Brufsky ◽  
A. Onisko ◽  
M. Chivukula
Keyword(s):  
Estrogen Receptor ◽  
Biological Markers ◽  
Hormone Receptors ◽  
Invasive Carcinoma ◽  
Ductal Carcinoma ◽  
Response To Therapy ◽  
Proliferation Index ◽  
Low Grade ◽  
Ki 67 ◽  
Mib 1

e11523 Background: DCIS is a heterogeneous pre-invasive carcinoma with a spectrum of clinical behavior. Patients with ER+ IC have better outcomes compared to ER- patients. FOXA1 and GATA 3 family of transcription factors have been shown to be associated with hormone receptors (ER and PR) and other variables of good prognosis with better overall and relapse-free survival rate. The specific aim of this study is to analyze the expression of these novel biological markers: FOXA1, GATA-3, with recognized markers: MIB-1(Ki-67) and HER2 /neu in DCIS patients with/without associated IC. Methods: Sixty-nine (69) cases of DCIS [(fifty two (52) cases in ER+; seventeen (17) in ER-] were retrieved from our Pathology database. The expressions of the biological markers are analyzed by using a panel of immunohistochemical stains. FOXA1, GATA 3, ER, PR are nuclear stains, a cumulative “H score” is derived based on proportionality (PS) and intensity scores (IS). A proliferation index (PI) is calculated for MIB-1 (Ki67) nuclear stain (low <10%, moderate 11–25%, high 26–50%, very high>50%). Her2/neu is scored as per guidelines for HercepTest (0, or 1+ =negative, 2+ =weakly positive, 3+ =strongly positive). Results: DCIS is categorized into low grade (LG) (nuclear grade 1 and 2), high grade (HG) (grade 3). In the HGDCIS (n=48), four (4) cases had IC after a mean of 7.75 years; three cases of recurrent DCIS after a mean 6 years. No recurrent IC or DCIS is seen in the LGDCIS (n=21) group. The results are shown in the Table . Conclusions: (1) Decreased expression of GATA 3 is observed in HGDCIS ER- group may be a contributor to higher recurrence observed in this group (14%) versus (0%) in ER+ group. (2) A strong expression of FOXA, GATA3, low Ki-67 index, absent Her 2 expression are characteristically seen in our ER+ DCIS group, as previously described in IC. 3. Comparing the response to therapy and outcome in the ER+ and ER- groups is on going. [Table: see text] No significant financial relationships to disclose.

Effect of Gestational and Lactational Exposure to DEHP, DINP, and DEP on Intestinal Morphology, Disaccharidases, and Alkaline Phosphatase in Rats during Postnatal Development

2018 ◽  
Vol 35 (13) ◽  
pp. 1251-1259 ◽  
Author(s):  
O. Kharoubi ◽  
A. Aoues ◽  
M. Bouchekara ◽  
B. Khaladi ◽  
M. Taleb ◽  
...  
Keyword(s):  
Corn Oil ◽  
Villous Atrophy ◽  
Wistar Rat ◽  
Treated Group ◽  
The Body ◽  
Intestinal Morphology ◽  
Proliferation Index ◽  
Postnatal Exposure ◽  
Ki 67 ◽  
The Impact

Introduction The diesters of 1,2-benzenedicarboxylic acid (phthalic acid), commonly known as phthalates, are used primarily as plasticizers of polyvinyl chloride and as additives in consumer and personal care products. Objective This study was designed to evaluate the impact of in utero and postnatal exposure to diisononyl phthalate (DINP), diethylhexyl phthalate (DEHP), and diethyl phthalate (DEP) on gut maturation in a Wistar rat model. Materials and Methods Pregnant females were gavaged from day 8 of gestation through postnatal day (pd) 30 with 0 (vehicle control), DEHP (380 mg/kg/d), DINP (380 mg/kg/d), or DEP (800 mg/kg/d) dissolved in corn oil. Intestinal samples have been collected at 0, 7, 14, 21, and 30 pd for histological and biochemical analysis. The mitotic index has been evaluated based on the expression of Ki-67 antigen. Results All tested phthalate treatments have significantly decreased the body as well as the organ's weight (p < 0.001). DINP exposure resulted in severe villous atrophy, while DEHP treated group was characterized by lymphoepithelial lesions. In addition, a significant decrease of the Ki-67 proliferation index was observed in the youngest rats (0 and 7 days) upon the various treatments (p < 0.0001), whereas at day 30, an increased numbers of Ki-67 positive cells were observed in DEHP and DEP but bot DINP group. Lactase and sucrase activities were inhibited by DEP in contrast to DINP and DEHP which increased enzymes activity (p < 0.05). Conclusion Our results suggest that exposure to phthalates during gestational and lactational phases negatively impacts the development of the small intestine.

scholarly journals Explainable information retrieval using deep learning for medical images

2021 ◽  
pp. 49-49
Author(s):  
Apoorva Singh ◽  
Husanbir Pannu ◽  
Avleen Malhi
Keyword(s):  
Deep Learning ◽  
Image Classification ◽  
Real Life ◽  
Region Of Interest ◽  
Image Features ◽  
Machine Learning Algorithms ◽  
Support Vector ◽  
Bleeding Detection ◽  
Real Life Situation ◽  
The Impact

Image segmentation is useful to extract valuable information for an efficient analysis on the region of interest. Mostly, the number of images generated from a real life situation such as streaming video, is large and not ideal for traditional segmentation with machine learning algorithms. This is due to the following factors (a) numerous image features (b) complex distribution of shapes, colors and textures (c) imbalance data ratio of underlying classes (d) movements of the camera, objects and (e) variations in luminance for site capture. So, we have proposed an efficient deep learning model for image classification and the proof-of-concept has been the case studied on gastrointestinal images for bleeding detection. The Ex plainable Artificial Intelligence (XAI) module has been utilized to reverse engineer the test results for the impact of features on a given test dataset. The architecture is generally applicable in other areas of image classification. The proposed method has been compared with state-of-the-art including Logistic Regression, Support Vector Machine, Artificial Neural Network and Random Forest. It has reported F1 score of 0.76 on the real world streaming dataset which is comparatively better than traditional methods.

Deep learning-based extended field of view computed tomography image reconstruction: influence of network design on image estimation outside the scan field of view

2022 ◽  
Author(s):  
Bhupinder Singh Khural ◽  
Matthias Baer-Beck ◽  
Eric Fournie ◽  
Karl Stierstorfer ◽  
Yixing Huang ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Deep Learning ◽  
Absolute Error ◽  
Field Of View ◽  
Reconstruction Algorithms ◽  
Ct Scanner ◽  
Real Patient ◽  
Data Set ◽  
Order Reconstruction ◽  
The Impact

Abstract The problem of data truncation in Computed Tomography (CT) is caused by the missing data when the patient exceeds the Scan Field of View (SFOV) of a CT scanner. The reconstruction of a truncated scan produces severe truncation artifacts both inside and outside the SFOV. We have employed a deep learning-based approach to extend the field of view and suppress truncation artifacts. Thereby, our aim is to generate a good estimate of the real patient data and not to provide a perfect and diagnostic image even in regions beyond the SFOV of the CT scanner. This estimate could then be used as an input to higher order reconstruction algorithms [1]. To evaluate the influence of the network structure and layout on the results, three convolutional neural networks (CNNs), in particular a general CNN called ConvNet, an autoencoder, and the U-Net architecture have been investigated in this paper. Additionally, the impact of L1, L2, structural dissimilarity and perceptual loss functions on the neural network’s learning have been assessed and evaluated. The evaluation of data set comprising 12 truncated test patients demonstrated that the U-Net in combination with the structural dissimilarity loss showed the best performance in terms of image restoration in regions beyond the SFOV of the CT scanner. Moreover, this network produced the best mean absolute error, L1, L2, and structural dissimilarity evaluation measures on the test set compared to other applied networks. Therefore, it is possible to achieve truncation artifact removal using deep learning techniques.

scholarly journals Automated Assessment of Ki‐67 Proliferation Index in Neuroendocrine Tumors by Deep Learning

Apmis ◽  
2021 ◽  
Author(s):  
Tiina Vesterinen ◽  
Jenni Säilä ◽  
Sami Blom ◽  
Mirkka Pennanen ◽  
Helena Leijon ◽  
...  
Keyword(s):  
Deep Learning ◽  
Neuroendocrine Tumors ◽  
Proliferation Index ◽  
Automated Assessment ◽  
Ki 67

scholarly journals Identifying Leaf Phenology of Deciduous Broadleaf Forests from PhenoCam Images Using a Convolutional Neural Network Regression Method

2021 ◽  
Vol 13 (12) ◽  
pp. 2331
Author(s):  
Mengying Cao ◽  
Ying Sun ◽  
Xin Jiang ◽  
Ziming Li ◽  
Qinchuan Xin
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Environmental Changes ◽  
Region Of Interest ◽  
Absolute Error ◽  
Leaf Phenology ◽  
Coefficient Of Determination ◽  
Learning Approaches ◽  
The One

Vegetation phenology plays a key role in influencing ecosystem processes and biosphere-atmosphere feedbacks. Digital cameras such as PhenoCam that monitor vegetation canopies in near real-time provide continuous images that record phenological and environmental changes. There is a need to develop methods for automated and effective detection of vegetation dynamics from PhenoCam images. Here we developed a method to predict leaf phenology of deciduous broadleaf forests from individual PhenoCam images using deep learning approaches. We tested four convolutional neural network regression (CNNR) networks on their ability to predict vegetation growing dates based on PhenoCam images at 56 sites in North America. In the one-site experiment, the predicted phenology dated to after the leaf-out events agree well with the observed data, with a coefficient of determination (R2) of nearly 0.999, a root mean square error (RMSE) of up to 3.7 days, and a mean absolute error (MAE) of up to 2.1 days. The method developed achieved lower accuracies in the all-site experiment than in the one-site experiment, and the achieved R2 was 0.843, RMSE was 25.2 days, and MAE was 9.3 days in the all-site experiment. The model accuracy increased when the deep networks used the region of interest images rather than the entire images as inputs. Compared to the existing methods that rely on time series of PhenoCam images for studying leaf phenology, we found that the deep learning method is a feasible solution to identify leaf phenology of deciduous broadleaf forests from individual PhenoCam images.

Sign in / Sign up

Export Citation Format

Share Document