Cell image classification based on ensemble features and random forest

Ko, Byoung Chul; Gim, Jawon; Nam, Jae-Yeal

doi:10.1049/el.2011.0831

Cited by 30 publications

(21 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The size of the datasets in some drug discovery applications ranges from ∼ 500 to ∼ 7 × 10 7 . The number of data samples varies from ∼ 250 to ∼ 6 × 10 4 in many digital pathology applications . Some cell biology applications have dataset sizes of the order ∼ 800.…”

Section: Discussionmentioning

confidence: 99%

Cell Image Classification: A Comparative Overview

et al. 2020

View full text Add to dashboard Cite

Cell image classification methods are currently being used in numerous applications in cell biology and medicine. Applications include understanding the effects of genes and drugs in screening experiments, understanding the role and subcellular localization of different proteins, as well as diagnosis and prognosis of cancer from images acquired using cytological and histological techniques. The article also reviews three main approaches for cell image classification most often used: numerical feature extraction, end-to-end classification with neural networks (NNs), and transport-based morphometry (TBM). In addition, we provide comparisons on four different cell imaging datasets to highlight the relative strength of each method. The results computed using four publicly available datasets show that numerical features tend to carry the best discriminative information for most of the classification tasks. Results also show that NN-based methods produce state-of-the-art results in the dataset that contains a relatively large number of training samples. Data augmentation or the choice of a more recently reported architecture does not necessarily improve the classification performance of NNs in the datasets with limited number of training samples. If understanding and visualization are desired aspects, TBM methods can offer the ability to invert classification functions, and thus can aid in the interpretation of results. These and other comparison outcomes are discussed with the aim of clarifying the advantages and disadvantages of each method. © 2020 International Society for Advancement of Cytometry Key terms image informatics; computational biology; cell biology; digital pathology Interpretation of images of cells has always played important roles in science and medicine. From their discovery in 1665, observation of the spatiotemporal characteristics of cells through microscopic technology has enabled us to better understand the structure of living cells, as well as how they perform certain functions (1,2). In addition, scientists have long used microscopes to evaluate the efficacy of different compounds for drug development (3-9). In medicine, as another example, the observation of cell morphology has long been used to discern malignancy in cancer cells (10-13).Cells are known to exhibit complex phenotypes such as differences in shape, gene expression, subcellular protein localization, and other qualities. In addition, cell cultures, tissues, and organs are known to exhibit complex heterogeneity of phenotypes. The combination of intricate phenotype differences together with their heterogeneous responses to different conditions (e.g., diseases) has made decoding biological processes an increasingly complex task. Thus computational approaches for analyzing images of cells have been used increasingly to aid in the task of decoding the complexity of biological processes. A common task useful in many practical situations is determining the category of a given cell or set of cells: a task known as cell classification.Auto...

show abstract

Section: Discussionmentioning

confidence: 99%

Cell Image Classification: A Comparative Overview

et al. 2020

View full text Add to dashboard Cite

show abstract

“…B. Ko et al showed that a Random Forest (RF) classifier was more accurate in classifying white blood cells compared to other machine learning models. The RF model is good at classifying white blood cells with a small amount of training data using ensemble features [ 16 ]. Chuanxin Zou proposed a framework for sequence descriptor-based protein function prediction using a SVM model which exploits the protein properties to assist with feature selection [ 45 ].…”

Section: Related Workmentioning

confidence: 99%

Comparison of Artificial Intelligence based approaches to cell function prediction

Padi

Manescu

Schaub

et al. 2020

Informatics in Medicine Unlocked

View full text Add to dashboard Cite

Predicting Retinal Pigment Epithelium (RPE) cell functions in stem cell implants using non-invasive bright field microscopy imaging is a critical task for clinical deployment of stem cell therapies. Such cell function predictions can be carried out using Artificial Intelligence (AI) based models. In this paper we used Traditional Machine Learning (TML) and Deep Learning (DL) based AI models for cell function prediction tasks. TML models depend on feature engineering and DL models perform feature engineering automatically but have higher modeling complexity. This work aims at exploring the tradeoffs between three approaches using TML and DL based models for RPE cell function prediction from microscopy images and at understanding the accuracy relationship between pixel-, cell feature-, and implant label-level accuracies of models. Among the three compared approaches to cell function prediction, the direct approach to cell function prediction from images is slightly more accurate in comparison to indirect approaches using intermediate segmentation and/or feature engineering steps. We also evaluated accuracy variations with respect to model selections (five TML models and two DL models) and model configurations (with and without transfer learning). Finally, we quantified the relationships between segmentation accuracy and the number of samples used for training a model, segmentation accuracy and cell feature error, and cell feature error and accuracy of implant labels. We concluded that for the RPE cell data set, there is a monotonic relationship between the number of training samples and image segmentation accuracy, and between segmentation accuracy and cell feature error, but there is no such a relationship between segmentation accuracy and accuracy of RPE implant labels.

show abstract

“…It's main purpose is build a proper classification model to analyze the effective information hidden in the students' achievements. As for now, there are many classification methods, such as moment invariants [1], random forest [2,3], support vector machine [4], principal component analysis [5], Markov random fields [6], particle swarm optimization [7], discrete wavelet transform [8], et al Additionally, in [9], a novel framework of complex network classifier is proposed to tackle the problem of network classification, which shows that the proposed method performs well on large-scale networks. In [10], it uses the domain-adversarial learning method to classify the low-resource text, and it can obtain good results.…”

Section: Introductionmentioning

confidence: 99%