Bigger Data Is Better for Molecular Diagnosis Tests Based on Decision Trees

Floares, Alexandru George; Calin, George A.; Manolache, Florin B.

doi:10.1007/978-3-319-40973-3_29

Cited by 4 publications

(3 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As evaluated by Gunduz et al 32, the random forest model substantially outperformed other techniques on both real life and simulated data regarding the task of robust classification in the high dimension low sample size context. Floares et al 33 further justified that the Random Forest method would derive accurate and robust model from omics data of small sample size. Such characteristic made random forest model more suitable to our study where radiomic pattern would be derived from high dimensional data (a total of 1826 features for each patient) of limited number of sample studies.…”

Section: Discussionmentioning

confidence: 99%

¹⁸F-FDG PET/CT Radiomic Analysis with Machine Learning for Identifying Bone Marrow Involvement in the Patients with Suspected Relapsed Acute Leukemia

Xin

et al. 2019

Theranostics

103

View full text Add to dashboard Cite

18 F-FDG PET / CT is used clinically for the detection of extramedullary lesions in patients with relapsed acute leukemia (AL). However, the visual analysis of 18 F-FDG diffuse bone marrow uptake in detecting bone marrow involvement (BMI) in routine clinical practice is still challenging. This study aims to improve the diagnostic performance of 18 F-FDG PET/CT in detecting BMI for patients with suspected relapsed AL. Methods : Forty-one patients (35 in training group and 6 in independent validation group) with suspected relapsed AL were retrospectively included in this study. All patients underwent both bone marrow biopsy (BMB) and 18 F-FDG PET/CT within one week. The BMB results were used as the gold standard or real “truth” for BMI. The bone marrow 18 F-FDG uptake was visually diagnosed as positive and negative by three nuclear medicine physicians. The skeletal volumes of interest were manually drawn on PET/CT images. A total of 781 PET and 1045 CT radiomic features were automatically extracted to provide a more comprehensive understanding of the embedded pattern. To select the most important and predictive features, an unsupervised consensus clustering method was first performed to analyze the feature correlations and then used to guide a random forest supervised machine learning model for feature importance analysis. Cross-validation and independent validation were conducted to justify the performance of our model. Results : The training group involved 16 BMB positive and 19 BMB negative patients. Based on the visual analysis of 18 F-FDG PET, 3 patients had focal uptake, 8 patients had normal uptake, and 24 patients had diffuse uptake. The sensitivity, specificity, and accuracy of visual analysis for BMI diagnosis were 62.5%, 73.7%, and 68.6%, respectively. With the cross-validation on the training group, the machine learning model correctly predicted 31 patients in BMI. The sensitivity, specificity, and accuracy of the machine learning model in BMI detection were 87.5%, 89.5%, and 88.6%, respectively, significantly higher than the ones in visual analysis ( P < 0.05). The evaluation on the independent validation group showed that the machine learning model could achieve 83.3% accuracy. Conclusions: 18 F-FDG PET/CT radiomic analysis with machine learning model provided a quantitative, objective and efficient mechanism for identifying BMI in the patients with suspected relapsed AL. It is suggested in particular for the diagnosis of BMI in the patients with 18 F-FDG diffuse uptake patterns.

show abstract

Section: Discussionmentioning

confidence: 99%

¹⁸F-FDG PET/CT Radiomic Analysis with Machine Learning for Identifying Bone Marrow Involvement in the Patients with Suspected Relapsed Acute Leukemia

Xin

et al. 2019

Theranostics

103

View full text Add to dashboard Cite

show abstract

“…We used this approach in Refs. [11,12] for TCGA miRNA data and Ensemble of Decision Trees algorithms like Random Forest and XGBoost. We iteratively increase the sample size, run the algorithm, and register the corresponding performance.…”

Section: The Limitations Of Traditional Biostatistical Methods In The...mentioning

confidence: 99%

i-Biomarker - Multi Cancer Early Detection as a Data Science with AI Problem

Floares,

Zety,

Floares

et al. 2023

Molecular Diagnostics of Cancer [Working Title]

View full text Add to dashboard Cite

Cancer is one of the most common and deadly diseases worldwide, claiming millions of lives yearly. Despite significant advances in treatment, the overall survival rate remains low, primarily due to late-stage diagnosis. In the high-throughput, high-dimensional omics data era, Biomedical Knowledge should be combined with Data Science best practices for real progress toward precision and personalized medicine. We intuitively and non-technically formulated the main problems or traps and suggested solutions. To illustrate them, we used our Biomedical Data Science platform, i-Biomarker, and its application to circulating miRNA for personalized Multi-Cancer Early Detection and treatment response monitoring, i-Biomarker CaDx. i-Biomarker combines and automates bioinformatics and Explainable AI/ML pipelines. i-Biomarker CaDx works on 32 types of cancer with 99–100% accuracy and is based on more than 30,000 cases.

show abstract

“…The usage of DNA alongside with decision tree is recognized in several articles that document the use of DNA sequences [18] and plasma DNA [19] for cancer detection. The profiling of gene expression using decision tree to search the gene signature of breast cancer cell is depicted in a number of studies where microRNA [20] and RNA-Seq datasets [21], [22] are used in cancer classifiers. The potential of using protein amino acid features for classification of breast cancer in cancer detection is presented in publications by Ali and Majid [23] and Ali, Majid, Javed and Sattar [24].…”

Section: Diagnosis Of Breast Cancermentioning

confidence: 99%

The use of Decision Tree in Breast Cancer-Related Research: a Scoping Analysis Based on Scopus-Indexed Articles

Badi’auzzaman¹,

Moey²,

Azemin³

et al. 2019

IJITEE

View full text Add to dashboard Cite

Breast cancer is the leading cancer that occurs in women globally. The use of machine learning has been introduced to supplement the work in breast cancer studies. There are undisputed pieces of evidence of the existence of publications pertaining to the use of decision tree in breast cancer-related research. However, little is known regarding the types and frequencies of the searched articles. The main objective of this paper is to unearth the broad variety of articles related to breast cancer research that utilized decision trees. The Scopus database was chosen to examine the trend, frequencies and themes of the related publications from the year 2013 until 2018. The study was also intended to disclose the categories of articles based on the areas of breast cancer that have employed the decision trees method. A total of 259 articles from Scopus database were found to meet the inclusion criteria. The analysis of the frequency of published articles generally shows an upward trend. The majority of articles targeted diagnosis of breast cancer (37.8%) in comparisons with other categories. Even though the number of articles found is adequate, several categories of breast cancer are lacking in publications specifically the survivability, incidence, and recurrence of breast cancer among patients. There is a need to redirect the focus of breast cancer research on these categories for future efforts

show abstract

Bigger Data Is Better for Molecular Diagnosis Tests Based on Decision Trees

Cited by 4 publications

References 8 publications

¹⁸F-FDG PET/CT Radiomic Analysis with Machine Learning for Identifying Bone Marrow Involvement in the Patients with Suspected Relapsed Acute Leukemia

¹⁸F-FDG PET/CT Radiomic Analysis with Machine Learning for Identifying Bone Marrow Involvement in the Patients with Suspected Relapsed Acute Leukemia

i-Biomarker - Multi Cancer Early Detection as a Data Science with AI Problem

The use of Decision Tree in Breast Cancer-Related Research: a Scoping Analysis Based on Scopus-Indexed Articles

Contact Info

Product

Resources

About

Bigger Data Is Better for Molecular Diagnosis Tests Based on Decision Trees

Cited by 4 publications

References 8 publications

18F-FDG PET/CT Radiomic Analysis with Machine Learning for Identifying Bone Marrow Involvement in the Patients with Suspected Relapsed Acute Leukemia

18F-FDG PET/CT Radiomic Analysis with Machine Learning for Identifying Bone Marrow Involvement in the Patients with Suspected Relapsed Acute Leukemia

i-Biomarker - Multi Cancer Early Detection as a Data Science with AI Problem

The use of Decision Tree in Breast Cancer-Related Research: a Scoping Analysis Based on Scopus-Indexed Articles

Contact Info

Product

Resources

About

¹⁸F-FDG PET/CT Radiomic Analysis with Machine Learning for Identifying Bone Marrow Involvement in the Patients with Suspected Relapsed Acute Leukemia

¹⁸F-FDG PET/CT Radiomic Analysis with Machine Learning for Identifying Bone Marrow Involvement in the Patients with Suspected Relapsed Acute Leukemia