Employment of Artificial Intelligence Based on Routine Laboratory Results for the Early Diagnosis of Multiple Myeloma

Yan, Wei; Shi, Hua; He, Tao; Chen, Jian; Wang, Chen; Liao, Aijun; Yang, Wei; Wang, Huihan

doi:10.3389/fonc.2021.608191

Cited by 12 publications

(13 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In their study, MM and non‐myeloma were mixed and randomly divided into training and test subsets. The best model in their study yielded a better performance (PPV = 92.9%, sensitivity = 90%, F‐measure = 91.50%, AUC = 97.50%) compared with ours (PPV = 88.50%, sensitivity = 85.20%, F‐measure = 86.80%, and AUC = 96.80%) 19 . However, the data in the prior study were derived from only a single healthcare institution, and immunoglobulin (A, G, and M) levels were also utilized as input features.…”

Section: Discussionmentioning

confidence: 58%

See 1 more Smart Citation

Routine blood biomarkers for the detection of multiple myeloma using machine learning

Fan

Cui

Zhang

et al. 2022

Int J Lab Hematology

View full text Add to dashboard Cite

Introduction: Primary laboratory tests performed in the diagnosis of multiple myeloma (MM) include bone marrow examination and free light chain assay; however, these may only be ordered after clinical suspicion of disease. In contrast, routine blood test results are readily available. Methods: Machine learning algorithms (ML) combined with routine blood tests were used to detect MM. Feature selection was performed to achieve improved classification performance. The robustness of the classification models was assessed in an internal and external validation data set. To minimize the divergence, the training and validation data sets were combined and used to assess the performance of the ML algorithms. Results:The AdaBoost-DecisionTable produced the best performance (accuracy =94.75%, sensitivity =87.70%, positive predictive value (PPV) =92.50%, F-measure =90.00%, and areas under the receiver operating characteristic curves (AUC) =97.50%) in the training data set using a 10-fold cross-validation. Performance in the validation data sets was affected by the divergence of the data sets, with accuracy greater than 85% and AUC greater than 90% in the validation data sets. The ML algorithm achieved a high accuracy of 92.61%, high AUC (96.80%), a sensitivity value of 85.20%, a PPV value of 88.50%, and an F-measure of 86.80% in a test set that was randomly selected from the combined data set. Conclusions:Combining ML and routine serum biomarkers hold a potential benefit in MM diagnosis.

show abstract

Section: Discussionmentioning

confidence: 58%

“…To our knowledge, to date, there has been only one published study that addresses the diagnosis of MM by combining ML and serum biomarkers 19 . In their study, MM and non‐myeloma were mixed and randomly divided into training and test subsets.…”

Section: Discussionmentioning

confidence: 99%

Routine blood biomarkers for the detection of multiple myeloma using machine learning

Fan

Cui

Zhang

et al. 2022

Int J Lab Hematology

View full text Add to dashboard Cite

show abstract

“…Machine learning is a part of AI and is related to the ability to learn from large amounts of data (a set of lessons) [65]. Recently, Yan et al [66] demonstrated that Cancer cachexia is a disorder characterized by wasting of muscle and adipose tissue, but changes in white adipose tissue (WAT) phenotype, i.e., the conversion of the white adipocytes in "beige" cells, was only described recently [57,58]. Interestingly, both high adipose tissue radiodensity and increased adipose tissue glucose uptake may be related to this phenomenon, and therefore, they may be early markers of cancer cachexia.…”

Section: Artificial Intelligence For Estimating Total Metabolic Tumor Volume In Multiple Myelomamentioning

confidence: 99%

“…Machine learning is a part of AI and is related to the ability to learn from large amounts of data (a set of lessons) [65]. Recently, Yan et al [66] demonstrated that machine learning models derived from routine laboratory results can accurately diagnose MM and can increase the rate of early diagnosis.…”

Section: Artificial Intelligence For Estimating Total Metabolic Tumor Volume In Multiple Myelomamentioning

confidence: 99%

“…CNN-based approaches have been demonstrated to effectively diagnose MM based exclusively on mass spectrometry data from peripheral blood [67]. Deep learning algorithms can use the full set of imaging data directly from the raw images in contrast to conventional machine learning approaches that require manual extraction of these features, a process that is tedious and may not fully capture the underlying imaging information for the task with potential selection bias [66].…”

Section: Artificial Intelligence For Estimating Total Metabolic Tumor Volume In Multiple Myelomamentioning

confidence: 99%

See 1 more Smart Citation

Metabolic Volume Measurements in Multiple Myeloma

et al. 2021

View full text Add to dashboard Cite

Multiple myeloma (MM) accounts for 10–15% of all hematologic malignancies, as well as 20% of deaths related to hematologic malignant tumors, predominantly affecting bone and bone marrow. Positron emission tomography/computed tomography with 18F-fluorodeoxyglucose (FDG-PET/CT) is an important method to assess the tumor burden of these patients. It is often challenging to classify the extent of disease involvement in the PET scans for many of these patients because both focal and diffuse bone lesions may coexist, with varying degrees of FDG uptake. Different metrics involving volumetric parameters and texture features have been proposed to objectively assess these images. Here, we review some metabolic parameters that can be extracted from FDG-PET/CT images of MM patients, including technical aspects and predicting MM outcome impact. Metabolic tumor volume (MTV) and total lesion glycolysis (TLG) are volumetric parameters known to be independent predictors of MM outcome. However, they have not been adopted in clinical practice due to the lack of measuring standards. CT-based segmentation allows automated, and therefore reproducible, calculation of bone metabolic metrics in patients with MM, such as maximum, mean and standard deviation of the standardized uptake values (SUV) for the entire skeleton. Intensity of bone involvement (IBI) is a new parameter that also takes advantage of this approach with promising results. Other indirect parameters obtained from FDG-PET/CT images, such as visceral adipose tissue glucose uptake and subcutaneous adipose tissue radiodensity, may also be useful to evaluate the prognosis of MM patients. Furthermore, the use and quantification of new radiotracers can address different metabolic aspects of MM and may have important prognostic implications.

show abstract

Clinical application of intelligent technologies and integration in medical laboratories

Huang

Ding

et al. 2023

iLABMED

View full text Add to dashboard Cite

With the development of scientific technology, the transition to the intelligent era of digitalization and automation is an irresistible trend for medical laboratories. Medical diagnosis systems have undergone significant changes as a result of intelligent technologies, such as machine learning, artificial intelligence, and the Internet of Things, from the collection, transmission, and detection of test samples to the review of reports and the provision of clinical feedback. In addition to significantly enhancing the efficiency, consistency, and accuracy of medical laboratory testing, these technologies also assist the improvement of individualized healthcare and medical expert systems, as well as the early detection and treatment of diseases. The future development of medical laboratories will focus on integrating big data and diverse intelligent resources, cooperating more closely with clinical departments, and realizing the effective pathway of patient‐centered care. The purpose of this review is to illustrate the current state of intelligent technology integration in medical laboratories and provide a preliminary discussion about the potential future influences of intelligent technology development on the evolution of medical laboratories.

show abstract

Employment of Artificial Intelligence Based on Routine Laboratory Results for the Early Diagnosis of Multiple Myeloma

Cited by 12 publications

References 39 publications

Routine blood biomarkers for the detection of multiple myeloma using machine learning

Routine blood biomarkers for the detection of multiple myeloma using machine learning

Metabolic Volume Measurements in Multiple Myeloma

Clinical application of intelligent technologies and integration in medical laboratories

Contact Info

Product

Resources

About