Colorectal Cancer Detected by Machine Learning Models Using Conventional Laboratory Test Data

Abstract

Background: Current diagnostic methods for colorectal cancer (CRC) are colonoscopy and sigmoidoscopy, which are invasive and complex procedures with possible complications. This study aimed to determine models for CRC identification that involve minimally invasive, affordable, portable, and accurate screening variables. Methods: This was a retrospective study that used data from electronic medical records of patients with CRC and healthy individuals between July 2017 and June 2018. Laboratory data, including liver enzymes, lipid profiles, complete blood counts, and tumor biomarkers, were extracted from the electronic medical records. Five machine learning models (logistic regression, random forest, k-nearest neighbors, support vector machine [SVM], and naïve Bayes) were used to identify CRC. The performances were evaluated using the areas under the curve (AUCs), sensitivity, specificity, positive predictive values (PPV), and negative predictive values (NPV). Results: A total of 1164 electronic medical records (CRC patients: 582; healthy controls: 582) were included. The logistic regression model achieved the highest performance in identifying CRC (AUC: 0.865, sensitivity: 89.5%, specificity: 83.5%, PPV: 84.4%, NPV: 88.9%). The first four weighted features in the model were carcinoembryonic antigen (CEA), hemoglobin (HGB), lipoprotein (a) (Lp(a)), and high-density lipoprotein (HDL). A diagnostic model for CRC was established based on the four indicators, with an AUC of 0.849 (0.840-0.860) for identifying all CRC patients, and it performed best in discriminating patients with late colon cancer from healthy individuals with an AUC of 0.905 (0.889-0.929). Conclusions: The logistic regression model based on CEA, HGB, Lp(a), and HDL might be a powerful, noninvasive, and cost-effective method to identify CRC.

A Retrospective Study from a Single Center to Identify Hematological Factors that Distinguish Between Patients with Colorectal Carcinoma and Colorectal Adenoma

“…A retrospective study in China stated that MPV-to-platelets ratio was used as an effective diagnostic index to distinguish benign from malignant CRC and distinguish early CRC from advanced CRC [ 19 ]. Hui et al [ 21 ] found that CEA, Hb, lipoprotein (a), and high-density lipoprotein might be powerful and noninvasive diagnostic indicators based on machine-learning approaches. Consistently, we have studied the diagnosis of CRC by hematological indicators, and there is also agreement on diagnostic indicators such as platelet-related parameters and CEA.…”

“…In addition, carcinoembryonic antigen (CEA) is a common tumor marker that has assisted in the diagnosis of CRC [ 20 ]. A machine-learning model using biochemical and hematological markers was built to identify CRC patients [ 21 ]. Several studies have established models of diagnosing CRC based on inflammation factors, platelet-related markers, or CEA [ 22 , 23 ], but most studies have excluded people with inflammatory diseases.…”

Clinlabomics: leveraging clinical laboratory data by data mining strategies

“…However, due to the simple form of the LR model (very similar to the linear model), it is difficult to fit the real distribution of data, so the accuracy is not high. Therefore, the LR model is currently widely used to predict the factors of disease pathogenesis [ 77 , 83 ]. We can use the LR model to analyze and predict disease risk through clinical laboratory testing items.…”

“…Li also used laboratory data, including liver enzymes, lipid profiles, complete blood counts, and tumor biomarkers to develop five machine learning models to identify colorectal cancer (CRC). The results showed that the logistic regression model achieved the highest performance in identifying CRC (AUC: 0.865, sensitivity: 89.5%, specificity: 83.5%, PPV: 84.4%, NPV: 88.9%) [ 83 ]. Studies of Clinlabomics in diagnosing disease have increased, both in general and severe diseases, and have achieved remarkable diagnostic results.…”

Potential roles of serum ATPase and AMPase in predicting diagnosis of colorectal cancer patients

“…Currently, colonoscopy is considered the gold standard tool for CRC detection; however, invasive procedures and complications preclude its widespread implementation [ 3 ]. Compared to colonoscopy, blood-based tumor biomarkers such as carcinoembryonic antigen (CEA) and carbohydrate antigen 19–9 (CA19-9), are widely used because they are noninvasive but show unsatisfactory performance [ 4 ]. Therefore, investigations of novel biomarkers, especially blood-based biomarkers for the early diagnosis of CRC, are highly warranted.…”