Machine learning applied to near-infrared spectra for clinical pleural effusion classification

Abstract

Lung cancer patients with malignant pleural effusions (MPE) have a particular poor prognosis. It is crucial to distinguish MPE from benign pleural effusion (BPE). The present study aims to develop a rapid, convenient and economical diagnostic method based on FTIR near-infrared spectroscopy (NIRS) combined with machine learning strategy for clinical pleural effusion classification. NIRS spectra were recorded for 47 MPE samples and 35 BPE samples. The sample data were randomly divided into train set (n = 62) and test set (n = 20). Partial least squares, random forest, support vector machine (SVM), and gradient boosting machine models were trained, and subsequent predictive performance were predicted on the test set. Besides the whole spectra used in modeling, selected features using SVM recursive feature elimination algorithm were also investigated in modeling. Among those models, NIRS combined with SVM showed the best predictive performance (accuracy: 1.0, kappa: 1.0, and AUCROC: 1.0). SVM with the top 50 feature wavenumbers also displayed a high predictive performance (accuracy: 0.95, kappa: 0.89, AUCROC: 0.99). Our study revealed that the combination of NIRS and machine learning is an innovative, rapid, and convenient method for clinical pleural effusion classification, and worth further evaluation.

Urine based near-infrared spectroscopy analysis reveals a noninvasive and convenient diagnosis method for cancers: a pilot study

“…Although NIRS has a wide range of applications in the medical field, such as brain function monitoring ( Grossmann, 2008 ; Holper et al, 2019 ), muscle oxygenation monitoring ( Klusiewicz et al, 2021 ), cardiac function monitoring ( Ortega-Loubon et al, 2019 ), neonatal care ( Tran et al, 2021 ), there are rare reports on liquid biopsy- based NIRS for cancer diagnosis. Our previous studies ( Zhu et al, 2023 ; Chen et al, 2021 ) showed a promising diagnostic value in cancer using plasma or pleural effusion. Again, this pilot study revealed that urine-based had a very good prediction value for cancers.…”

“…To discover the most significant NIRS features in SVM model, SVM-recursive feature elimination (SVM-RFE) algorithm was used to rank the NIRS features. The method was according to our previous studies ( Zhu et al, 2023 ; Chen et al, 2021 ). Moreover, to optimize the SVM model through involving less variables, modeling progressively with different numbers of NIRS features (from Top 1 to Top N) were investigated of their prediction performance.…”

Development and validation a Nomogram for identify malignant pleural effusion

“…Previous studies have developed several models for predicting MPE, and it was pointed out by the authors that these models could improve the diagnosis of MPE. For example, Chen et al (14) revealed that the combination of FTIR nearinfrared spectroscopy (NIRS) and machine learning is an innovative, rapid, and convenient method for clinical PE classi cation. Wang et al (15,16) reported a classi cation model to identify BPE and MPE using a CT volume and its 3D PE mask as inputs based on arti cial intelligence.…”

Digital-Twin-Enabled IoMT System for Surgical Simulation Using rAC-GAN

“…The predictive results demonstrated that the proposed model can not only maintain a high accuracy on LC patients, which reaches the experts' level but also decrease the mistake diagnostic rate of the normal. Chen et al [32] used machine learning to classify different pathological conditions of lung cancer patients. The experiments show that combining medical data and AI methods such as machine learning is an innovative, fast and convenient way to classify the corresponding pathologies.…”