Turbidivision: a machine vision application for estimating turbidity from underwater images.

The measurement of turbidity serves as a key indicator of water quality and purity, crucial for informing decisions related to industrial, ecological, and public health applications. As existing processes require both additional expenses and steps to be taken during data collection relative to photography, we seek to generate accurate estimations of turbidity from underwater images. Such a process could give new insight to historical image datasets and provide an alternative to measuring turbidity when lower accuracy is acceptable, such as in citizen science and education applications. We used a two-step approach to a machine vision model, creating an image classification model trained on image data and their corresponding turbidity values recorded from a turbidimeter that is then used to generate continuous values through multiple linear regression. To create a robust model, we collected data for model training from a combination of in situ field sites and lab mesocosms across suspended sediment and colorimetric profiles, with and without a Secchi disk for visual standard, and binned images into 11 classes 0-55 Formazin Nephelometric Units (FNU). Our resulting classification model is highly accurate with 100% of predictions within one class of the expected class, and 84% of predictions matching the expected class. Regression results provide a continuous value that is accurate to ±0.7 FNU of true values below 2.5 FNU and ±33% between 2.5 and 55 FNU; values that are less accurate than conventional turbidimeters but comparable to field-based test kits frequently used in classroom and citizen science applications. To make the model widely accessible, we have implemented it as a free and open-source user-friendly web, computer, and Google Play application that enables anyone with a modern device to make use of the tool, the model, or our repository of training images for data collection or future model development.

External Validation of a Limited Sampling Strategy for the Estimation of Mycophenolic Acid Exposure Between Different Assay Methods: PETINIA and HPLC Methods.

INTRODUCTION: A limited sampling strategy (LSS) for estimating the area under the plasma concentration-time curve (AUC0-12) of the immunosuppressant mycophenolic acid (MPA) is used for therapeutic drug monitoring (TDM) in clinical practice. Our study delves into the applicability of the MPA AUC0-12 LSS, originally developed using particle-enhanced turbidimetric inhibition immunoassay (PETINIA) measurements, to those obtained via high-performance liquid chromatography with ultraviolet detection (HPLC-UV). METHODS: We developed an LSS for estimating MPA AUC0-12 based on PETINIA measurements in 32 adult kidney transplant patients who were receiving mycophenolate mofetil. Validation of this strategy was conducted in an additional 14 adult kidney transplant patients (validation sets) through measurements obtained by both PETINIA and HPLC-UV. Predictive performance was assessed using mean absolute error (MAE), root mean squared error (RMSE), and "good guess" defined as predicted AUC within observed AUC ± 15%. RESULTS: The three time point equation (0, 2, and 6 h) emerged as optimal for estimating MPA AUC0-12, balancing predictive performance and usefulness in clinical settings. In validation sets, the coefficient of determination for observed versus predicted AUC0-12 was consistent between PETINIA (0.978) and HPLC-UV (0.958) measurements. Comparable MAE, RMSE, and "good guess" outcomes were observed for PETINIA (6.4%, 8.1%, and 85.7%, respectively) and HPLC-UV (7.6%, 9.4%, and 85.7%, respectively) measurements. CONCLUSION: Our findings support the application of the MPA AUC0-12 LSS, originally developed using PETINIA measurements, to those obtained via HPLC-UV.

Influence of Hemoglobin Strasbourg, a Rare High Oxygen Affinity Hemoglobin Variant, on Different Methods of HbA1c Measurement.

Hemoglobin Strasbourg is a rare high oxygen affinity hemoglobin variant which leads to secondary erythrocytosis. This variant is caused by a HBB gene mutation c.71T > A resulting in an amino acid exchange on position 23 of the ß globin chain (p.Val23Asp.). The influence of Hb Strasbourg on HbA1c measurement has not been studied to date. For patients with hemoglobin variants it is important to know whether possible interferences exist with the measurement of HbA1c. We therefore investigated the influence of Hb Strasbourg on HbA1c measurement with two different HPLC (high-performance liquid chromatography) systems and one turbidimetric immunoassay in two non-diabetic brothers who are heterozygous carriers of Hb Strasbourg. The examined tests are all used in routine diagnostics. In the case of Hb Strasbourg, the HbA1c measured by HPLC showed lower results than those obtained by the immunoassay. We conclude that HbA1c is underestimated when measured with these methods as glycated Hb Strasbourg is most likely not co-eluting with HbA1c in HPLC.

The different serum albumin assays influence the prescriptions in children with primary nephrotic syndrome.

The differences between the serum albumin determined by bromocresol green (BCG) and immunonephelometry (IN) were inconsistent in past studies, and the samples were all adults. We sought to determine the differences in children and reveal the impacts of these differences on the clinical diagnosis and treatments of primary nephrotic syndrome (PNS). Repeated measurements from 576 PNS children showed that albumin measured by BCG and IN (ALB-B and ALB-I) were 19.95 (11.15) g/L and 15.30 (11.05) g/L, respectively, and the mean difference was 4.68 g/L (P < 0.001). The cut-offs we calculated for hypoalbuminemia and severe hypoalbuminemia based on the IN were 25 and 15 g/L, which were 5 g/L lower than the cut-offs recommended by KIDGO, respectively. A pair of historical control samples (206 vs. 216) with ALB-B or ALB-I showed that the proportion of severe hypoalbuminemia was 14.60% greater in IN group (75.20% vs. 60.60%, P < 0.001). The misdiagnosis rate of severe hypoalbuminemia by IN was 33.77% when 20 g/L rather than 15 g/L was used as the cut-off. Furthermore, the proportion of patients receiving albumin injections increased by 10.20%, and the average consumption increased by 97.06% (P = 0.01) along with the use of IN. So, our results suggested that the difference between ALB-B and ALB-I led to misdiagnosis and prescription abuse in PNS children.

Comparative Analysis of Protein Quantification by the SomaScan Assay versus Orthogonal Methods in Urine from People with Diabetic Kidney Disease.

To our knowledge, calibration curves or other validations for thousands of SomaScan aptamers are not publicly available. Moreover, the abundance of urine proteins obtained from these assays is not routinely validated with orthogonal methods (OMs). We report an in-depth comparison of SomaScan readout for 23 proteins in urine samples from patients with diabetic kidney disease (n = 118) vs OMs, including liquid chromatography-targeted mass spectrometry (LC-MS), ELISA, and nephelometry. Pearson correlation between urine abundance of the 23 proteins from SomaScan 3.2 vs OMs ranged from -0.58 to 0.86, with a median (interquartile ratio, [IQR]) of 0.49 (0.18, 0.53). In multivariable linear regression, the SomaScan readout for 6 of the 23 examined proteins (26%) was most strongly associated with the OM-derived abundance of the same (target) protein. For 3 of 23 (13%), the SomaScan and OM-derived abundance of each protein were significantly associated, but the SomaScan readout was more strongly associated with OM-derived abundance of one or more "off-target" proteins. For the remaining 14 proteins (61%), the SomaScan readouts were not significantly associated with the OM-derived abundance of the targeted proteins. In 6 of the latest group, the SomaScan readout was not associated with urine abundance of any of the 23 quantified proteins. To sum, over half of the SomaScan results could not be confirmed by independent orthogonal methods.

Nephelometry vs. Immunoturbidimetry assay: Analytical performance on IgG subclasses.

Interest in measuring immunoglobulin G Subclasses (IgG Subclasses) is increasing as more information is gathered and understanding regarding conditions associated with deficiencies of each IgG Subclass grows. Different methodologies are available for the measurement of IgG Subclasses, but their specificities vary. As a result, laboratories choose the methodology that better suits their routine, but which may not necessarily align with the needs of their population. In addition, the lack of standardization for the quantification of IgG Subclasses causes diagnostic gaps when comparing results provided by different methodologies. Thus, the purpose of our research is to compare the analytical performance of The Binding Site's (TBS) Optilite® human Immunoglobulin G (IgG) and IgG Subclasses Immunoturbidimetry assay, with the Nephelometry method routinely used in our clinical laboratory, Siemens BNII®. Our results show that the Immunoturbidimetry assay appears to be the most reliable to evaluate IgG Subclasses: the sum of IgG Subclasses and Total IgG correlate better than by Nephelometry. Although these methodologies share a similar principle, the comparison of results appears to be compromised. Therefore, prior to switching methodologies, further studies should be conducted to assess which methodology could be better applied to specific populations. It is also essential to standardise IgG Subclasses assays to reduce discrepancies that arise from comparing results.

The 'Erlenmeter': a low-cost, open-source turbidimeter for no-sampling phenotyping of microorganism growth.

This work presents a low-cost, open-source turbidimeter, the 'Erlenmeter', designed to monitor the growth of microorganisms in batch cultures. It is easy to build, based exclusively on inexpensive off-the-shelf electronic components and 3D-printed parts. The Erlenmeter allows measuring the optical density of cultures on standard Erlenmeyer flasks without the need to open the flasks to collect aliquots, ensuring speed, minimal use of consumables, and elimination of the risk of contamination. These features make it particularly well-suited not just for routine research assays but also for experimental teaching. Here we illustrate the use of the Erlenmeter turbidimeter to record the growth of the microalga Phaeodactylum tricornutum, of the bacterium Escherichia coli, and of the yeast Saccharomyces cerevisiae, model organisms that are widely used in research and teaching. The Erlenmeter allows a detailed characterization of the growth curves of all organisms, confirming its usefulness for studying microbial populations dynamics both for research purposes and in classroom settings.

Real-time triplex loop-mediated isothermal amplification (LAMP) using a turbidimeter for detection of shrimp infectious hypodermal and hematopoietic necrosis virus (IHHNV).

OBJECTIVE: The World Organization for Animal Health still regulates the infectious hypodermal and hematopoietic necrosis virus (IHHNV) in shrimp. The existing disease identification approach is time consuming, necessitates expensive equipment, and requires specialized expertise, thereby limiting the accessibility of shrimp disease screening on farms. Loop-mediated isothermal amplification (LAMP) is recognized for its ability to detect inhibitory substances with high sensitivity and specificity. METHODS: We developed a real-time triplex LAMP assay that combines the simplicity of point-of-care testing with the accuracy of a turbidimeter. Using a set of three LAMP primers, our technology enables rapid DNA amplification in a single reaction within 45 min and with a low detection limit (10 copies/reaction). RESULT: We tested 192 shrimp samples from different sources and demonstrated the clinical utility of our method, achieving 100% specificity (95% confidence interval = 93.40-100.00%), 100% sensitivity (97.36-100.00%), and 100% accuracy (98.10-100.00%) in detecting IHHNV DNA, with a high Cohen's kappa value (1) compared to the standard quantitative polymerase chain reaction assay. CONCLUSION: The high technology readiness level of our method makes it a versatile platform for any real-time LAMP assay, and its low cost and simplicity make it well suited for fast deployment and use in shrimp farming.

Depolarization diagrams for circularly polarized light scattering for biological particle monitoring.

Significance: The depolarization of circularly polarized light (CPL) caused by scattering in turbid media reveals structural information about the dispersed particles, such as their size, density, and distribution, which is useful for investigating the state of biological tissue. However, the correlation between depolarization strength and tissue parameters is unclear. Aim: We aimed to examine the generalized correlations of depolarization strength with the particle size and wavelength, yielding depolarization diagrams. Approach: The correlation between depolarization intensity and size parameter was examined for single and multiple scattering using the Monte Carlo simulation method. Expanding the wavelength width allows us to obtain depolarization distribution diagrams as functions of wavelength and particle diameter for reflection and transparent geometries. Results: CPL suffers intensive depolarization in a single scattering against particles of various specific sizes for its wavelength, which becomes more noticeable in the multiple scattering regime. Conclusions: The depolarization diagrams with particle size and wavelength as independent variables were obtained, which are particularly helpful for investigating the feasibility of various particle-monitoring methods. Based on the obtained diagrams, several applications have been proposed, including blood cell monitoring, early embryogenesis, and antigen-antibody interactions.

Turbidimetric bioassays: A solution to antimicrobial activity detection in asymptomatic bacteriuria isolates against uropathogenic Escherichia coli.

Traditional bacteriocin screening methods often face limitations due to diffusion-related challenges in agar matrices, which can prevent the peptides from reaching their target organism. Turbidimetric techniques offer a solution to these issues, eliminating diffusion-related problems and providing an initial quantification of bacteriocin efficacy in producer organisms. This study involved screening the cell-free supernatant (CFS) from eight uncharacterized asymptomatic bacteriuria (ABU) isolates and Escherichia coli 83972 for antimicrobial activity against clinical uropathogenic E. coli (UPEC) strains using turbidimetric growth methods. ABU isolates exhibiting activity against five or more UPEC strains were further characterized (PUTS 37, PUTS 58, PUTS 59, S-07-4, and SK-106-1). The inhibition of the CFS by proteinase K suggested that the antimicrobial activity was proteinaceous in nature, potentially bacteriocins. The activity of E. coli PUTS 58 and SK-106-1 was enhanced in an artificial urine medium, with both inhibiting all eight UPECs. A putative microcin H47 operon was identified in E. coli SK-106-1, along with a previously identified microcin V and colicin E7 in E. coli PUTS 37 and PUTS 58, respectively. These findings indicate that ABU bacteriocin-producers could serve as viable prophylactics and therapeutics in the face of increasing antibiotic resistance among uropathogens.

Clinical Performance of Immunonephelometric Assay and Chemiluminescent Immunoassay for Detection of IgG Subclasses in Chinese.

BACKGROUND: Detection of IgG subclasses (IgGSc) is vital for the diagnosis and management of disease, especially IgG4-related diseases (IgG4-RD). This study aimed to evaluate the performances of the chemiluminescent immunoassay (CLIA) for detecting IgGSc and diagnosing IgG4-RD by IgGSc. METHODS: A total of 40 individuals with IgG4-RD, 40 with primary Sjogren's syndrome (pSS), and 40 healthy controls (HCs) were enrolled. Serum samples were collected for the simultaneous detection of IgG1, IgG2, IgG3, and IgG4 by the Siemens immunonephelometric assay and the CLIA. The correlation analysis was performed, and diagnostic value was analyzed by the receiver operating characteristic (ROC) curve. RESULTS: Patients with IgG4-RD had higher IgG4 (p < 0.001) and lower IgG1 (p < 0.001) than those with pSS, and HC. The results by the Siemens immunonephelometric assay and the CLIA showed a strong correlation in detecting IgG1, IgG2, IgG3, and IgG4 (r = 0.937, r = 0.847, r = 0.871, r = 0.990, all p < 0.001, respectively). The sum of IgG1, IgG2, IgG3, and IgG4 using two assays strongly correlated with total IgG by the IMMAGE 800 (r = 0.866, r = 0.811, both p < 0.001, respectively). For discriminating IgG4-RD from pSS and HC, no significant differences were observed in CLIA IgG4 and Siemens immunonephelometric assay IgG4 (z = 0.138, p = 0.891), which provided the area under the curves (AUCs) of 0.951 (p < 0.001) and 0.950 (p < 0.001), respectively. The AUCs of CLIA IgG1 and Siemens immunonephelometric assay IgG1 in distinguishing pSS from IgG4-RD and HC were 0.761 (p < 0.001) and 0.765 (p < 0.001), respectively, with no significant differences (z = 0.228, p = 0.820). CONCLUSIONS: The CLIA and the Siemens immunonephelometric assay appeared to have good consistency with comparable diagnostic value in detecting IgGSc, especially IgG4, and IgG1 that can accurately identify IgG4-RD or pSS in clinical practice.

High-throughput nephelometry methodology for qualitative determination of aqueous solubility of chemical libraries.

The purpose of the protocol reported in this work is the solubility profiling of large chemical libraries using nephelometry. This technique allows the qualitative classification of compounds as highly, moderately, or poorly water-soluble. The described methodology is not intended to yield quantitative solubility values of the studied compounds but can be used as a primary solubility assessment of large chemical libraries, to guide hit prioritization after High Throughput Screening (HTS) campaigns.

Novel automated antifungal susceptibility testing system for yeasts based on dual-detection algorithm of turbidimetry and colorimetry.

Introduction. The increasing prevalence and growing resistance of fungi present a significant peril to public health. There are only four classes of antifungal medicines available today, and few candidates are in clinical trials.Hypothesis/Gap Statement. Rapid and sensitive diagnostic techniques are lacking for most fungal pathogens, and those that do exist are expensive or hard to obtain.Aim. This study aimed to evaluate the feasibility of a novel automated antifungal susceptibility testing system, Fungus AST, in comparison to the broth microdilution method (BMD) recommended by the Clinical and Laboratory Standards Institute (CLSI).Methodology. A total of 101 clinical Candida spp. isolates were collected from the Zengcheng Branch of Nanfang Hospital and subjected to antifungal susceptibility testing. Antifungal susceptibility was assessed using the Fungus AST method and the BMD.Results. In this study, we introduce a novel automated antifungal susceptibility testing system, Fungus AST, which detects the turbidity and/or colour intensity of microdilution wells using a four-wavelength detection technology in real time and is designed to match the growth characteristics of strains over time. Based on our analysis, all reportable ranges of Fungus AST were suitable for clinical fungal isolates in PR China. Within ±twofold dilutions, reproducibility was 100 %. Considering the BMD as a referenced method, ten antifungal agents (anidulafungin, caspofungin, micafungin, fluconazole, voriconazole, posaconazole, itraconazole, amphotericin B, 5-flucytosine and nystatin) showed an essential agreement of >95 %. The category agreement of five antifungal agents (anidulafungin, caspofungin, micafungin, fluconazole and voriconazole) was excellent at >90 %. One Candida albicans isolate and voriconazole showed a major error (ME) (1.7 %), and no other ME or very ME agents were found.Conclusion. Given the above, it can be argued that the utilization of Fungus AST is a discretionary automated approach. More improvements are needed in Fungus AST compared to the BMD system for a wider range of clinical isolates, including different types of fungi.

Laser Sources for Traditional and Spectral Flow Cytometry.

Lasers for light scattering measurement and fluorescence excitation are essential components of all flow cytometers. Flow cytometers now typically rely on multiple laser wavelengths allowing excitation of a constantly increasing variety of fluorescent probes. The expanding use of spectral flow cytometry to increase the magnitude of multiparametric analysis is also changing the significance of laser choice in cytometry. In this chapter, we review the lasers available for flow cytometry and provide guidance in choosing laser wavelengths and characteristics to best match the needs of modern cell analysis by both conventional and spectral cytometry. We also discuss the recent advances in laser technology as the push to expand the palette of laser wavelength for cytometry continues.

MiniRead: A simple and inexpensive do-it-yourself device for multiple analyses of micro-organism growth kinetics.

Fitness in micro-organisms can be proxied by growth parameters on different media and/or temperatures. This is achieved by measuring optical density at 600 nm using a spectrophotometer, which measures the effect of absorbance and side scattering due to turbidity of cells suspensions. However, when growth kinetics must be monitored in many 96-well plates at the same time, buying several 96-channel spectrophotometers is often beyond budgets. The MiniRead device presented here is a simple and inexpensive do-it-yourself 96-well temperature-controlled turbidimeter designed to measure the interception of white light via absorption or side scattering through liquid culture medium. Turbidity is automatically recorded in each well at regular time intervals for up to several days or weeks. Output tabulated text files are recorded into a micro-SD memory card to be easily transferred to a computer. We propose also an R package which allows (1) to compute the nonlinear calibration curves required to convert raw readings into cell concentration values, and (2) to analyze growth kinetics output files to automatically estimate proxies of growth parameters such as lag time, maximum growth rate, or cell concentration at the plateau.

In Situ Oral Metabolism Analysis of Astringent Compounds in Tea by Paper Spray Mass Spectrometry, Electrospray Mass Spectrometry, Turbidimetry, and Sensory Evaluation.

The phenolic compounds (PCs) are the primary components responsible for the astringency of tea infusions, and this astringency is intricately linked to the in situ oral metabolism of PCs in saliva. Initially, a total of 54 PCs were identified in tea infusions by electrospray mass spectrometry (ESI-MS). Subsequently, an in vivo metabolism analysis of PCs during varying drinking times and oral locations was conducted by both paper spray mass spectrometry (PS-MS) and sensory evaluation. The metabolism of PCs within oral saliva was a prolonged process, the residual PCs were distributed across diverse oral regions after drinking tea infusion, and the higher residual PC content reflected the stronger astringency intensity. Furthermore, an in vitro metabolism analysis of PCs under varied reaction temperatures and durations was performed by ESI-MS and turbidimetry. As the reaction time extended, more PCs in tea was interacting with saliva. Moreover, the higher temperatures facilitated this interaction between PCs and saliva. Therefore, this investigation establishes a foundation for further elucidating the mechanisms underlying astringency formation.

In vitro bacterial penetration and dissemination through dentinal tubules in roots filled with calcium silicate-based cements.

The aim of this study was to assess whether calcium silicate root fillings prevent bacterial penetration and to determine how bacteria penetrate roots. Extracted single-rooted, single-canal human teeth were decoronated, prepared and filled with ProRootMTA or Biodentine (n = 12 each). Positive and negative (n = 2 each) controls were not filled. A two-chamber model was used with Streptococcus gordonii. The lower compartment was evaluated for turbidity over 150 days. Roots were split and examined for bacteria via SEM. The chi-squared test was used for comparisons (α = 0.05). Experimental groups had bacteria in their coronal thirds. Tubules contained bacteria in 90.9% and 91.7% of areas examined in the Biodentine and ProRootMTA groups, respectively, with no significant difference (p = 0.914). Experimental and negative roots had no turbidity with no significant difference between Biodentine and ProRootMTA (p = 1.000). Positive controls had turbidity. Bacteria penetrate roots via dentine tubules of root-filled teeth. Biodentine was comparable to ProRoot MTA.

Measuring of Alpha-1 Antitrypsin Concentration by Nephelometry or Turbidimetry.

Most clinical laboratories quantify alpha-1 antitrypsin using either nephelometry or turbidimetry techniques because they are commercially available, amenable to automation, and precise. Both methods are based on light scatter. The foundation of both techniques is based on incubation of the specimen with anti-AAT polyclonal antibody solution, a polymer matrix between endogenous AAT and the reagent antibodies forms, leading to production of light-scattering large particles. Although these two terms are sometimes used synonymously, technically speaking they are not.Nephelometry measures the amount of turbidity or cloudiness of a solution by directly quantifying the intensity of the light scattered by insoluble particles in the sample. Therefore, this technique measures the light that passes through the sample, with the detector being placed at an angle from the sample. Turbidimetry is the process of measuring the loss of intensity of the light transmitted linearly through a sample caused by the scattering effect of insoluble particles. The decrease in light transmission is measured compared to a reference, and the absorbed light is quantified.Beyond specific technical differences between both techniques, there are two major differences between the two procedures that may influence the results. First, the concentration of the sample and the resulting intensity of scattered light relative to the intensity of the light source is one major factor. Second, the size of the scattering particles is also a key differentiating factor. This chapter describes the technical requirements, the different protocols, and the clinical applicability of these two techniques in the diagnosis of alpha-1 antitrypsin deficiency.

Individual polysaccharide quantification in polyvalent pneumococcal conjugate vaccine using rate nephelometry.

The multivalent pneumococcal conjugate vaccine (PCV) contains purified polysaccharides of different serotypes conjugated to a carrier protein. Testing the final formulated product for individual serotype polysaccharide content is critical in vaccine quality control which requires an assay specific to each serotype polysaccharide present in the formulated product. Antibodies specific to the serotypes specific polysaccharide were used in rate nephelometry assay for quantifying individual serotype polysaccharides in the formulated vaccine. Generally, native polysaccharide (NP) have been used as reference standard. However, the polysaccharide antigen in the vaccine product is in the conjugate form (CRM197 linked) and hence using NP as a reference standard may not be suitable. Activated quenched polysaccharide (AQP) as a reference standard in rate nephelometry would be more appropriate. The epitope structure of AQP closely represents the polysaccharide-protein conjugate drug product (DP) after trypsin digestion. Hence, AQP was evaluated as a novel reference standard for the accurate and precise determination of individual polysaccharides in the multivalent DP. Rate nephelometry assay using AQP could be used for DP release and stability for monitoring time-dependent changes in the product and establishing the shelf life. A similar strategy could be applied to test and release monovalent or multivalent polysaccharide-protein conjugate vaccines (Meningococcal, Haemophilus influenza Type B, Typhoidal, and non-typhoidal salmonella).

Shape Discrimination of Individual Aerosol Particles Using Light Scattering.

We established an experimental apparatus by combining polarized light scattering and angle-resolved light scattering measurement technology to rapidly identify the shape of an individual aerosol particle. The experimental data of scattered light of Oleic acid, rod-shaped Silicon dioxide, and other particles with typical shape characteristics were analyzed statistically. To better study the relationship between the shape of particles and the properties of scattered light, the partial least squares discriminant analysis (PLS-DA) method was used to analyze the scattered light of aerosol samples based on the size screening of particles, and the shape recognition and classification method of the individual aerosol particle was established based on the analysis of the spectral data after nonlinear processing and grouping by particle size with the area under the receiver operating characteristic curve (AUC) as reference. The experimental results show that the proposed classification method has a good discrimination ability for spherical, rod-shaped, and other non-spherical particles, which can provide more information for atmospheric aerosol measurement, and has application value for traceability and exposure hazard assessment of aerosol particles.