Comparative Analysis of Automatic Fecal Analyzer versus Direct Wet Smear Microscopy for Detecting Parasitic Infections in Stool Samples

Summary

Parasitic infections pose a significant public health challenge, especially in developing countries and regions. This study compares an Automatic Fecal Analyzer to assess its diagnostic efficacy with that of the traditional Direct Wet Smear Microscopy method.

Abstract

With socio-economic development, the prevalence of intestinal parasitic diseases has significantly decreased year by year. However, parasitic infections remain a major public health issue globally, particularly in developing countries and regions. Timely diagnosis and treatment are crucial for controlling the spread of these diseases. The traditional Direct Wet Smear Microscopy method, while widely used, is labor-intensive, prone to contamination, and dependent on the skills of the technician. This paper introduces an Automatic Fecal Analyzer, which automates the stool sample processing, offering advantages over the traditional Direct Wet Smear Microscopy method, such as ease of operation, rapid detection, a clean and hygienic working environment, and high sensitivity and specificity, thus enhancing diagnostic efficiency and accuracy.

We compared three different methods for fecal analysis: direct wet smear microscopy method, automatic fecal analyzer (AI report), and automatic fecal analyzer (user audit). The AI report uses automated image analysis and machine learning algorithms to identify components like parasites and eggs in fecal samples. This method can process a large number of samples quickly, increasing efficiency. The User Audit also uses an automatic fecal analyzer but includes an additional step of user audit. Experienced technicians review the AI report to enhance the accuracy and reliability of the results.The analyzer demonstrated a sensitivity of 84.31% for AI report and 94.12% for user audits, along with a specificity of 98.71% for AI reports and 99.69% for user audits, making it an invaluable tool for the clinical diagnosis and treatment of parasitic infections.

Introduction

Globally, intestinal parasitic diseases persist as a substantial public health issue, predominantly in developing nations. Historical data from the Health and Welfare Commission of the People's Republic of China reveal a marked decline in infection rates from 62.63% in 1992 to 5.96% in 2015^1,2,3. Despite its dramatic decline¹, with a weighted national prevalence of 0.47% and an estimated 5.98 million infected individuals³, the prevalence of Clonorchis sinensis infection in the Guangdong population remains notably high at 4.90%⁴. Specific regions like Guangdong still exhibit higher rates of C. sinensis infection, underscoring the need for precise diagnostic methods.

Diagnosing parasitic diseases can be achieved through microscopic examination, medical imaging, serological, and molecular methods. Serological analysis using enzyme-linked immunosorbent assay (ELISA) is helpful in the diagnosis of parasitic disease, but the sensitivity is moderate, at approximately 80%⁵. Molecular biology diagnostic techniques have high sensitivity and specificity⁶, but they are hindered by cumbersome, time-consuming operations, and high test costs, making them unsuitable for large-scale routine testing.

The detection of C. sinensis eggs in the patient's feces is the basis for confirming the diagnosis. The Kato-Katz thick smear method, which looks for C. sinensis eggs, is the classic method commonly used to diagnose parasitic diseases^7,8. Although inexpensive, it is time-consuming, labor-intensive, and unsuitable for clinical practice in the face of a high volume of specimens to be screened. Most primary hospitals currently employ the Direct Wet Smear Microscopy method. However, this method is still time-consuming and labor-intensive, with unpleasant fecal odor, and is prone to cross-contamination, high biosecurity risks, and results that are contingent upon the expertise and diligence of the testing personnel⁹.

Conventional diagnostic techniques, while valuable, face limitations in sensitivity and practicality. However, with the ongoing advancements in science and technology, fecal testing is becoming more automated and standardized. Automated instruments that utilize digital image-based methods for quantitatively detecting parasites or parasite eggs in feces are gaining popularity in clinical laboratories. They offer advantages such as ease of operation, rapid detection, a clean and hygienic working environment, and high sensitivity and specificity. The Automatic Fecal Analyzer uses digital imaging technology to document microscopic observations by capturing a series of low and high-magnification digital photographs through the microscope and its integrated digital imaging system. The system then transmits these images to a computer data processing system, which identifies and analyzes them accordingly. Recently, we have attempted to diagnose a spectrum of parasitic infections, focusing on the detection of parasites and their eggs in fecal samples. Positive samples are identified by the presence of their morphological characteristics. We evaluated the Automatic Fecal Analyzer for detecting parasites in fecal samples and confirmed its potential as an adjunct to conventional microscopy methods. The Automatic Fecal Analyzer is composed of a mainframe and accompanying software featuring several specialized units, including an automatic sampling unit, a puncture sampling unit, a sample characteristics and color observation unit, a camera unit, and a reagent card unit (Figure 1).

Figure 1: Components of the instrument. Please click here to view a larger version of this figure.

In this study, 2,305 fresh stool samples from inpatients of Guangdong Provincial People's Hospital in July 2024 were evaluated in a double-blind manner using the Automatic Fecal Analyzer and the Direct Wet Smear Microscopy method (Figure 2). One group conducted tests using a Direct Wet Smear Microscopy method, while the other used an Automatic Fecal Analyzer with an AI-generated report. After documenting the outcomes from each group, a third group of experimenters manually reviewed the images captured by the automatic analyzer. They scrutinized for suspected parasites or eggs, images that deviated from the automatic analysis, blurred or missing images, and those with complex backgrounds containing excessive residue. This process involved correcting the results, which were then referred to as the Automatic Fecal Analyzer's user audit. Ultimately, a fourth individual summarized and statistically analyzed the findings. The efficacy of the fully automated fecal instrumentation was assessed through a double-blind detection of parasites/eggs, comparing the detection time and results across the three methodologies.

Figure 2: Schematic illustration of the research plan. Please click here to view a larger version of this figure.

Protocol

This study was approved by the Ethical Review Committee of Guangdong Provincial People's Hospital, Southern Medical University, Guangzhou, China (Approval number: KY2024-445-02). All participants signed a written informed consent before enrollment. The reagents and equipment used for the study are listed in the Table of Materials.

1. Specimen collection

1. Utilize the Automatic Fecal Analyzer Manufacturer's Specimen Collection Tubes, ensuring they are capped, sealable, clean, dry, non-leaking, and have an appropriate opening and volume for multi-point sampling.
2. Select as much fresh abnormal fecal matter as possible, containing mucus, pus, or blood, with a target amount of approximately 0.5 g. When feces are dilute or watery, employ a disposable pipette to collect the necessary sample volume.
   NOTE: Care must be taken to prevent contamination from all non-patient fecal samples. This includes, but is not limited to, feces from enemas or oil-based laxatives, samples from bedpans or toilets, feces contaminated by plants, soil, sewage, disinfectants, detergents, urine, menstrual blood, or leucorrhea.

2. Fecal testing methods

1. Direct Wet Smear Microscopy method^10,11
   1. Apply 1-2 drops of saline solution to a clean microscope slide.
   2. Take a small sample from the abnormal area or from multiple areas of the feces for smearing.
   3. Spread the sample evenly on the slide. Adjust the thickness of the smear to ensure visibility of the details.
   4. Place a coverslip over the smear, avoiding air bubbles.
   5. Scan more than 10 fields of view at low magnification (10x). If parasite eggs or parasites are detected, switch to high magnification (40x) for detailed morphological observation.
      NOTE: A low-magnification examination of the slide in a 'battlements' sequence is recommended to minimize the risk of misses.
2. Automatic Fecal Analyzer (Figure 3)
   1. Power on the computer and analyzer, click to start the software, enter the username and password login, and wait for the instrument to perform a self-check.
   2. Collect the qualified feces using the appropriate sample collection tube.
      NOTE: Fresh abnormal feces containing mucus, pus, or blood weighing approximately 0.5 g is the preferred sample.
   3. Place the sample collection tube in the sample holder of the analyzer, place the sample rack on the automatic sample inlet, and click Start for the first test after the self-check.
      NOTE: Subsequent tests will be automatically detected and conducted by the instrument. The instrument robot will photograph the sample for documentation.
   4. Add 5 mL of diluent to the sample tube, mix well by pneumatic mixing, and allow the sample to settle before analysis.
      NOTE: The specimen collection tube is lined with a special filter to remove residues of no pathological significance, such as large amounts of plant fibers, seeds, and undigested food residues.
   5. Let the sample flow into the counting cell for analysis where the analyzer will capture high-resolution microscopic images under the microscope and the software will process and recognize the images.
   6. Switch to the Result interface, check the test images, and send the result to Laboratory Information Management System (LIS) or print them.
   7. Click Stop to end the test.
   8. Select Flush and Shutdown.
   9. Load the rack of concentrated clearing liquid. The instrument will automatically flush and shutdown.

Figure 3: Automated Fecal Analyzer System Workflow. Please click here to view a larger version of this figure.

3. Methodological evaluation

1. Calculation of detection time
   NOTE: Detection times were calculated by recording the start and end times of each test procedure. Test results were compared by analyzing the sensitivity and specificity of each method in detecting parasites.
   1. Direct Wet Smear Microscopy method
      1. For each fecal sample, measure the time taken to pick up rice grain-sized feces and spread them evenly on a glass slide.
      2. Measure the time taken to read the results for each participant.
      3. Add the swabbing and reading times for each participant.
      4. Multiply the sum from step 3.1.1.3 by the number of participants to get the total test time for the Direct Wet Smear Microscopy method (T₁).
      5. Divide T₁ from step 3.1.1.4 by the number of samples tested to get the average test time for the Direct Wet Smear Microscopy method (t₁).
   2. The Automatic Fecal Analyzer (AI report)
      NOTE: The process can be performed by a single person.
      1. Measure the Automatic Fecal Analyzer test time for each sample (T₂).
      2. Divide T₂ by the number of samples tested to get the average test time for the Automatic fecal analyzer (AI Report) (t₂).
   3. The Automatic Fecal Analyzer (user audit)
      1. Measure the Automatic Fecal Analyzer test time for each sample (see step 3.1.2).
      2. Measure the manual image review time for each sample. Start the timer (stopwatch or digital timer) as soon as the review of the images captured by the Automatic Fecal Analyzer from a sample starts. Carefully examine the images to identify any relevant components or abnormalities. Stop the timer once the review for that sample is complete, note down the time taken for each sample review, and enter the data into a spreadsheet for further analysis.
      3. Add the Automatic Fecal Analyzer test time and manual image review time for each sample.
      4. Multiply the sum from step 3.1.3.3 by the number of participants to get the total test time for the Automatic Fecal Analyzer (user audit) (T₃).
      5. Divide T₃ by the number of samples tested to get the average test time for the Automatic Fecal Analyzer (user audit) (t₃).
2. Comparison of test results
   NOTE: The performance of the Automatic Fecal Analyzer's AI-generated report and the user audit results were assessed using the direct wet smear method as the benchmark.
   1. Collect the results from the Direct Wet Smear Microscopy method, which will serve as the benchmark.
   2. Collect the results from the Automatic Fecal Analyzer's AI-generated report.
   3. Collect the results from the user audit of the Automatic Fecal Analyzer.
3. Calculate the Evaluation Metrics.
   1. Consider a sample to be a True Positive (TP) when the results of the Automatic Fecal Analyzer (AI report or user audit) and the Direct Wet Smear Microscopy method both indicate the presence of the target analyte.
   2. Consider a sample to be a True Negative (TN) when both the results of the Automatic Fecal Analyzer (AI report or user audit) and the Direct Wet Smear Microscopy method indicate the absence of the target analyte.
   3. Consider a sample to be a False positive (FP) when the AI report or user audit of the Automatic Fecal Analyzer indicates the presence of the target analyte, but the Direct Wet Smear Microscopy method does not confirm its presence.
   4. Consider a sample to be a False Negative (FN) when the AI report or user audit of the Automatic Fecal Analyzer indicates the absence of the target analyte, while the Direct Wet Smear Microscopy method actually detects its presence.
   5. Calculate Specificity as follows:
      ​Specificity= True Negatives (TN) / (True Negatives + False Positives)
   6. Calculate Sensitivity as follows.
      ​Sensitivity= True Positives (TP) / (True Positives + False Negatives)
   7. Calculate Positive Predictive Value (PPV) as follows:
      PPV = True Positives (TP) / (True Positives + False Positives)
   8. Calculate Negative Predictive Value (NPV) as follows:
      ​NPV = True Negatives (TN) / (True Negatives + False Negatives)
   9. Calculate the Kappa Value using statistical software.
   10. Calculate the Compliance Rate as follows:
       ​Compliance Rate = Agreements (both AI and user report match the benchmark) / Total Samples.
   11. Calculate Positive detection rate as follows:
       Positive detection rate = Number of Positives / Total Samples.

4. Analyze the results

1. Enter the data in spreadsheets and analyze them statistically.
2. For count data, use rates and proportions for description, and compare between groups using the paired four-tailed chi-squared test.
3. Compare mean detection times between the two methods using the independent samples T-test.
4. Assess the consistency in parasite species detection between techniques via Kappa (κ) agreement.
5. Consider P < 0.05 to be statistically significant.

Results

Comparison of detection times between the three detection methods
The average detection time was 1.55 min per person per sample for the Direct Wet Smear Microscopy method, 1.30 min per person per sample for the Automatic Fecal Analyzer (AI report), and 1.47 min per person per sample for the Automatic Fecal Analyzer (user audit). The Direct Wet Smear Microscopy method required 1.1x the duration compared to the Automatic Fecal Analyzer (AI report) (t = 2.442, P = 0.020). Conversely,...

Discussion

Parasitic diseases represent a significant global public health concern¹², particularly in developing countries where they lead to health issues such as malnutrition, stunting, anemia, and biliary obstruction¹³, and are also linked to inadequate socio-economic development and education. Fecal testing for parasites has become a routine part of clinical practice, with accuracy, speed, and efficiency being key requirements for such tests.

Parasite d...

Materials

Name	Company	Catalog Number	Comments
Automatic Fecal Analyzer (FA280S) software	ORIENTER	v.1.028(RecoV2.9.14)	The software of the Automatic Fecal Analyzer (FA280S) system
Concentrated Cleaning Fluid	ORIENTER		Suitable for pipeline cleaning
Cover GLASS	SAIL BRAND		Protect the specimen from damage and contamination.
Disposable plastic straw	Nantong Lingke Biotechnology		Disposable plastic straw used for aspirating physiological saline
FA280S	ORIENTER	A022201035	Fully automatic feces analyzer
Flushing Fluid	ORIENTER	Q/WWT-019-2021	Flushing pipeline of fully automatic feces analyzer produced
Microscope	OLYMPUS	3H51278	Observation of cells
Microscope slides		WS-7101	They provide a stable, flat surface to mount specimens for examination under a microscope.
Sample Diluent	ORIENTER	Sichuan Rong Machinery 20150025	Suitable for dilution of stool samples
Sample rack	ORIENTER		Loading racks
Sample tube	ORIENTER	Sichuan Rong Machinery 20200204	Faecal Collector
Sodium Chloeide Injection	Guangdong Daxiang Pharmaceutical Co.	H14020786	Dilution
SPSS statistical software	IBM	SPSS Statistics 22.0	Suitable for a variety of research and analytical tasks
Wooden stick			A disposable wooden stick used to pick an appropriate amount of fecal sample.