Identificación y cuantificación automática de huevos de helmintos en muestras de agua residual (in Spanish) - Automatic identification and quantification of helminth eggs in wastewater samples

Pérez Sánchez, J. D. (2013)

Published in: 2013

Publisher:
MSc thesis, Universidad Nacional Autónoma de México (UNAM), Mexico

Author:
Pérez Sánchez, J. D.

Uploaded by:
SuSanA secretariat


3993 Views
107 Downloads

Content - Summary

Summary in English (in Spanish further down below):

Nowadays, the treatment applied to residual water can greatly reduce the concentration of pathogens, but it is still difficult to ensure complete removal, so the water is the vehicle for transmission of pathogenic organisms such as the helminths. Helminths are parasitic worms that infect communities with low levels of sanitation, causing intestinal diseases, hence the importance of identifying and quantifying them as accurately as possible.

Since the conventional method of identifying these organisms depends on the ability of humans to recognize them, is always needed the assistance of an expert in such identification, because it requires a very good training and a continual practice on analysis.

Considering this, we propose a new methodology based on image processing and classification for automatic identification and quantification of helminth eggs through optical microscope captured images, which aims to simplify the established analytical technique, reduce costs and analysis time. This thesis focused on the development of software for identifying and quantifying automatically helminth eggs images of different species, in order to develop a system for environmental control laboratories and research, not also at national level, it goes internationally.

Given the complexity of biological automatic image analysis, we performed a comparative study of protocols for image processing techniques applicable for this study. The development of automatic identification technique used 360 images of different species (Ascaris lumbricoides (fertile and infertile), Hymenolepis nana, H. diminuta, Schistosoma mansoni, Taenia sp., Toxocara canis and Trichuris trichiura), which allowed train system and establish a range of values for each property classification.

System validation is always performed with residual water samples, we commonly used three different qualities based on the total suspended solids (TSS). Class I was called to water 150 mg / L (TSS), where quality is typical of an untreated wastewater. This allowed us to validate the results according to the amount of solids present in water. For Class I and Class II, results were obtained identifying specificity 0.99 and 0.98, respectively, indicating that the system is able to distinguish between significant accuracy helminth egg and different objects. For the same quality of water, yielded a sensitivity of 0.83 and 0.80, respectively, indicating that the system's ability to identify a species among other exceeds 80%. The Class III identification efficiency was considerably lower (15%) than samples I or II. So in the current conditions of the system, must be carried out a prior dilution of the samples before identification and quantification through software.

The advantages of the developed system versus the traditional technique are: a) No specific skills required for the recognition of the species. b) Samples with different amount of total suspended solids can be identified in approximately 10 minutes against three hours or more of the traditional technique. c) The initial cost of the software is 30,000 compared to 20,000 dls of the traditional technique, but the operating cost per sample is less in the case of software with 2 dlls per sample versus 10 - 35 dlls for traditional technique. Regarding the specificity and sensitivity of the developed software, both features exceed 80%, while the traditional technique depends on previously acquired human skills.

++++++++++++++

RESUMEN (in Spanish)

En la actualidad el tratamiento aplicado al agua residual puede reducir en mucho las concentraciones de organismos patógenos; sin embargo, aún es muy difícil asegurar su completa eliminación, por lo que este tipo de agua se convierte en un vehículo de transmisión de organismos patógenos, como son los helmintos. Los helmintos son gusanos parásitos que en su estadio de huevo infectan a comunidades con bajo nivel de saneamiento, causando enfermedades intestinales, por ello la importancia de identificarlos y cuantificarlos con la mayor precisión posible. Dado que el método convencional de identificación de estos organismos depende de la capacidad del ser humano para reconocerlos, es necesaria siempre la intervención de personal experto en dicha identificación, pues al llevarse a cabo de manera visual, representa la principal fuente de incertidumbre y requiere de un muy buen entrenamiento y práctica continua por parte del analista. Considerando esto se propone una nueva metodología basada en el procesamiento y clasificación de imágenes para la identificación y cuantificación automática de huevos de helmintos a través de imágenes capturadas en microscopio óptico, con lo cual se busca simplificar la técnica analítica establecida y reducir los costos y el tiempo del análisis. Esta tesis se centró en el desarrollo de un software para identificar y cuantificar de manera automática imágenes de huevos de helmintos de diferentes especies, con la finalidad de que se pueda desarrollar en un futuro un sistema para laboratorios de control ambiental y de investigación no sólo a nivel nacional sino que pueda ser distribuido internacionalmente.

Dada la complejidad del análisis automático de imágenes biológicas, se realizó un estudio comparativo de los protocolos de reconocimiento disponibles así como de las técnicas de procesamiento de imágenes aplicables a este estudio. Para el desarrollo de la técnica de identificación automática se emplearon 360 imágenes de especies (Ascaris lumbricoides (fértil e infértil), Hymenolepis nana, H. diminuta, Schistosoma mansoni, Taenia sp., Toxocara canis y Trichuris trichiura), las cuales permitieron entrenar al sistema y establecer un intervalo de valores para cada propiedad de clasificación. La validación del sistema siempre se llevó a cabo con muestras de agua residual y se utilizaron comúnmente tres calidades diferentes con base al contenido de sólidos suspendidos totales (SST). La clase I se denominó para agua con 150 mg/L SST, donde la calidad es la propia de una agua residual sin tratamiento; esto permitió validar los resultados según la cantidad de sólidos presentes.

Para el agua clase I y clase II, se obtuvieron resultados de especificidad de la identificación de 0.99 y 0.98, respectivamente, lo que indica que el sistema es capaz de distinguir con significativa exactitud entre un huevo de helminto y objetos diferentes. Para esas mismas calidades de agua, se obtuvo una sensibilidad de 0.83 y 0.80, respectivamente, lo que indica que la capacidad del sistema para identificar una especie entre las demás supera el 80%. Para el caso del agua clase III, la eficacia de identificación fue considerablemente menor (15%) que para las muestras de clase I o II. Así que en las condiciones actuales del sistema y para el caso de muestras de aguas residuales con alto contenido de sólidos, se debe llevar a cabo, una dilución previa de las muestras antes de su identificación y cuantificación a través del software.

Las ventajas del sistema desarrollado frente a la técnica tradicional son: a) No se requieren habilidades específicas del personal para el reconocimiento de las especies. b) Muestras con diversa cantidad de sólidos suspendidos totales pueden ser identificadas en aproximadamente 10 minutos contra tres horas o más de la técnica tradicional. c) El costo inicial del software es de 30,000 dls frente a 20,000 dls de la técnica tradicional, pero el costo operativo por muestra es menor en el caso del software, con 2 dls por muestra contra un mínimo de 10 dls y máximo 35 dls de la técnica tradicional. En cuanto a la especificidad y sensibilidad del software desarrollado, ambas características superan el 80%, mientras que la técnica tradicional depende de las habilidades humanas previamente adquiridas.

Bibliographic information

Pérez Sánchez, J. D. (2013). Identificación y cuantificación automática de huevos de helmintos en muestras de agua residual (in Spanish) - Automatic identification and quantification of helminth eggs in wastewater samples. MSc thesis, Universidad Nacional Autónoma de México (UNAM), Mexico

Filter tags

Latin America & Caribbean Spanish

Download

Identificación y cuantificación automática de huevos de helmintos en muestras de agua residual (MSc thesis)

Format: pdf file
Size: 3.9 MB

Share this page on    


Follow us on    

SuSanA Partners currently 351 partners

Sanitation events

SuSanA newsletter

Stay informed about the activities of SuSanA and its partners. The SuSanA newsletter is sent out around four times per year. It contains information about news, events, new partners, projects, discussions and publications of the SuSanA network.

Subscribe to newsletter »

Latest tweets mentioning SuSanA

Tweets mentioning SuSanA »

28th SuSanA meeting, Stockholm

The next SuSanA meeting will take place in Stockholm, Sweden. We will meet on the 24th of August, just before the World Water Week 2019 begins. More information will be available soon.

Find out more about the meeting! 

SuSanA meetings

Jobs

 


close  

 

Resources and publications

Our library has more than 2300 publications, factsheets, presentations, drawings etc. from many different organisations. It continues to grow thanks to the contributions from our partners.

Add item to library »

The three links below take you to special groups of items in the library for more convenient access:

Projects

The project database contains nearly 400 sanitation projects of many different organizations dealing with research, implementation, advocacy, capacity development etc. Advanced filtering functions and a global map are also available. Information on how and why this database was created is here.

People working for SuSanA partners can add their own projects through their partner profile page. You might need your SuSanA login upgraded for this purpose. Please contact us if you would like to add a project.


Trainings, conference and events materials

Missed important conferences or courses? Catch up by using their materials for self study. These materials have been kindly provided by SuSanA partners.

Regional chapters

Use the map or the search tool to access the most relevant information and knowledge products for your region or country. This includes relevant resources, events, partners or projects.

Shit flow diagrams (SFDs), excreta flow diagrams

Shit flow diagrams (SFDs) help to visualize excreta management in urban settings. Access SFDs and more through the SFD Portal.

 


close  

 

Discussion forum

Share knowledge, exchange experiences, discuss challenges, make announcements, ask questions and more. Hint: Your discussion forum login is the same as your SuSanA login. More about the forum's philosophy »


Integrated content

We are hosting content from some other communities of practice and information-sharing portals. This section also provides a link to SuSanA's Sanitation Wikipedia initiative.

Suggest content to add »

SuSanA partners

Not yet a SuSanA partner? Show your organisation's support to SuSanA's vision and engage in  knowledge sharing by becoming partners.

Apply to become a partner »


Individual membership

Register as an individual member of SuSanA free of charge. As a member you can interact with thousands of sanitation enthusiasts on the discussion forum.  You can also get engaged in one of our 13 working groups and our regional chapters. Our FAQs explain the benefits further.

By getting a SuSanA login you can fully participate in the SuSanA community!

Register as a member

Login


Forgot your password?

 


close