WO2021247408A1 - Système d'ia et de données pour surveiller des agents pathogènes dans des eaux usées et procédés d'utilisation - Google Patents

Système d'ia et de données pour surveiller des agents pathogènes dans des eaux usées et procédés d'utilisation Download PDF

Info

Publication number
WO2021247408A1
WO2021247408A1 PCT/US2021/034837 US2021034837W WO2021247408A1 WO 2021247408 A1 WO2021247408 A1 WO 2021247408A1 US 2021034837 W US2021034837 W US 2021034837W WO 2021247408 A1 WO2021247408 A1 WO 2021247408A1
Authority
WO
WIPO (PCT)
Prior art keywords
pathogen
wastewater
data
learning
levels
Prior art date
Application number
PCT/US2021/034837
Other languages
English (en)
Inventor
Bryan Walser
Original Assignee
Pangolin Llc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Pangolin Llc filed Critical Pangolin Llc
Publication of WO2021247408A1 publication Critical patent/WO2021247408A1/fr

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/80—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for detecting, monitoring or modelling epidemics or pandemics, e.g. flu
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Definitions

  • the invention generally relates to wastewater sampling and pathogen surveillance.
  • COVID-19 is a highly pathogenic respiratory disease, which exhibited an outbreak after its first appearance in Wuhan, China in December 2019.
  • COVID-19 is caused by a novel coronavirus namely SARS-CoV-2, which causes respiratory illness with elevated fatality rate in patients, including patients with one or more of comorbidities such as obesity, hypertension and diabetes.
  • SARS-CoV-2 a novel coronavirus
  • Cases of COVID-19 in which the patient shows no symptoms of infection appear asymptomatic but may still infect or transmit the virus to the community, state, or country.
  • the Centers for Disease Control and Prevention confirms that the virus has been found in the feces of patients diagnosed with COVID-19, and is thus gathering in city sewers.
  • the surveillance of sewage or waste water for COVID-19 may provide details on the epidemiology and accelerate governments’ efforts to contain the viral outbreaks, and save human lives.
  • the present invention attempts to solve these problems, as well as others.
  • the invention does not intend to encompass within the scope of the invention any product, process, or making of the product or method of using the product, which does not meet the written description and enablement requirements of the USPTO (35 U.S.C. ⁇ 112, first paragraph) or the EPO (Article 83 of the EPC), such that Applicants reserve the right and hereby disclose a disclaimer of any previously described product, process of making the product, or method of using the product. It may be advantageous in the practice of the invention to be in compliance with Art. 53(c) EPC and Rule 28(b) and (c) EPC.
  • FIG. 1A is a perspective view of a wastewater system.
  • FIG. IB is a schematic flow chart of the AI system coupled with wastewater pathogen detection system.
  • FIG. 1C is a schematic flow chart for a wastewater detection system.
  • FIG. 2A is a schematic flow chart of a neural network training structure.
  • FIG. 2B is a schematic flow chart of a single node ANN.
  • FIG. 2C is a schematic flow chart of the different stages of the supervised learning.
  • FIG. 2D is a schematic flow chart of the different stages of the unsupervised learning.
  • FIG. 2E is a schematic flow chart of the different stages of the semi-supervised learning.
  • FIG. 3A is a schematic flow chart of the Transfer Learning Approach.
  • FIG. 3B is a schematic flow chart of the Convolutional Neural Network.
  • FIG. 3C is a schematic flow chart of the architecture of the ANFIS.
  • FIG. 4A is a schematic flow chart a continuous policy feedback loop.
  • FIG. 4B is a schematic flow chart for an afferent clinical feedback loop.
  • references to “one embodiment,” “an embodiment,” “example embodiment,” “various embodiments,” etc., may indicate that the embodiment(s) of the invention so described may include a particular feature, structure, or characteristic, but not every embodiment necessarily includes the particular feature, structure, or characteristic. Further, repeated use of the phrase “in one embodiment,” or “in an exemplary embodiment,” do not necessarily refer to the same embodiment, although they may.
  • the term “method” refers to manners, means, techniques and procedures for accomplishing a given task including, but not limited to, those manners, means, techniques and procedures either known to, or readily developed from known manners, means, techniques and procedures by practitioners of the chemical, biological, industrial, electrical, software, and mechanical arts, as well as public health, public policy, and healthcare. Unless otherwise expressly stated, it is in no way intended that any method or aspect set forth herein be construed as requiring that its steps be performed in a specific order. Accordingly, where a method claim does not specifically state in the claims or descriptions that the steps are to be limited to a specific order, it is no way intended that an order be inferred, in any respect. This holds for any possible non-express basis for interpretation, including matters of logic with respect to arrangement of steps or operational flow, plain meaning derived from grammatical organization or punctuation, or the number or type of aspects described in the specification.
  • AI Artificial Intelligence
  • the AI system provides real-time sampling, modeling, analysis, and/or recommended policy interventions.
  • the AI system identifies active cases in an outbreak through pro-active pathogen wastewater sampling in any geographical region and including high risk locations, such as hospitals, schools, airports, churches, production facilities, or crowded commercial areas, where pathogen transmission can occur rapidly.
  • the AI system provides containment measures for surrounding cities, states, or countries to slow the transmission of the pathogen and measures pathogen level in wastewaters to ensure when containment measures may be removed or other restrictions lessened.
  • the AI system comprises computer modeling and simulation for predictive analysis.
  • computer models successfully track a citywide, a s nationwide, or a nationwide spread of pathogen in the wastewater.
  • the AI system may use previous tracked pathogens in the wastewater to predict the spread of the pathogen in wastewater in a period of time thereafter.
  • the AI system provides up-to-date data on the spread of pathogen in geographical locations, due to transportation and human movement, and applies optimal control strategies to early pathogen transmission so as to prevent an epidemic or a pandemic through public health interventions and changes to population behavior, including with respect to vaccines, therapeutics, or other modalities for intervention.
  • the present invention is an AI system that applies to separate sanitary sewer systems, combined sewer systems, or standalone sewer systems.
  • Separate sanitary sewer systems are designed to transport sewage alone in underground pipe or tunnel system from houses and commercial buildings to treatment facilities or disposal. In municipalities served by sanitary sewers, separate storm drains may convey surface runoff directly to surface waters.
  • Sanitary sewers are part of an overall system called a sewage system or sewerage. Sanitary sewers are distinguished from combined sewers, which combine sewage with storm water runoff in one pipe.
  • a combined sewer system is a sewage collection system of pipes and tunnels designed to simultaneously collect surface runoff and sewage water in a shared system.
  • a standalone sewer system may be an open system or septic tank where sewage is collected for home or small building where there is no separate or combined sewer system.
  • Wastewater system herein means a separate sewer system, a combined sewer system, or a standalone sewer system that is defined by a geographical region, generally shown with the wastewater system in FIG. 1 A.
  • Wastewater herein means the type of wastewater that is produced by a community of people. Wastewater is characterized by volume or rate of flow, pressure within the enclosed pipe, physical condition, chemical and toxic constituents, and its bacteriologic, virological, or pathogenic status (which organisms it contains and in what quantities). Wastewater consists mostly of greywater (from sinks, bathtubs, showers, dishwashers, and clothes washers), blackwater (the water used to flush toilets, combined with the human waste that it flushes away); soaps and detergents; and toilet paper.
  • the wastewater system can be a pressurized system with variable wastewater flow rates due to levels of diluent in the system, blockages, etc.
  • the system may incorporated for testing finished drinking water, treated water, disinfected water, irrigation water, and water obtained from wells, rivers, lakes and recreational waters such as swimming pools.
  • the AI system samples and analyzes food (such as fruits, vegetables, meat and prepared food items), swabs taken from slaughter lines, and meat surfaces, as well as swabs taken from environmental surfaces from slaughter houses, and meat preparation facilities, soil and clinical and veterinary samples including stool and biopsy samples.
  • the AI system 10 to monitor sewage or untreated wastewater for pathogen states in a city, state, or country and comprises detecting a pathogen in the wastewater system 12, storing the pathogen detection data in a database or computer readable medium 14, analyzing the pathogen detection data by data analysis, machine learning, algorithms, and modeling 16; and providing temporal information on pathogen increasing or decreasing levels, geographical information on the pathogen transmission, and containment measure information 18, and providing feedback 20 for containment measures or to adjust to varying parameters to decrease pathogen transmission.
  • a pathogen is a bacterium, virus, or other microorganism that can cause disease.
  • Pathogens detectable by the system include, but are not limited to: prions, bacterium, viruses, fungi, algae, or parasites, listed in Table 2.
  • the pathogen is the virus SARS-CoV-2.
  • the AI system comprises interpreting data through an algorithm or platform to interpret data over time and assess foreground variables and background variables for the detection of the pathogen in wastewater.
  • the AI system 100 may be implemented in a wastewater system comprising accessing the wastewater system 110, retrieving wastewater samples 120, detecting the pathogen in wastewater 130, data generation 140, data retrieval 150, data interpretation 160, data presentation 170, implementation 180, implementation evaluation 190, and updating the detector 132.
  • the methods and steps for the AI system 100 may be combined in any particular order or combination to provide for wastewater pathogen detection and appropriate response for population behavior, adaptation of the pathogen detection, and penetration of pathogen transmission in wastewater after containment measures are deployed geographically.
  • the AI system comprises using a mapping algorithm to provide geolocation in the wastewater system to sample for the pathogen or where to place the detectors in the wastewater system, including separate standalone locations and topographical locations factoring volumetric wastewater flow rates and velocities of the wastewater.
  • accessing the AI system 100 comprises using a timing algorithm to process and compute the flow of wastewater, pressure of wastewater, the season of detection, and contaminants in the wastewater.
  • the algorithms may estimate wastewater loading or fecal loading using historic wastewater meter data and population density, according to one embodiment. Algorithms are further described below.
  • the AI system 100 is operably coupled with a server and a network communicating data and information from the retrieving wastewater samples 120, detecting the pathogen in wastewater 130, data generation 140, data retrieval 150, data interpretation 160, data presentation 170, implementation 180, implementation evaluation 190, and updating the detector 132, as shown in FIG. 1C.
  • Each part of the AI system may operate through a computer or program module to execute a set of instructions, protocols, or parameters and be updated through the database or a distributed Internet of Things (IoT) system.
  • the server may be operably coupled with a database to store information or retrieve information. Wastewater systems and IoT systems used with the AI system are disclosed in U.S. provisional application serial no. 63/030,017, filed May 26, 2020. PCR methods and non-PCR methods to detect the pathogen in wastewater may be used as described in commonly assigned U.S. provisional application serial no 63/032,565, filed May 30, 2020
  • the AI system comprises a data generator 140 to indicate the level of the pathogen in a city, state, or country, accounting for the population and levels of the wastewater.
  • the data generator 140 may indicate rising levels of the pathogen above an infection level, threshold or rate of detection indicates that the city, state, or country may require containment measures to delay or stop the spread of the pathogen.
  • the threshold level of pathogen considers the population above N people, N% of relevant population, and size of population.
  • the infection level or threshold of the pathogen may be above about 0.1 million, about 0.2 million, about 0.3 million, about 0.4 million, about 0.5 million, or about 0.6 million genome units fL of wastewater, according to one embodiment.
  • the infection level or threshold of the pathogen may be above about 10, about 20, about 30, about 40, or about 50 genome units per mL of wastewater.
  • the infection level or threshold of the pathogen may be above about 10, about 20, about 30, about 40, or about 50 genome units per mL of wastewater.
  • for the COVID-19 pathogen as many as about 600,000 to about 30,000,000 viral genomes of SARS-CoV-2 per mL of fecal material, assuming a fecal load of about 100-400 g feces/day/person with a density of about 1.06 g/mL. Normalization of population of the pathogen’s infection level in the wastewater ensures that a significant increase in pathogen concentration in a wastewater sample does not correspond to an increase in population in the serviced wastewater area or a decrease in the amount of diluent as might occur during a dry season or drought.
  • Normalization of the pathogen’s infection level includes factoring cultural practices and hygiene and sanitation, concentration of the pathogen in the wastewater, seasonal and rain effects on the levels of diluent (which may be higher due to hurricanes, monsoons, or rainy seasons, or lower due to drought, cultural practices, or dry season) as well as other factors affecting concentration of the pathogen without influencing the amount of pathogen actually delivered to the system.
  • the AI system includes the data generator 140 correlating with clinical data and containment measures.
  • Containment measures may include stay at home orders from the government, social distancing in public places, allowing only essential businesses to remain open, a requirement to wear masks in public, creation of non-acute and acute care sites for hospital patient overflow, pharmaceutical treatments, production and demand for pharmaceuticals with pathogen treatment, activation of vaccine development and production, or a requirement for all health care workers to wear Personal Protective Equipment (PPE) whenever interacting with an admitted or sick patient. Falling levels of the pathogen below an infection level or particular threshold indicates that the community, state, or country no longer has the risk of pathogen transmission and may remove some or all of the containment measures for the city, state, or country.
  • PPE Personal Protective Equipment
  • the infection levels of the pathogen may correlate with clinical data from the number of positive cases or deaths in the community, state, or country.
  • the containment measures are to ensure the epidemiological curve of the pathogen remains low and the number of critically ill patients does not overburden the health system of the city, state, or country. Containment measures and pathogen detection may ensure that a second wave of epidemics or pandemics are prevented or delayed.
  • the AI system provides for the acquisition and analysis of pathogen detection in wastewater over a period of time in combination with continuous policy feedback, a clinical afferent loop, a biological afferent loop, and a virological afferent loop.
  • the AI system includes inputs from clinical systems, biological systems, or virological systems.
  • This automated artificial intelligence system encompasses monitoring, screening, diagnosing, and performing prognosis of disease(s) and condition(s) by integrating primary and secondary genomic information of the pathogen, pathogen profiles, environmental information, wastewater profiles, geographical information, and disease models, using proprietary algorithms. The resulting information can be used for population analysis as well as for containment measures, particularly preventing outbreaks and pandemics.
  • the AI system links to external clinical data bases; processes information in real-time; uses the Internet and other wireless technologies to transmit or receive information; provides access to information that is useful in managing disease outbreaks and emergency situations; provides tiered information to cities, states, or countries; performs simultaneous, multi-dimensional analysis; and analyzes pathogen information by city, region, state, country, and other geographies.
  • the AI system is a real time, dynamic decision making tool that can be used with a clinical system, and with the information obtained from environment contexts. Access to this AI system will allow the analysis of both clinical and non- clinical information, such as that contained within GIS mapping applications.
  • the AI system comprises providing a continuous policy feedback system by the detection of the pathogen in the wastewater system and establishing a clinical afferent loop, according to one embodiment.
  • the continuous policy feedback 200 is shown in FIG. 4A, where the detection of the pathogen in the wastewater 210 is continuous and in real time and allows for the detection of the levels of the pathogen indicating a disease state 220 in the city, state, or country. If the detection of the pathogen is above an infection level or threshold 230, then containment measures are deployed in real time 240, as indicated above. In the case of asymptomatic conditions, clinical testing of the pathogen of admitted or deceased patients 250 proceeds to confirm the presence of the pathogen in admitted or deceased patients.
  • the AI system comprises a clinical afferent loop coincides with the levels or thresholds of the pathogen in the wastewater and provides a city, state, or country actionable or containment measures to curb the spread of the pathogen and prevent outbreaks and transmission of the pathogen.
  • the clinical afferent loop 300 is shown in FIG. 4B, and comprises detection of the pathogen in the wastewater 310 and allows for the detection of the infection levels of the pathogen indicating a disease state 320 in the city, state, or country. If the detection of the pathogen is above a level or threshold 320, then clinical testing of the pathogen of admitted patients 330 proceeds to confirm the presence of the pathogen in admitted patients. Testing of admitted patients or deceased patients for the pathogen confirms presence of the pathogen correlated with the pathogen detection in the wastewater. Comparison and correlation with clinical data in the surrounding cities, states, or countries confirms the spread or containment of the pathogen, where the comparison or correlation is processed by Artificial Intelligence systems operably coupled with the server or database.
  • AI systems may provide prospective information or retrospective information regarding the levels of the pathogen, considering factors of the wastewater conditions, population, type of pathogen, clinical data, hospital capacity, population, and the like. If the detection of the pathogen is not above the level or threshold 320, then continuous pathogen detection 310 proceeds or adapts to the detection levels through sampling, attributes, AI systems, and detector feedback, as indicated herein. If the detection of the pathogen a level or threshold is not below an infection level or disease state 340, then additional clinical testing of admitted patients initializes 330, as well as other proactive measures such as creating non-acute and acute care sites. Once the levels of the pathogen are below are threshold or level 340, then the clinical testing of admitted patients requirement may be potentially removed.
  • the AI system comprises data retrieval 150 detecting at the most distal location of the city, state, or country and includes non-wastewater system collection (septic tanks), non-wastewater system customers, for the most distal (far from the treatment center) location of the city, state, or country.
  • the timing of the pathogen detection is less than about 12 hours, less than about 24 hours, less than about 48 hours, since pathogens have a rapid generation time.
  • the generation time N is 8 generations, alternatively about 10 generations, alternatively about 12, alternatively N ⁇ X, where X is arbitrary or distinct for the pathogen.
  • the data interpretation 160 comprises interpreting data through an algorithm or Artificial Intelligence (AI) platform.
  • the AI platform may interpret data over time and assess foreground variables and background variables for the detection of the pathogen in wastewater.
  • the data interpretation 160 comprises digital health options for the population or at upstream facilities.
  • the AI system comprises detecting the pathogen or sampling the wastewater at each interlock or interceptor and mapping each interceptor to the specific neighborhoods it serves, to provide pathogen occurrence data collection representative of each serviced area of the city, state, or country. If pathogen concentrations detected are higher in one interceptor than the rest in the wastewater system, the corresponding serviced area could be subject to containment measures, subject to pharmaceutical or non-pharmaceutical treatments, and nearby areas subject to the containment measures for risk of a potential viral outbreak. Detecting the pathogen or sampling wastewater in rural areas, the system determines where in the environment to sample based upon watershed modeling and microbial source tracking.
  • the AI system interprets the information obtained from the pathogen detection in the wastewater system and applies containment measures and provides feedback on the containment measures effectiveness, geographical implications of the containment measures, and epidemiology of the pathogen.
  • the AI system is divided into at least one central data processing system or one or more processing systems, linked by encrypted network connections or similar links.
  • Each processing system collects pathogen detection data from wastewater, an AI system to convert the pathogen detection data into a set of parameters or thresholds, a connection to a data network, and an interface to deploy containment measures and provide feedback on containment measures with the pathogen levels in the wastewater.
  • the AI system comprises machine learning, algorithms, and modeling methods.
  • Machine learning may be selected from the group consisting of: 1) Supervised learning; 2) Unsupervised Learning; 3) Semi-supervised learning; 4) Reinforcement learning; and 5) Deep Learning.
  • the AI system comprises supervised learning including training datasets at which data is presented in pairs, an input and its corresponding correct output.
  • One example for the data is the actual level of pathogens provided the wastewater conditions associated with each detector by which the level of pathogens is directly affected.
  • This process helps to train the algorithm and thus develop a model capable of predicting the rise of pathogens with new pathogen detectors that were not included in the training dataset.
  • unsupervised leaning comprises finding the common points between the inputs in the dataset. This process can be described as clustering of inputs that are greatly correlated under one general label based on their statistical properties.
  • Semi-supervised learning is a combination of the supervised and unsupervised learning and is an algorithm trained using a dataset that contains both labeled and unlabeled input points. This type is used mainly to enhance the performance of the algorithm through the use of both types of inputs. Reinforcement learning depends on the trial and error process to uncover the set of actions that maximizes a cumulative reward metric, which is used to make the algorithm understand whether it’s going in the right direction or not. Deep learning attempts to model the abstractions found in the dataset using a graph with multiple processing layers. These layers aim to mimic the neural network found within the human brain.
  • the AI system comprises supervised learning including using labeled data to train the learning algorithm.
  • the labeled data consists of pairs, an input that can be represented by a vector and its corresponding desired output, which is a supervisory signal.
  • the learning mechanism includes a known correct output and the learning algorithm attempts to iteratively predict this output and is corrected to reduce the variation gap between its predicted and the actual output. Analyzing the training data allows the supervised learning algorithm to produce a function that is called a classifier function if the output was discrete, and a regression function if the output was continuous.
  • the learning algorithm generalizes detected patterns and features from the training data to a new input data in a reasonable way and thus the produced function predicts the output corresponding to any provided input.
  • FIG. 2C illustrates the different stages of the supervised machine learning method, according to one embodiment.
  • Regression algorithms continuous output
  • classification algorithms discrete output
  • Regression algorithms uncover the best function that fits points in the training dataset.
  • Regression algorithms include the following main types: linear regression, multiple linear regression, lasso regression, and polynomial regression.
  • Lasso regression is a type of linear regression that uses shrinkage. Shrinkage is where data values are shrunk towards a central point, like the mean. The lasso procedure encourages simple, sparse models (i.e. models with fewer parameters).
  • Classification algorithms uncover the best fit class for the input data through assigning each input to its correct class. In this case, the output of the predictive function is in the discrete form and its value is one of the different classes available.
  • the AI system comprises unsupervised learning using an input dataset without any labeled outputs to train the learning algorithm. There is no right or wrong output to each input object and no human intervention to correct or adjust as in supervised learning. Unsupervised learning learns more about the data through identifying the fundamental structure or distribution patterns that are found in the data itself. Learning by itself, the algorithm attempts to represent a particular identified input pattern while reflecting it on the overall structure of input patterns. Thus, the different inputs are clustered into groups based on the features that were extracted from each input object.
  • FIG. 2D represents the different stages of unsupervised machine learning method, according to one embodiment.
  • the unsupervised machine learning method produces and differentiates among the result clusters and use some of them to assign new examples into other clusters. This approach is driven by input data and work well when there is adequate data available for use.
  • An example of that is filtering algorithms, to recommend containment measures to governments, which the algorithms are based on finding similar groups of pathogens detected, then adding new pathogens to these groups.
  • Algorithms included in unsupervised learning comprise three main categories, which are: clustering, dimensionality reduction and anomaly detection.
  • the AI system comprises semi-supervised learning where there is a large amount of input data, some of which are labeled and the rest are not labeled.
  • Semi-supervised requires less human intervention since it utilizes very small amount of labeled data and a large amount of unlabeled data. Utilizing less labeled datasets is more appealing since such datasets are very hard to collect as well as expensive and may require access to domain experts. Unlabeled datasets on the other hand are cheaper and easier to get access to.
  • Both supervised and unsupervised learning methods can be utilized to train the learning algorithm in semi-supervised learning.
  • Unsupervised learning methods can be used to unfold hidden structures and patterns in the input dataset.
  • supervised learning methods can be utilized to make guess predictions on the unlabeled data, feed the data back to the learning algorithm as training data and use gained knowledge to make predictions on new sets of data.
  • unlabeled data is used to modify or reprioritize prediction or hypothesis obtained from labeled data.
  • Fig. 2E illustrates the different stages of a semi-supervised machine learning method, according to one embodiment.
  • all semi- supervised learning algorithms include at least one of the following assumptions: smoothness assumption, cluster assumption and manifold assumption.
  • the AI system comprises reinforcement learning including learning by interacting with the problem environment in wastewater pathogen detection.
  • a reinforcement learning agent learns from its own actions rather than being specifically taught what to do. Reinforcement learning selects current actions based on past experiences (exploitation) and new choices (exploration). Thus, reinforcement learning is a trial and error learning process.
  • the success of an action is determined through a signal received by the reinforcement learning agent in the form of a numerical reward value.
  • the agent aims to learn to select actions that maximize the value of the numerical reward. Actions may affect not only the current situation and current reward value, but also affect successive situations and reward values.
  • the AI system comprises reinforcement learning of past pathogen detection levels in wastewater and new choices with respect to containment measures and environmental conditions.
  • Learning agents include goals set and can sense, to some extent, the state of the environment it is in and thus take actions that affect the state and bring it closer to the set goals.
  • Reinforcement learning uses direct interactions with the problem environment to gain knowledge.
  • the AI system including reinforcement learning uses direct interactions with population behavior in response to the pathogen levels in the wastewater to gain knowledge on containment measures.
  • the AI system comprises deep learning based on algorithms that learn from multiple of levels in order to provide a model that represents complex relations among data.
  • a hierarchy of features is present such that high level features are defined in terms of lower level features.
  • Deep learning models are based on unsupervised learning representations. Deep learning is the intersection point between neural networks, graphical modeling, optimization, artificial intelligence, and pattern recognition as well as signal processing.
  • Shallow structured architectures contain one or two layers at most of non-linear feature transformation. Examples of these shallow architectures include: Gaussian mixture models (GMMs), the support vector machines (SVMs) and linear or nonlinear dynamical systems.
  • GMMs Gaussian mixture models
  • SVMs support vector machines
  • the AI system comprises a Decision Tree algorithm, Support Vector Machine algorithm, and Gaussian Process Regression model to project the data and capture the possible deviation for the pathogen detection in wastewater and application of containment measures, and quantify the risk in a city, state, or country as a high risk, low risk, and moderate risk of pathogen transmission.
  • the AI system comprises deep learning based on an Artificial Neural Network (ANN).
  • ANN is an information processing system that is inspired by the way biological nervous systems work, e.g. the brain.
  • the neural network computes output values from input values by some internal calculations.
  • the neural network is trained to perform a particular function by adjusting the values of the connections (weights) between elements (based on a comparison of the output and the target) until the network output matches the target, so that the network can predict the correct outputs for a given set of inputs.
  • FIG. 2A illustrates an ANN embodiment.
  • the neural is trained to perform a pathogen transmission function by adjusting the values of the pathogen detection in the wastewater (weights) between elements based on location (based on a comparison of the output and the target) until the network output matches the target, and the neural network predicts the correct outputs of containment measures for a given set of inputs.
  • neural networks are trained to perform complex functions in various fields of the wastewater pathogen detection system, including pattern recognition of pathogen detection, identification of containment measures, classification of pathogen transmissions, and control systems due to environmental factors.
  • BPNN can approximate any continuous function and have robust non-linear mapping capabilities.
  • the neural network includes a neuron, also called “node”, as shown in Fig. 2B illustrates a single node of a neural network. Inputs are represented by a , a and a , and the output by O . There can be many input signals to a node.
  • the node manipulates these inputs to give a single output signal.
  • the values Wy, Wy, and Wnj are weight factors associated with the inputs to the node. Weights are adaptive coefficients within the network that determine the intensity of the input signal. Every input (a , a , ... , a soup) is multiplied by its corresponding weight factor (Wy, Wy, ... , Wnj), and the node uses summation of these weighted inputs (Wij * a , Wij * a , . . . , Wnj * a n ) to estimate an output signal using a transfer function.
  • the node’s output is determined using a mathematical operation on the node’s net input. This operation is called a transfer function.
  • the transfer function can transform the node’s net input in a linear or non-linear manner.
  • the AI system comprises Support Vector Machines (SVMs) including a data-based machine learning model, which is based on structural risk minimization (SRM).
  • SVMs Support Vector Machines
  • the SRM minimizes the empirical error and model complexity simultaneously.
  • SRM improves the generalization ability of the classification or regression problems.
  • SVM models comprise two types: linear support vector regression and nonlinear support vector regression.
  • the AI system comprises deep neural networks (DNN) including feed forward neural networks.
  • DNN deep neural networks
  • Back-propagation (BP) is an algorithm used for learning the parameters of these networks.
  • BP does not work well for learning networks that contain more than a small number of hidden layers.
  • Optimization with the deep models is alleviated when an unsupervised learning algorithm is introduced.
  • Deep belief networks (DBN) consists of a stack of restricted Boltzmann machines (RBMs).
  • RBMs restricted Boltzmann machines
  • the AI system comprises deep learning, including synthesis or recognition, generation or classification, including three different classes of networks consisting of: deep network intended for unsupervised (generative) learning; deep networks for supervised learning; and hybrid deep networks.
  • Deep networks intended for unsupervised (generative) learning capture high order correlation of the visible data for the purpose of synthesis or pattern analysis given that no information is available about the target class labels.
  • Deep networks for supervised learning directly provide the discriminative power for the purpose of pattern classification. Labeled data are always present in the direct or indirect form for supervised learning.
  • the hybrid deep networks perform discriminations which are often assisted with the outcomes of generative or unsupervised deep networks and by better optimization of the deep networks in deep networks for supervised learning.
  • the AI system comprises a Convolutional Neural Network, including a discriminative deep architecture in which every model contains a convolutional layer and a pooling layer and are stacked on top of each other. Many weights are shared in the convolutional layer, the pooling layer on the other hand sub-samples the output coming from the convolutional layer and decreases the data rate of the below layer. The weight sharing together with properly chosen pooling schemes, results in invariance properties of the CNN.
  • the drive for using the convolution operator in such applications which is considered a specialized linear operator, is selected from: sparse interactions, parameter sharing, and equivariant representation.
  • the AI system comprises Recurrent Neural Networks (RNNs) for the use in unsupervised learning in the cases where the depth of the input data sequence can be as large as the length, since RNNs allow parameter sharing through the different layers of the network.
  • RNNs use of the same set of weights in a recursive manner over a tree like structure, and the tree is traversed in topological order.
  • the RNN is used for predicting the future data sequence through the use of previous data samples.
  • the RNN predicts sequence data for PCR methods of pathogen detection in wastewater.
  • ANN For the ANN embodiment, measurements of the pathogen level are collected over a period of time and over a geographical location to cover all probable seasonal variations in the wastewater variables. The input and output data is analyzed statistically by means of one-way analysis of variance ANOVA1 function in MATLAB software. Moreover, ANOVA1 function is applied before ANN in order to reject the inaccurate measured raw data. ANN can predict the wastewater system pathogen detection performance with correlation coefficient (R) between the observed and predicted output variables reached up to 0.90. Moreover, ANN provides an effective analyzing and diagnosing tool to understand and simulate the non-linear behavior of the pathogen detection in wastewater, and is used as a valuable performance assessment tool for decision makers in healthcare policy.
  • the AI system comprises a supervised machine learning method comprising a Lasso regression for predicting the pathogen levels rising in the wastewater and model selection for the epidemiology of the pathogen, and providing inferential statistics to provide for containment measures.
  • a Lasso regression for predicting the pathogen levels rising in the wastewater and model selection for the epidemiology of the pathogen, and providing inferential statistics to provide for containment measures.
  • three separate selection models are available in StataMP v.16, including Cross-validation (CV), Adaptive lasso and minimum Bayesian Information Criterion (BIC).
  • CV Cross-validation
  • BIC minimum Bayesian Information Criterion
  • the dataset may be split into two sample groups; Group 1 are training datasets to select the model, and Group 2 are testing datasets test the prediction. Lasso includes a greater prediction accuracy and increases model interpretability.
  • the Cross-fit partialing out Lasso technique to estimate the coefficients, robust standard errors, p-values and confidence intervals of the specified variables of interest while the other covari
  • the AI system comprises unsupervised machine learning algorithms (k-means) to define data-driven clusters of geography; the algorithm is informed by disease prevalence estimates, environmental metrics, socio-economic status and health system coverage.
  • k-means unsupervised machine learning algorithms
  • pathogen clusters are compared by number of COVID-19 clinical cases, number of deaths, case fatality rate and order in which a geography reported the first case.
  • Different data sources to build a dataset with information on COVID-19 (Table 1).
  • the variables used to develop the clustering model include different values between them, thus each of them carries a different variance. Because of this characteristic, it is relevant to standardize these variables to set reliable clusters without losing information. Consequently, before running the unsupervised clustering algorithms, the predictors were treated with an orthogonal transformation and then with principal component analysis (PCA).
  • PCA principal component analysis
  • PCA is a method of unsupervised machine learning algorithms. PCA follows an orthogonal transformation, which turns correlated variables into an uncorrelated set of variables. The PCA creates a set of characteristics, or components, that represents the relevant information from the original group of variables. The PCA seeks to reduce the number of predictors while maximizing the variance.
  • K-means seeks to group heterogeneous elements into homogenous clusters. This method is unsupervised machine learning, because it assigns the elements into clusters which were unknown at the beginning of the analysis.
  • the centroid-based algorithm: k-means works well when the clusters have similar size, similar densities and follow a globular shape.
  • Convolutional neural networks To train a CNN from scratch to differentiate between regions with low and high pathogen levels in wastewater prevalence, data may be selected from Table 1. If training data is unavailable, a transfer learning approach (as shown in Fig. 3A), which involves applying a previously trained network to a dataset of detected pathogens in wastewater systems to make inferences about containment measures for wastewater systems where the pathogen is still undetected. A network, which has been previously trained (hereafter referred to as, pre-trained) for pathogen recognition on the challenge dataset to an adjacent wastewater system or geography should be able to identify the correct class for every detection within its top 5 predictions, or predictions with highest probability.
  • Fig. 3A shows the model trained for domain A has learnt how to interpret containment measures. Knowledge acquired by the model is to understand containment measures effectiveness. When the containment measures are fed as input to the model, a feature map encodes this knowledge. These feature maps are used in a regression model to predict pathogen values which is represented as domain B.
  • the convolutional neural network consists of five convolutional layers and three fully connected layers.
  • Each convolutional layer is composed of several two-dimensional filters which activate the features required for classifying an object correctly.
  • the neural network learns to extract gradients, edges and patterns that aid in accurate pathogen detection.
  • the fully connected layers further process these features and convert them into single dimensional vectors.
  • the output layer (final fully connected layer) is originally designed for classifying between, i.e. 1000 object categories.
  • this neural network transforms a large two-dimensional data into a single vector of fixed dimension, containing only the most important descriptors of the image. This feature vector is extracted by deploying the network and making a forward pass through it for each data point.
  • a forward pass through the network is made for some data and examines the output maps from convolutional layers of the CNN, to see if built environment features are being highlighted by these filters and transmitted to the succeeding layers.
  • the output maps are single channel images which can be plotted and compared to the original data point for interpretation of the outputs. The results from this process are shown in Fig. 3B.
  • the AI system comprises an adaptive neuro-fuzzy inference system (ANFIS) including an artificial neural network integrating neural networks and fuzzy logic.
  • ANFIS includes nonlinear functions.
  • Fig. 3C shows the architecture of the ANFIS model.
  • ANFIS comprises five main layers.
  • the first layer is the inputs layer which takes the parameters and imports them to the model. This layer is also called as the input layer of the fuzzy system.
  • the outputs of the first layer imports to the second layer and carries prior values of Membership Functions (MFs). Fuzzy rules are concluded from the nodes on second layer related degree of activity.
  • the third layer normalizes the degree of activity of any rules.
  • the fourth layer adopts the nodes and function and produces the outputs and send them to the output layer.
  • the important factors for determining the accuracy of ANFIS is the number and type of MFs, the optimum method and the output MF type.
  • ANFIS model is developed by an ANFIS toolbox on MATLAB. Input parameters are independent variables of each scenario and the output variable was the number of cases or mortality rate.
  • the ANFIS model is trained with three triangular, trapoizidal and gaussian MFs. This step was performed in order to select the best MF.
  • the output membership function type selected was linear type in one embodiment, because of its ability to further reduce of errors. Training of FIS may be done with a backpropagation optimum method, and 0 value of error tolerance.
  • Multi-layered-Perceptron is the frequently used ANN method for prediction and modeling purposes. This technique is a single method to provide an acceptable accuracy for prediction tasks in simple and semi-complex dataset. But, in case of doing modeling tasks in complex dataset, there is a need for more robust techniques. For this reason, hybrid methods have been used more frequently. Hybrid methods contain a predictor and one or more optimizer. The present study develops a hybrid MLP-ICA method as a robust hybrid algorithm for developing a platform for predicting the COVID-19 cases and rate of transmission. The ICA is a method in the field of evolutionary calculations that seeks to find the optimal answer to various optimization problems. This algorithm, by mathematical modeling, provides a socio-political evolutionary algorithm for solving mathematical optimization problems.
  • the ICA constitutes a primary set of possible answers. These answers are known as countries in the ICA.
  • the ICA gradually improves the initial responses (countries) and ultimately provides the appropriate answer to the optimization problem.
  • Each of these algorithms are based on daily positive new cases, population, population density, deceased cases, availability of hospitals, no. of tests conducted, recovered cases, and availability of essential delivery services.
  • the cases considered positive define the intensity of COVID-19 in a wastewater-defined region.
  • Population data reflects possible asymptomatic carriers.
  • Population density impacts the risk of spread/community transmission.
  • the number of deceased and recovered cases in the region is impacted by the quality of health care facilities.
  • the availability of hospitals defines the reserve capacity to handle this detrimental situation.
  • the risk associated with a region can be understood by considering outputs from each of these algorithms or others that may be developed.
  • the availability of daily-needs-to-the- doorsteps may be confined to only a few of the areas. These services help to minimize mass gatherings, and mitigate the crowding at retail units. Frequent data corresponding to each of these variables is collected from various sources and is compiled in a data repository.
  • Different sources of the datasets to train the AI system may be based upon open datasets SARS-CoV-2 provided by a variety of sources, such as the Johns Hopkins University, the World Health Organization, the Chinese Center for Disease Control and Prevention, the National Health Commission, and the US Centers for Disease Control and Prevention (CDC). Additional data from socio-demographic sources and policies and regulations implemented by countries may be derived from resources of the World Bank, UNICEF, WHO, and CDC. Table 1 lists the datasets according to their appearances on the respective websites.
  • the AI system may incorporate epidemiology models to track the location and spread of the pathogen and provide containment measures for the same.
  • Susceptible-infectious-susceptible SIS
  • susceptible-infected-recovered-deceased-model SIRD
  • Maternally derived immunity- Susceptible-Infectious-Recovered MSIR
  • Susceptible-exposed-infectious-recovered SEIR
  • SEIS Susceptible-exposed-infectious-susceptible
  • SEIR Susceptible-exposed-infectious-susceptible
  • SEIS Susceptible-exposed-infectious-susceptible
  • MSEIR Maternally derived immunity-Susceptible- Exposed-Infectious-Recovered-Susceptible
  • MSEIRS Maternally derived immunity-Susceptible- Exposed-Infectious-Recovered-Susceptible
  • MSEIRS Maternally derived immunity-Susceptible- Exposed-Infectious-Recovered-S
  • SEIR models have been reported among the most popular tools to predict the outbreak. SEIR models involve the significant incubation period of an infected person, and are reported to present relatively more accurate predictions. In the case of Varicella and Zika outbreaks the SEIR models showed increased model accuracy. SEIR models assume that the incubation period is a random variable and similarly to the
  • a wide variety of pathogenic organisms pass through municipal waste-water treatment systems. Any type of infection within a community is likely to lead to pathogen excretion in bodily fluids/sub stances and therefore, transportation into the community sewage system. Infections may be classified into symptomatic infections and asymptomatic infections. Symptomatic infections may result in death, severe illness, moderate severity, and mild illness-all of which are clinical diseases. Asymptomatic infections may be infection without clinical illness and exposure, including colonization showing no illness.
  • Table 2 classifies pathogen in categories Category A pathogens require the most intensive public preparedness efforts due to the potential for mass causalities, public fear, and civil disruption.
  • Category B pathogens are also moderately easy to spread, but have lower mortality rates.
  • Category C pathogens do not present a high public health threat, but could emerge as future threats
  • Table 2 The center for disease control select agents
  • Agents causing enteric and respiratory infections are released in large numbers in feces and respiratory secretions. Many of the enteric viruses such as the enteroviruses and adenoviruses may replicate both in the intestinal and respiratory tract. The number of enteric viruses detected can approach peak concentrations of 10 12 organisms per gram of stool while protozoa can approach 10 6 -10 7 per gram. Cultivatable enteric bacterial pathogens such as Salmonella may also occur in concentrations as large as 10 11 per gram. The concentration of respiratory viruses ranges from 10 5 to 10 7 per ml of respiratory secretion.
  • Blood-borne viruses such as HIV will be found in the feces of infected persons and many viruses will occur in the urine during infection of the host, although these excreted viruses may not be infectious.
  • the total amount of virus released by a person is, of course, also related to the amount of feces, urine, respiratory secretion, and skin that is released by the person. On average, a person excretes between 100 to 400 g of feces and 700- 2000 ml of urine per day [0111] Examples
  • system may refer to a computer-related entity, either hardware, a combination of hardware and software, software, or software in execution.
  • a system can be, but is not limited to being, a process running on a processor, a processor, an object, an executable, a thread of execution, a program, and/or a computer.
  • an application running on a server and the server can be a component.
  • One or more components can reside within a process and/or thread of execution, and a component can be localized on one computer and/or distributed between two or more computers.
  • program modules include routines, programs, components, data structures, etc., that perform particular tasks or implement particular abstract data types.
  • inventive methods can be practiced with other computer system configurations, including single-processor or multiprocessor computer systems, minicomputers, mainframe computers, as well as personal computers, hand-held computing devices, microprocessor-based or programmable consumer electronics, and the like, each of which can be operatively coupled to one or more associated devices.
  • the illustrated aspects of the innovation may also be practiced in distributed computing environments where certain tasks are performed by remote processing devices that are linked through a communications network.
  • program modules can be located in both local and remote memory storage devices.
  • a computer typically includes a variety of computer-readable media.
  • Computer-readable media can be any available media that can be accessed by the computer and includes both volatile and nonvolatile media, removable and non-removable media.
  • Computer-readable media can comprise computer storage media and communication media.
  • Computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer- readable instructions, data structures, program modules or other data.
  • Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by the computer.
  • Communication media typically embodies computer-readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism, and includes any information delivery media.
  • modulated data signal means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal.
  • communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared and other wireless media. Combinations of the any of the above should also be included within the scope of computer-readable media.
  • Software includes applications and algorithms.
  • Software may be implemented in a smart phone, tablet, or personal computer, in the cloud, on a wearable device, or other computing or processing device.
  • Software may include logs, journals, tables, games, recordings, communications, SMS messages, Web sites, charts, interactive tools, social networks, VOIP (Voice Over Internet Protocol), e-mails, and videos.
  • VOIP Voice Over Internet Protocol
  • ⁇ includes any type of computer code, including source code, object code, executable code, firmware, software, etc.
  • computer readable medium includes any type of medium capable of being accessed by a computer, such as read only memory (ROM), random access memory (RAM), a hard disk drive, a compact disc (CD), a digital video disc (DVD), or any other type of memory.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Medical Informatics (AREA)
  • Data Mining & Analysis (AREA)
  • Biomedical Technology (AREA)
  • Databases & Information Systems (AREA)
  • Pathology (AREA)
  • Epidemiology (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Abstract

L'invention concerne des systèmes et des procédés de surveillance d'eaux usées, de détection d'un agent pathogène et de contrôle de l'apparition et de la propagation de maladies associées à l'agent pathogène.
PCT/US2021/034837 2020-06-02 2021-05-28 Système d'ia et de données pour surveiller des agents pathogènes dans des eaux usées et procédés d'utilisation WO2021247408A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US202063033627P 2020-06-02 2020-06-02
US63/033,627 2020-06-02

Publications (1)

Publication Number Publication Date
WO2021247408A1 true WO2021247408A1 (fr) 2021-12-09

Family

ID=78829866

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2021/034837 WO2021247408A1 (fr) 2020-06-02 2021-05-28 Système d'ia et de données pour surveiller des agents pathogènes dans des eaux usées et procédés d'utilisation

Country Status (1)

Country Link
WO (1) WO2021247408A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114509939A (zh) * 2022-01-12 2022-05-17 江南大学 一种i-moead算法的污水处理过程优化控制方法
CN115099321A (zh) * 2022-06-17 2022-09-23 杭州电子科技大学 双向自回归无监督预训练微调式排污异常监控方法及应用

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080030730A1 (en) * 2006-08-03 2008-02-07 The United States Of America As Represented By The United States Environmental Protection Agency Water contamination measurement apparatus
JP2009000673A (ja) * 2007-05-24 2009-01-08 Metawater Co Ltd 浄水プロセスの監視装置及び監視方法
US20100175006A1 (en) * 2008-08-28 2010-07-08 Georgetown University System and method for detecting, collecting, analyzing, and communicating event-related information
US20190251453A1 (en) * 2018-02-15 2019-08-15 Wipro Limited Method and system for determining occurrence of sewer flooding in a geographical area
US20190347921A1 (en) * 2012-05-24 2019-11-14 Douglas Lundy THREAT DETECTION SYSTEM HAVING CLOUD or LOCAL MONITORING UNIT FOR COMMUNICATING WITH INTERNET ADDRESSABLE WIRELESS DETECTOR UNITS, ASSOCIATED WIRELESS SENSOR AND OPTIONAL WIRELESS SUB-SENSOR DEVICES, THAT BLANKET A VENUE USING A BROAD RANGE OF NETWORK ARRANGEMENTS AND REPORTING NON-COMPLIANT SENSOR VALUES OR CONDITIONS, DATA RATE OF CHANGE AND/OR DEVICE LOCATION AND DEVICE NON-RESPONSE, AND DELIVERING NOTIFICATIONS, WHILE RELATED DEVICES ARE ENABLED, AS AND WHEN REQUIRED.

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080030730A1 (en) * 2006-08-03 2008-02-07 The United States Of America As Represented By The United States Environmental Protection Agency Water contamination measurement apparatus
JP2009000673A (ja) * 2007-05-24 2009-01-08 Metawater Co Ltd 浄水プロセスの監視装置及び監視方法
US20100175006A1 (en) * 2008-08-28 2010-07-08 Georgetown University System and method for detecting, collecting, analyzing, and communicating event-related information
US20190347921A1 (en) * 2012-05-24 2019-11-14 Douglas Lundy THREAT DETECTION SYSTEM HAVING CLOUD or LOCAL MONITORING UNIT FOR COMMUNICATING WITH INTERNET ADDRESSABLE WIRELESS DETECTOR UNITS, ASSOCIATED WIRELESS SENSOR AND OPTIONAL WIRELESS SUB-SENSOR DEVICES, THAT BLANKET A VENUE USING A BROAD RANGE OF NETWORK ARRANGEMENTS AND REPORTING NON-COMPLIANT SENSOR VALUES OR CONDITIONS, DATA RATE OF CHANGE AND/OR DEVICE LOCATION AND DEVICE NON-RESPONSE, AND DELIVERING NOTIFICATIONS, WHILE RELATED DEVICES ARE ENABLED, AS AND WHEN REQUIRED.
US20190251453A1 (en) * 2018-02-15 2019-08-15 Wipro Limited Method and system for determining occurrence of sewer flooding in a geographical area

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
FRANCESCO GRANATA, ET AL.: "Machine Learning Algorithms for the Forecasting of Wastewater Quality Indicators", WATER, vol. 9, no. 105, 1 January 2017 (2017-01-01), pages 1 - 13, XP055659378, DOI: 10.3390/w9020105 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114509939A (zh) * 2022-01-12 2022-05-17 江南大学 一种i-moead算法的污水处理过程优化控制方法
CN114509939B (zh) * 2022-01-12 2023-04-07 江南大学 一种i-moead算法的污水处理过程优化控制方法
CN115099321A (zh) * 2022-06-17 2022-09-23 杭州电子科技大学 双向自回归无监督预训练微调式排污异常监控方法及应用
CN115099321B (zh) * 2022-06-17 2023-08-04 杭州电子科技大学 双向自回归无监督预训练微调式排污异常监控方法及应用

Similar Documents

Publication Publication Date Title
Shaban et al. A new COVID-19 Patients Detection Strategy (CPDS) based on hybrid feature selection and enhanced KNN classifier
Kogilavani et al. COVID-19 detection based on lung CT scan using deep learning techniques
Fotovatikhah et al. Survey of computational intelligence as basis to big flood management: Challenges, research directions and future work
Goethals et al. Applications of artificial neural networks predicting macroinvertebrates in freshwaters
Nkiruka et al. Prediction of malaria incidence using climate variability and machine learning
Garain et al. Detection of COVID-19 from CT scan images: A spiking neural network-based approach
Jahan et al. Automated invasive cervical cancer disease detection at early stage through suitable machine learning model
Madhiarasan et al. Analysis of artificial neural network: architecture, types, and forecasting applications
WO2021247408A1 (fr) Système d'ia et de données pour surveiller des agents pathogènes dans des eaux usées et procédés d'utilisation
Tseng et al. Computational intelligence techniques for combating COVID-19: A survey
Motamarri et al. Development of a neural-based forecasting tool to classify recreational water quality using fecal indicator organisms
Srinivas et al. Survey on prediction of heart morbidity using data mining techniques
Paul et al. Combating Covid-19 using machine learning and deep learning: Applications, challenges, and future perspectives
Weller et al. Predictive models may complement or provide an alternative to existing strategies for assessing the enteric pathogen contamination status of northeastern streams used to provide water for produce production
Pedrycz et al. The design of granular classifiers: a study in the synergy of interval calculus and fuzzy sets in pattern recognition
Darapaneni et al. Pneumonia detection in Chest X-rays using neural networks
Bhoi et al. Estimating the water quality class of a major irrigation canal in Odisha, India: a supervised machine learning approach
Murugan et al. Impact of Internet of Health Things (IoHT) on COVID-19 disease detection and its treatment using single hidden layer feed forward neural networks (SIFN)
Islam et al. Machine learning to promote health management through lifestyle changes for hypertension patients
Kok et al. Recent trends in air quality prediction: An artificial intelligence perspective
Pontin et al. Determining factors that influence the dispersal of a pelagic species: A comparison between artificial neural networks and evolutionary algorithms
El-Attar et al. Performance of artificial intelligence models in analysis and prediction of water potability
Leke-Betechuoh et al. Prediction of HIV status from demographic data using neural networks
GurjotKour et al. Hybrid voting classifier model for COVID-19 prediction by embedding machine learning techniques
Senturk Amphibian species detection in water reservoirs using artificial neural networks for ecology-friendly city planning

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21817543

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21817543

Country of ref document: EP

Kind code of ref document: A1