ANTICIPATING DIABETIC FOOT ULCER USING GENERATIVE FUZZY EXPERT SYSTEM FRAMEWORK

Transcription

1 International Journal of Civil Engineering and Technology (IJCIET) Volume 8, Issue 12, December 2017, pp , Article ID: IJCIET_08_12_070 Available online at ISSN Print: and ISSN Online: IAEME Publication Scopus Indexed ANTICIPATING DIABETIC FOOT ULCER USING GENERATIVE FUZZY EXPERT SYSTEM FRAMEWORK J. Jayashree, J. Vijayashree School of Computer Science and Engineering, VIT, Vellore, India ABSTRACT A standout amongst the most genuine complexities of diabetes is Diabetic foot. The frequency of diabetic foot increments similarly on an ascent of diabetic's increments and it is regularly disregarded by individuals prompting evacuation of foot. A Diabetic foot expert system which constructs large scale knowledge framework is proposed using fuzzy. Fuzzification is used to build knowledge. Key words: Diabetic Mellitus, Neuropathy, Foot Ulcer, Fuzzy, Wagner. Cite this Article: J. Jayashree, J. Vijayashree. Anticipating Diabetic Foot Ulcer Using Generative Fuzzy Expert System Framework. International Journal of Civil Engineering and Technology, 8(12), 2017, pp INTRODUCTION Diabetic mellitus is one of the oldest diseases from 1500 BC which was described by both the physicians of India and Egypt in their manuscript. In AC physicians of India have discovered the classifications of diabetics as type1 and type 2. The chronic disease which is found in the young as type 1 and the disease found in overweight as type 2. Body cell attacks the pancreas beta cells which results in zero production of insulin which does not have any cure will come under type 1 diabetics. If a human body cannot make use of the insulin being produced in the pancreas will come under the type 2 diabetics. Gastric surgery medication is the cure for type 2 diabetics million People were affected by diabetics in India in 2015 and which is expected to reach a328.4 million people in Due the modern life style of people diabetics statistics shows the diabetic population in more in urban. In 2015, million people have diabetics in china and it is expected to reach million in 2040.india has the second largest diabetic population with 69.2 million in 2015 and expected to reach a unbelievable population of million diabetic people in United states spend a good amount of funds in diabetic related centers buy still the fall under the third country with largest population being affected by diabetics with 29.3 million in 2015 and expected 35.1 million in editor@iaeme.com

2 Anticipating Diabetic Foot Ulcer Using Generative Fuzzy Expert System Framework 2. NEUROPATHY Over a period, diabetics will affect the organs of human and based on the organs being affected, they are classified into four types of neuropathy and they are 1. Peripheral Neuropathy 2. Focal Neuropathy 3. Proximal Neuropathy 4. Autonomic Neuropathy Peripheral Neuropathy Organs Affected 1. Toe 2. Feet 3. Legs 4. Hands 5. Arms Autonomic Neuropathy Organs Affected 1. Heart & Blood vessel 2. Urinary track 3. Sex organ 4. Sweat gland 5. Lungs Focal Neuropathy Organs Affected 1. Eyes 2. Ears 3. Chest 4. Abdomen 5. Facial muscle 6. Pelvis Proximal Neuropathy Organs Affected 1. Hip 2. Buttocks 3. Thighs Figure 1 Types of Neuropathy 2.1. Diabetic Neuropathy Complications Diabetic Neuropathy has some of the following complications like 1. Loss of Limbs Lower: leg or toe gets amputated because of infection caused due to the unnoticed ulcer.2. Charcot Joint: Swelling and insensitivity due to damage of nerve makes the joints to deteriorate. 3. Ulcer tract infections and urinary incontinence: Urinary bladder controlling nerve gets damage leading to infection. 4. Hypoglycaemia unawareness: High heart beat due to sugar drop. 5. Low blood pressure: Fainting due to pressure drop caused by nerve damage of blood circulation. 6. Digestive problem: 7. Sexual Dysfunction: Men (erectile) and women (lubrication) problem 8. Increase or decrease sweating: Temperature regulation due to damage of sweat gland editor@iaeme.com

3 J. Jayashree, J. Vijayashree 2.2. Human Foot and Diabetics Blood vessel damage and nerve damage is caused by both T1 and T2 diabetes which leads to feet and leg problem. Diabetic people foot problem is because of main two conditions; Peripheral Vascular Diabetics and Peripheral Neuropathy. Feet nerves and leg nerves sensation will get reduced which makes a diabetic patient could not even recognize the injury in legs and feet. Few areas of feet get more pressure due to the changes in the alignment of feet. Due to the reduction of oxygen flow to the feet and reduction of sensation of feet ulcer is formed. Diabetic Neuropathy test are given below in which Monofilament test is very common which checks the sensitivity of the diabetic foot. Table 1 Diabetic Neuropathy Test and Diagnosis Diabetic Neuropathy Test and Diagnosis Test Description Diagram Test Description Diagram Filament Test Nylon fiber (monofilament) sensitivity test Electromyograph Muscles y (EMG) electrical discharge is tested Nerve conduction studies Diagnose (carpal tunnel syndrome) Electrical signal test on nerves in arm and legs Quantitative sensory test Temperature and vibration test -. To test nerve response Foot infection can be localized and generalized. Localized Foot infection is an area of foot swells and become reddish due to infection and antibiotic can be applied as medicine. These infected areas can even get ulcer formation. In Generalized Foot infection whole foot become swelled and reddish due to infection and which may even spread to tissue which can t be cured by antibiotic editor@iaeme.com

4 Anticipating Diabetic Foot Ulcer Using Generative Fuzzy Expert System Framework 1. Poor circulation 2. Hyperglycaemia 3. Nerve damage 4. Wounded feet 1. Poorly fitted shoe 2. Improper toe trimming 3. Alcohol usage 4. Tobacco usage 5. Obesity Causes Risk Factor 1. Silver sulphadiazine cream 2. Polyhexamethylene biguanide gel 3. Iodine (povidone or cadexomer) 4. Medical grade honey in gel Treatm ent Prevent ion 1. Foot baths 2. Disinfecting ulcer skin 3. Keeping the ulcer dry with frequent dressing 4. Enzyme treatments 5. Calcium alginates dressing Figure 2 Foot Ulcer 3. FUZZY INFERENCE SYSTEM Neural Networks are great in speculation and taking in a mapping from information set of qualities (components) to yield esteem. Take it as a discovery, you give it a preparation set containing various columns of input - > output, and the container figures a general govern with these preparation tests. Presently, when you give qualities to an indistinguishable component from information, the discovery advises the yield as indicated by the standards it has learnt. Fuzzy Logic is utilized to model instabilities as observed in numerous genuine situations. It is a superset of Boolean rationale which has strict boundaries between various classes. Utilizing fuzzy if-then standards to express input output connections and model the subjective sources of information and thinking process for making the yield. The fuzzy inference systems consolidate an arrangement of predecessor and subsequent fuzzy participation works and in addition an arrangement of Fuzzy IF THEN rules which considered a firm reason for building up the center of any system which may be utilized for settling on choices in obscure and incorrect circumstances. A fuzzy inference system, truncated as FIS, copy human insight for evaluating the equivocalness/imprecision in editor@iaeme.com

5 J. Jayashree, J. Vijayashree certifiable situation through legitimate displaying of the fluffiness of genuine information by suitable lead bases. To process the yield of this FIS given the sources of information, six steps must be followed: 1. Deciding an arrangement of fuzzy standards 2. Fuzzifying the information sources utilizing the information enrolment capacities 3. Consolidating the fuzzified inputs as per the fuzzy principles to build up a manage quality (Fuzzy Operations) 4. Finding the result of the run the show by joining the administer quality and the yield participation work (suggestion) 5. Joining the results to get a yield conveyance (collection) 6. Defuzzifying the yield dissemination (this progression is just if a fresh yield (class) is required). This paper proposes a model for anticipating the risk DFU (Diabetic Foot Ulcer) utilizing independent agents for various information parameters and a blend of fuzzy expert systems. The model has fundamentally two sections. The initial segment comprises of agents who separately work on information parameters. The second part, Fuzzy system comprises of multi-layered design comprising of three FISs (Fuzzy Inference Systems) in the main layer. The second layer has yield from the primary layer and this Fuzzy inference system utilizes information base to give the yield. Essentially, the model should take input as qualities comprising of variables like age, duration of diabetes, Systolic blood pressure, and diastolic blood pressure. These information parameters are ordered in view of the watchwords the agent sends the information qualities to legitimate fuzzy inference system for calibration. This agent is known as symptom classification agent Symptom Classifier In light of calculations examined in review done by Aggarwal and classifiers talked about by Rajeswari, a classifier was executed to choose vital traits from an extensive arrangement of characteristics. By utilizing data mining methods and reducing the characteristics, the accuracy of classifier can be improved significantly. The aforementioned classifier was changed to separate information traits in three unique classes, to be specific: foot related, body related and demographic related. This classifier then sends to the individual agent Normalizing Layer The system in this manner has three agents, in particular, foot related agent, body composition agent and statistic based agent. Foot related agent is concerned with elements like state of foot, callus of foot, and so on. Second agent, body composition agent deals with the parameters like blood pressure, duration of diabetes, age, co-morbidity, and so on and the third agent works at attitude of patient (whether the patient is positive while taking the drugs and treatments), learning of diabetes, and so on. The agents have the undertaking of changing over the qualities to fuzzy value from crisp and furthermore calibrating as indicated by the fuzzy inference systems which will work on them. For instance, duration of diabetes is changed over into information which fluctuates from zero to one. The qualities from these agents are transferred to separate fuzzy expert systems, which shapes the main layer editor@iaeme.com

6 Anticipating Diabetic Foot Ulcer Using Generative Fuzzy Expert System Framework 3.3. Fuzzy System The fuzzy inference systems in the main layer are named (Fuzzy Inference System one) FIS1, FIS2 and FIS3. Each of these systems has its own particular rule base, which is utilized to give an intermediate yield. The rules are created from a knowledge base taken from domain experts. This yield is utilized as input for another fuzzy inference system which is utilized to give the yield of risk of DFU. This structures the second layer of the fuzzy system Inference Systems In light of the rules and input, the risk of DFU is assessed. These rules may have diverse weights and priorities. For instance, the weight of property callus of foot is higher than property age, on the grounds that the hazard is significantly higher for a man who has callus exhibit than a man who has age more than 45. Mamdani Inference system was utilized for its intuitiveness and across the board acknowledgement. Centroid method is utilized for de-fuzzification. MATLAB FIS editor is utilized to choose the different parameters. "AND method utilized is minimum (min), "OR" method utilized is maximum(max), implication method utilized is minimum(min) and aggregation method utilized is maximum(max). 4. PROPOSED FRAMEWORK Figure 3 Proposed Framework 4.1. Fuzzy Inference Process Fuzzify Inputs The first and foremost steps includes converting the crisp input values into fuzzy values. The inputs must be fuzzified according to linguistic sets. For example, let us take ratings of a car in terms of looks. This rating is crisp value. The linguistic sets can be good-looking, mediocre and simple. This step maps the crisp value to one of the value from the linguistic set Apply fuzzy Operator After the inputs are fuzzified, the degree of satisfied antecedent is known for each rule. If there are more than one parts, fuzzy operator is applieed to obtain one number that represents the result of that rule for a particular antecedent. The inputs to a fuzzy operator can be many but the output is a single truth value. For example, the rating for looks of car and ratings for driver satisfaction are 3 and 8. After first step let us assume that we get membership values as editor@iaeme.com

7 J. Jayashree, J. Vijayashree 0.1 and 0.7. Considering the system has a rule with OR, and the system applies OR operator MAX, the operator simple selects Apply Implication Method The weights of the rules should be determined before this step. This weight determines the importance of the rule in finding out the output. Generally, the weight is 1, but may go down to 0 in case of trivial rules. The input of implication method is a single number given by antecedent. A consequent is a fuzzy set that is represented by a membership function, which weights appropriately the linguistic characteristics that are attributed to it. The methods used are min (Minimum) and prod (product) to calculate the output fuzzy set. Implication is implemented for each rule Aggregate all outputs Since, each rule will generate an output, all these should be combined to make a single decision. Aggregation is the process which combines the fuzzy sets that represent the outputs of each rules into a single fuzzy set. Generally methods for aggregation include, max (Maximum), probor (Probablistic OR) and sum Defuzzify Finally, the fuzzy set is converted into a single number in this step. Several methods can used to find this single value. Some of them are centroid, bisector, middle of maximum, largest of maximum and smallest of the maximum Diabetic Foot Classifications The well-established widely used Wagner wound classification system, the new University of Texas (UT) diabetic wound classification system PEDIS classification. All the wound classification systems are easy to use among health care providers, and both can provide a guide to planning treatment strategies. The Wagner system assesses ulcer depth and the presence of osteomyelitis or gangrene by using the following grades: grade 0 (pre- or post ulcerative lesion), grade 1 (partial/ full thickness ulcer), grade 2 (probing to tendon or capsule), grade 3 (deep with osteitis), grade 4 (partial foot gangrene), and grade 5 (whole foot gangrene). The UT system assesses ulcer depth, the presence of wound infection, and the presence of clinical signs of lower-extremity ischemia. This system uses a matrix of grade on the horizontal axis and stage on the vertical axis. The grades of the UT system are as follows: grade 0 (pre- or post ulcerative site that has healed), grade 1 (superficial wound not involving tendon, capsule, or bone), grade 2 (wound penetrating to tendon or capsule) and grade 3 (wound penetrating bone or joint). Within each wound grade there are four stages: clean wounds (stage A), non ischemic infected wounds (stage B), ischemic non infected wounds (stage C), and ischemic infected wounds (stage D). Perfusion, extent, depth, infection and sensation are the components of interest in the PEDIS classification Treatment If the wagner grade >= 3, then the risk of DFU- severe and if the grades are 1 and 2, then the risk of DFU is moderate and grade 0 illustrates the risk of DFU as mild. From the study [18], Ceftobiprole/ Vincomycin and Clinafloxacin / Piperacillin - tazobactam are the treatment/ control used for severe foot ulcer with a clinical cure rate of 86.2/81.8 and 51.7/48. Moxifloxacin / Piperacillin - tazobactam and Daptomycin / Vincomycin are the treatment/ control used for Moderate foot ulcer with a clinical cure rate of 67.6/61and 66/70. Pexiganan topical / Ofloxacan and Piperacillin - tazobactam / Ampicillin - Sulbactam are the treatment/ control used for Mild foot ulcer with a clinical cure rate of 86.8/90.4 and 81.3/ editor@iaeme.com

8 Anticipating Diabetic Foot Ulcer Using Generative Fuzzy Expert System Framework 5. CONCLUSIONS This paper proposes a model for predicting the risk DFU (Diabetic Foot Ulcer) using independent agents for different input parameters and a combination of fuzzy expert systems. To describe the severity of diabetic foot ulcer, the proposed system uses [17] Wagner foot classification system. REFERENCES [1] Rajeswari, K., V. Vaithiyanathan, and Shailaja V. Pede. "Feature selection for classification in medical data mining." International Journal of Emerging Trends and Technology in Computer Science (IJETTCS) 2.2 (2013): [2] Montgomery, Alan A., et al. "Evaluation of computer based clinical decision support system and risk chart for management of hypertension in primary care: randomised controlled trial." Bmj (2000): [3] Martins-Mendes, Daniela, et al. "The independent contribution of diabetic foot ulcer on lower extremity amputation and mortality risk." Journal of Diabetes and its Complications 28.5 (2014): [4] Cowley, Matthew S., et al. "Foot ulcer risk and location in relation to prospective clinical assessment of foot shape and mobility among persons with diabetes." Diabetes research and clinical practice 82.2 (2008): [5] Acharya, S., et al. "Conservative management of diabetic foot osteomyelitis."diabetes research and clinical practice (2013): e18-e20. [6] Moura, Liane IF, et al. "Recent advances on the development of wound dressings for diabetic foot ulcer treatment a review." Acta biomaterialia 9.7 (2013): [7] Sanders, Antal P., et al. "Patient and professional delay in the referral trajectory of patients with diabetic foot ulcers." Diabetes research and clinical practice (2013): [8] Bennetts, Craig J., et al. "Clustering and classification of regional peak plantar pressures of diabetic feet." Journal of biomechanics 46.1 (2013): [9] Molvær, Anne Karin, et al. "Diabetes-related foot ulcers and associated factors: Results from the Nord-Trøndelag Health Survey (HUNT3)( )." Journal of Diabetes and its Complications 28.2 (2014): [10] Ledoux, William R., et al. "Diabetic foot ulcer incidence in relation to plantar pressure magnitude and measurement location." Journal of Diabetes and its Complications 27.6 (2013): [11] McEwen, Laura N., et al. "Prevalence and risk factors for diabetes-related foot complications in Translating Research Into Action for Diabetes (TRIAD)."Journal of diabetes and its complications 27.6 (2013): [12] Tabur, Assist Prof Dr Suzan, et al. "The major predictors of amputation and length of stay in diabetic patients with acute foot ulceration." Wiener klinische Wochenschrift (2015): [13] Tin, Si Thu Win, et al. "Diabetes related amputations in Pacific Islands countries: A root cause analysis of precipitating events." Diabetes research and clinical practice (2013): [14] Sun, Jui-Hung, et al. "Risk factors for lower extremity amputation in diabetic foot disease categorized by Wagner classification." Diabetes research and clinical practice 95.3 (2012): editor@iaeme.com

9 J. Jayashree, J. Vijayashree [15] Li, Xiang, et al. "Incidence, risk factors for amputation among patients with diabetic foot ulcer in a Chinese tertiary hospital." Diabetes research and clinical practice 93.1 (2011): [16] Imran, Shaikh, Rajab Ali, and Ghulam Mahboob. "Frequency of lower extremity amputation in diabetics with reference to glycemic control and Wagner's grades." Journal of the College of Physicians and Surgeons--Pakistan: JCPSP 16.2 (2006): [17] Abad, C., and N. Safdar. "From ulcer to infection: an update on clinical practice and adjunctive treatments of diabetic foot ulcers." Current infectious disease reports 14.5 (2012): [18] J. Jayashree., N.Ch. Sriman Narayana Iyengar., "A Review on Mobile Based Systems for Homecare Monitoring of Diabetic Mellitus Foot Ulcer", International Journal of Smart Home, 10.8 (2016): [19] Preethi N Patil and G. G. Rajput. Detection and Classification of Non Proliferative Diabetic Retinopathy Stages Using Morphological Operations and Svm Classifier. International Journal of Computer Engineering and Technology, 4 (6), 2013, pp [20] Sorna Percy. G and Dr. T. Arumuga Maria Devi, An Efficiently Identify The Diabetic Foot Ulcer Based On Foot Anthropometry Using Hyper Spectral Imaging. International Journal of Information Technology & Management Information System, 7 (2), 2016, pp [21] [22] [23] [24] [25] [26] [27] [28] editor@iaeme.com