Journal of Biomedical Imaging and Bioengineering

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.
Reach Us +44-7360-538437

Research Article - Journal of Biomedical Imaging and Bioengineering (2022) Volume 6, Issue 8

A survey on recent cancer using various technologies.

Udayakumar E1*, Gowrishankar R1, Vetrivelan P2

1Department of Electronics & Communication Engineering, Kalaignarkarunanidhi Institute of Technology, Tamil Nadu, India

2Department of Electronics & Communication Engineering, PSG Institute of Technology and Applied Research, Coimbatore, Tamil Nadu, India

Corresponding Author:
Udayakumar E
Department of Electronics & Communication Engineering
Kalaignarkarunanidhi Institute Of Technology
Tamil Nadu, India
E-mail:
[email protected]

Received: 18-Aug-2022, Manuscript No. AABIB-22-72260; Editor assigned: 22-Aug-2022, Pre QC No. AABIB-22-72260 (PQ); Reviewed: 5-Sep-2022, QC No. AABIB-22-72260; Revised: 12-Sep-2022; AABIB-22-72260(R); Published: 19-Sep-2022, DOI: 10.35841/aabib-6.8.136

Citation: Udayakumar E, Gowrishankar R, Vetrivelan P. A survey on recent cancer using various technologies. J Biomed Imag Bioeng.2022;6(8):136

Visit for more related articles at Journal of Biomedical Imaging and Bioengineering

Abstract

Malignant growth prompts the most passing’s all around. It is conceivable to effortlessly lessen passing’s brought about by malignant growth, by identifying it early and implementing preventive measures. Natural and hereditary elements assume a critical job in concocting new techniques for the discovery and anticipation of malignant growth. The methodology for identifying and forestalling malignancy which is portrayed here depends on an investigation which utilizes affiliation rule mapping. The information for investigation is identified with history of enslavement, clinical indications, survivability and co-grim conditions identified with malignant growth patients. It is conceivable to find designs in huge arrangements of information through information mining which is a computational procedure. Information mining utilizes a blend of procedures identified with database frameworks, insights, Artificial Intelligence and AI. Information mining assists with removing and change information in an informational collection. Huge and complex information in clinical informational indexes can be examined utilizing AI systems and calculations. In the on-going occasions, various scientists have utilized AI approaches in the Healthcare business. This survey shows the information mining advances being used is bunching which assists with identifying and group potential non-malignant growth and oral disease patients.

Keywords

Skin Cancer, Artificial intelligence, Machine learning, K-means clustering, Neural networks and deep learning.

Introduction

The focal point of the current survey is to structure and execute an oral malignant growth distinguishing proof and discovery framework utilizing Deep Neural based Adaptive Fuzzy System (DNAFS). DNAFS utilizes both fluffy rationale and profound neural system based calculation. The preparing is dealt with by first starting information assortment by social occasion online data for the clinical informational collection. This is trailed by pre-processing which helps in information cleaning, evacuation of undesirable information and void qualities. Next, information is grouped by Fuzzy C-Means bunching for additional handling. Information examples [1]. can be recognized by highlight determination. This procedure helps feature similitudes and contrasts. Next, the strategies in DNAFS and information digging are investigated for distinguishing proper procedures and techniques for characterizing information productively. At long last, information is grouped utilizing DNAFS, an AI method which is a famous classifier in the field of AI identified with recognition of oral malignant growth. A famous classifier utilized in various fields relating to Medical Research is DNN which is straightforward just as simple to execute. Neural systems depend on the capacities and structure of natural neural systems which comprises of neurons. Fluffy rationale relates to the characterization of gigantic measures of information for getting exact outcomes. In the ebb and flow explore, a standard based arrangement procedure is proposed for recognition of oral disease. This is one of the quickest fluffy methods dependent on setting rules for clinical procedures and ascribes relating to separation between oral disease and ordinary cases. In the current examination, about 94% exactness in characterization was acquired.

There are two classes of grouping calculations. These are hard grouping and delicate or fluffy bunching. In hard grouping, information is partitioned into bunches with the end goal that each datum component has a place with just one group. These calculations have accomplished great outcomes for various genuine applications [2]. In this stage, the extricated include is adjusted as contribution to the proposed Fuzzy C-implies for the grouping of mind tumors. K-Means, KNN and Fast Fuzzy C-Means bunching strategies were utilized to group dangerous just as favourable tumors. To all the more likely comprehend the outcome; a correlation was made between the exhibitions of the Fuzzy C Means, Fast Fuzzy C-Means and KNN calculations. Improved characterization result was gotten by the half breed model proposed in the examination in correlation with the ordinary calculations being used in the past [3]. In the clinical science field, the grouping and bunching of malignant growth information has demonstrated great outcomes [4]. Exact outcomes can be acquired using the Fuzzy Cmeans and K-implies calculations in the proposed arrangement. The current survey delivers the issue identified with characterization of malignant growth information utilizing the techniques and data got from the periods of testing and preparing. A bunching calculation were looked at further for their exhibitions utilizing outline scores.

The K-implies calculation was at first examined with the assistance of various bunches and separation measurements. Later on, the DBSCAN calculation was investigated utilizing diverse least focuses that are required to shape bunches and separation measurements. Based on this examination, the correlation of the exhibition of the K-implies calculation was better [5]. than DB filter. In particular, better qualities were accomplished as far as execution time and bunching precision. The manner by which groundwater influences human wellbeing was broke down by utilizing the K-implies grouping calculation. For this situation, the hazard factors for substance of fluoride in water were dissected. Important concealed examples were uncovered with the assistance of this investigation which can aid sensible choices on a network level. It conceived a calculation for anticipating Heart Attack through fluffy C-implies. This calculation utilizes unaided bunching where one explicit information item can have a place with in excess of a solitary group. For this situation, the framework can assist doctors with the effective determination of respiratory failure. The bio-profile idea was utilized. K-implies calculation for bunching was utilized for recognizing creature ailment early and its arrangement into non-pathologic and pathologic gatherings. Specialists showed that the clinical space information as for grouping will in general improve the algorithmic exhibition. Towards this end, the grouping calculation execution is assessed with either Neural Gas or Fuzzy C-Means [6]. for the location of tumors. Three distinct advances are utilized for this situation. In the initial step, the three calculations are assessed under the states of commotion. Afterward, tumor division results are contrasted and district developing calculation. At long last, results are physically contrasted and division results [7].

Skin malignancy is a usually and locally dangerous harmful development of the skin. There are three significant kinds of skin Image handling is one of the generally utilized techniques for skin malignant growth location. The majority of the analyst taken as dermoscopy pictures be a non-obtrusive assessment strategy bolstered the reasons for episode light bar and oil submersion method to type of potential the visual examination of surface structures of the skin. Melanoma can spread out all pieces of the body through lymphatic framework or blood [1]. There are some particular highlights that recognize harmful melanoma from other three kinds of amiable melanoma. The skin disease Image. In this proposed strategy a PC can conclusion of melanoma by utilizing dermoscopic pictures and furthermore characterizing its sort as whether picture is dangerous or non-destructive. This strategy utilizes both the picture preparing for include extraction and fake neural system for order reason. It is a piece of man-made consciousness and has acknowledged as another procedures in software engineering [8].

The ANN figuring’s made for dealing with the clinical picture and assessment every now and again become more sharp than show methodology [2]. By and large, AI models are set up to perform supportive tasks subject to truly organized features rough data, or features direct AI schemes. The significant supportive depictions and features thusly, honestly unrefined data, by this manual and irksome development. As far the most generally perceived significant various varieties of phony frameworks, anyway there are others. The central standard characteristic of significant learning methodologies is their accentuation on incorporate learning: normally learning depictions of data. This is the fundamental qualification between significant learning draws near and anything is possible from that point "old style" AI. Discovering features a task joined likewise during a comparative getting ready procedure.

Related work

This thought got the masters excitement for the continuous years for its extraordinary execution to transform into the best course of action in various issues in clinical picture assessment applications, for instance, picture denoising, division selection and gathering [9]. Our proposed way of thinking subject to the DNN getting the hang of building for gathering where the classifier is recognizing the cerebrum tumors as a top priority MRIs. The proposed way of thinking for describing the psyche tumors in cerebrum MRIs is according to the accompanying. The features for every cerebrum MRI. Regardless of the way that this number isn't so gigantic stood out from the amount of feature maps occurred by the convolution channels of CNNs yet we used the fundamental parts assessment (PCA) to inferred the primary isolated features with lower dimensional component vectors. Game plan after the features are expelled and picked, the portrayal step using DNN is performed on the came about segment vector. Game plan is performed by using 7-overlay cross endorsement framework for building and setting up the DNN of 7 covered layers structure. Moreover for evaluating the introduction of the picked classifier, we used other AI portrayal computations from WEKA using comparable norms, the picked plan computations [10].

Counterfeit neural frameworks Artificial neural frameworks (ANNs) is one of the most notable AI models, introduced starting at now during the 1950s, and viably mulled over. Around, a neural framework involves different related computational units, called neurons, arranged in layers. There's a data layer where data enters the framework, trailed by at any rate one covered layers changing the data as it travels through, before conclusion at a yield layer that makes the neural framework's estimates. The framework is set up to yield accommodating desires by perceiving structures in a great deal of stamped getting ready data, dealt with through the framework while the yields are differentiated and the genuine names by an objective work. During preparing the system's parameters the quality of every neuron is tuned until the examples recognized by the system bring about great forecasts for the preparation information. When the examples are found out, the system can be utilized to make forecasts on new, inconspicuous information, for example sum up to new information. It has for quite some time been realized that ANNs are truly adaptable, ready to demonstrate and take care of muddled issues, yet in addition that they are troublesome and computationally costly to prepare.

This has brought down their pragmatic utility and driven individuals to, as of not long ago, center around other AI models. Be that as it may, at this point fake neural systems structure one of the prevailing techniques in AI, and the most seriously examined. This change is because of the development of huge information, ground-breaking processors for equal calculations (specifically, GPUs), some significant changes to the calculations used to build and train the systems, and the advancement of simple to-utilize programming structures. The flood of enthusiasm for ANNs prompts an unfathomable pace of improvements, which additionally drives different pieces of AI with it. The uninhibitedly accessible books are two of the numerous fantastic sources to become familiar with fake neural systems. We'll just give a concise sign of how they are developed and prepared. The fundamental type of counterfeit neural systems, the feed forward neural systems, are parametrized scientific capacities y = f(x; θ) that maps an info x to a yield y by taking care of it through various nonlinear changes. Here every segment fk, called a system layer, comprises of a basic direct change of the past part's yield, trailed by a nonlinear capacity: The nonlinear capacities σk are normally sigmoid capacities or ReLUs, as talked about beneath, and the θk are networks of numbers, called the model's loads.

During the preparation stage, the system is taken care of preparing information and entrusted with making forecasts at the yield layer that coordinate the known names, every part of the system creating a catalyst portrayal of its information. It needs to figure out how to best use the moderate portrayals to frame a complex various leveled portrayal of the information, finishing off with right expectations at the yield layer. Preparing a neural system implies changing its loads to improve the yields of the system. This is finished utilizing an improvement calculation, called inclination plunge, on a capacity estimating the rightness of the yields, called a cost capacity or misfortune work. The fundamental thoughts behind preparing neural systems are straightforward: as preparing information is taken care of through the system, register the angle of the misfortune work as for each weight utilizing the chain rule, and diminish the misfortune by changing these loads utilizing inclination plummet. Be that as it may, one rapidly addresses enormous computational difficulties when confronted with entangled systems with thousands or a large number of parameters and an exponential number of ways between the hubs and the system yield. The methods intended to beat these difficulties gets very convoluted [11].

The discriminative intensity of the scholarly highlights is shown accomplishment ongoing frameworks in coordinating execution spaces recently excessively aspiring. Investigation of the consequently separated data might give novel bits of knowledge and encourage explore on pathologies for which minimal earlier information is as of now accessible. While trying to outline this, we investigate what examples have been realized consequently for the sore division errands. We envision the initiations of Deep Medic's handling. Many showing up designs are hard decipher, particularly more profound. One of the most intriguing discoveries is that the system figures out how to distinguish the ventricles, CSF, white and dim issue. This uncovers separation of is valuable for sore division. This accordance discoveries writing, division execution with conventional fundamentally consolidation. It is instinctive that various sorts of injuries influence various pieces of the cerebrum relying upon the basic systems of the pathology. A thorough investigation of spatial signs removed by the system may uncover relationships that are not all around characterized at this point. So also captivating is the data separated in the low resolution pathway. As they process more noteworthy setting, these neurons increase extra restriction abilities. The actuations of specific FMs structure fields in the encompassing regions of the mind. These examples are saved in the most profound concealed layers, which demonstrates they are useful for the last division [12].

System design

The key goals of danger gauge and expectation are specific from the targets of infection area and end. In the essential case, one is endeavouring to envision the likelihood of working up a sort of threatening development going before the occasion of the sickness. At any rate a disorder conjecture can simply come after a clinical finding and a prognostic gauge must think about something past a fundamental end. The 100% accuracy is accomplished for an anticipated desire is significant. Without a doubt more likely than not, this ANN was over-arranged. The size of a given instructive assortment in like manner basically impacts the model per-feature extent. If all else fails, the sampleper-incorporate extent should be at any rate 5-10. The block diagram of skin cancer identification is shown in Figure 1.

Figure 1: Block diagram of skin cancer identification.

A model for each segment extent this little is uncommonly frail to the issues of overtraining. Also, with a model for each part extent. Information size isn't the main impediment for viable AI. Informational index quality and cautious element choice are likewise similarly significant (review: "trash in=garbage out"). For enormous informational collections information section and information confirmation are of vital significance. Regularly indiscreet information section can prompt straightforward off-by-one mistakes in which all the qualities for a specific variable are moved up or somewhere around one line in a table. This is the reason free confirmation by [13] a second information passage caretaker or information checker is constantly valuable. Further confirmation or spot checking of information respectability by an educated master, not only an information passage representative, is additionally an important exercise.

The strategies utilized to guarantee information quality and [14]. honesty are infrequently talked about in most AI surveys. Similarly as information is significant. In a perfect world highlights ought to be picked that are reproducible and correctly quantifiable from one lab. One examination utilized "essential site code" and "site explicit medical procedure code" as highlights to foresee bosom disease survivability. While these clinical highlights might be useful in deciding the result for bosom malignant growth patients at this specific emergency clinic, for this second in time, they may get immaterial after some time. Far more terrible, or site explicit medical procedure made, should be re-prepared. Comparative component determination issues regularly happen with histological evaluations [15]. The sample input images is shown in Figure 2.

Figure 2: Sample input images.

On a par with numerous pathologists are there is in every case some irregularity (up to 30% by and large) between various histopathological appraisals from various destinations or various pathologists. When in doubt, the best highlights exceptionally. Indeed, even with these apparently vigorous highlights critical to recollect informational indexes substances. With time significance pertinence develop after some certain highlights included, altered erased. Accordingly a classifier [10] should likewise have the option to adjust to various capabilities after some time as well. Another significant exercise that was found out from surveying huge numbers of these AI surveys was the benefit of utilizing various indicator models dependent on various AI strategies.

While ANNs are regularly considered to be complex and propelled AI techniques, ANNs are not generally the most ideally equipped apparatuses for the activity. Here and there more straightforward AI strategies, similar to the guileless Bayes and choice tree techniques can significantly beat ANNs. Surveying the exhibition of an AI indicator against different indicators is basic to picking the ideal apparatus. It is likewise basic to choosing if the technique is any superior to already existing plans. In a perfect world, any recently distributed AI model ought to be analyzed, for example, arranging framework. Here and there the more modern AI techniques don't prompt the best indicators. Now and again, customary measurements really beat machine [16].

For most human ailments, anticipation is commonly desirable over finding a fix, both to improve the odds of fruitful treatment and as the more affordable methodology. In malignant growth, early tumor location with non-obtrusive imaging, for example, CT, MRI and PET sweeps – known as radiomics – is demonstrating progressively encouraging for segregating among lethargic and forceful tumors. Medical procedure specifically can turn out to be particularly successful, even therapeudic [17]. To control specialists, improved imaging holds the guarantee of better characterizing and expelling tumor edges, and picture acknowledgment programming would now be able to arrange skin melanomas. Be that as it may, these are costly intercessions that likewise increment a person's introduction to radiation and can have a high bogus constructive rate. The glucose simple 18F-fludeoxyglucose, utilized for PET imaging, gauges the glycolytic motion of disease cells dependent on their trademark expanded glucose take-up. This, related to different tracers, allows the recognition of metabolic changes before tissue-level modifications happen. Further open door ought to be taken to abuse these assets to lighten the requirement for biomarkers while surveying applicant helpful targets.

Investigation of the clonal elements of tumor inception (for model, in hereditarily built mouse models (GEMMs) utilizing advances [18], for example, information examination and exome sequencing could likewise encourage early conclusion. Promising anticancer methodologies, for example, the developing field of immunotherapies, may demonstrate increasingly successful during prior phases of the ailment, before a patient's resistant framework is battered by cutting edge ailment and by earlier therapeutics. A key test is budgetary: the present clinical protection framework in the US at any rate offers no monetary help for refining avoidance systems, for example, malignant growth forestalling antibodies, a procedure that has demonstrated effective for HPV. By and large, moving toward malignant growth as a network challenge can yield significant advantages and is significant for essential scientists to consider in their disease examines. Better access to assets. The Mask detection of primary and secondary region is shown in Figure 3.

Figure 3: Mask detection of primary and secondary region.

The eventual fate of malignant growth research will progressively depend on improved coordinated effort as we grow increasingly [19] advanced restorative, analytic and prognostic apparatuses. Improving open accessibility of information will require concession to explicit measures for information taking care of, preparing and sharing.

The DREAM mammography imaging challenge and a few medication recombination challenges are a genuine case of the intensity of publicly supporting utilizing openly accessible. The difficulties we are confronting are perplexing and will require an inexorably expansive range of skill. From fundamental research to clinical preliminaries and infection counteraction, the disease field needs to elevate expanded open access to assets, for example, quality articulation and radiomics information, and patient tissue tests. Moreover, considering fast upgrades in the improvement of organoids, ex vivo culture models and patient-determined xenograft models [20], new ways to deal with saving tissue, for example, cryopreservation, would permit future work on refined tumors/models and along these lines increasingly thorough examination. By normalizing cryopreservation, tissues can be 'revived' and contemplated, moving disease pathology from absolutely basic examinations to utilitarian investigations that could decidedly affect patients just as research. Growing better models one continuous inquiry during this workshop was the manner by which to join model living beings into the medication improvement process. The beginning stage of another malignancy study ought to be grounded in information from human patients or from epidemiological information; creature or cell models ought to be intended to catch however much tumor multifaceted nature as could reasonably be expected. For instance, tumors are significantly more than DNA changes; discovering pathway-based associations, for example, between the oncogenes Ras and AKT helps catch significant parts of malignancy science. Competitor therapeutics ought to be tried in various models, on the grounds that the hereditary foundation can be basic in clarifying the impact of a specific change. In any case, mouse models speak to the key mammalian model utilized preceding moving medication hits into human preliminaries. A key point that rose up out of the workshop is the significance of appropriately deciphering information from model frameworks: we ought [20] not anticipate that patients should react more heartily than our models do. Models that are utilized for patients ought to be utilized when testing treatments in mice; for instance, mouse xenograft information ought to be shown as cascade plots, and in any event 30% tumor shrinkage ought to be seen in the dominant part [14].

Each AI application needs purpose behind underfitting ordinarily lies either in the model, which comes up short on the capacity to communicate the unpredictability of the information, or in the highlights, which don't sufficiently portray the information. This unavoidably prompts a high preparing blunder. Then again, an overfitted model deciphers some portion of the clamor in the preparation information as significant data, in this way neglecting to dependably foresee new information [21]. Normally catch intricacy with preparation information. So as to abstain from overfitting, it is fundamental to screen during preparing the preparation blunder as well as the mistake of the approval set. When the approval blunder quits diminishing, an AI model can begin to overfit. This issue is likewise examined as the predisposition change exchange. In unique circumstance, inclination a mistake dependent suspicions prepared high difference blunder coming about because of an excessive amount of affectability to commotion in the preparation information. In that capacity, it must be assessed on already inconspicuous information, signified as test set, to gauge its speculation and extrapolation capacity.

Various techniques extending a straightforward k-overlay approval, forget about one approval, cross validation, to forget about one-bunch cross-validation73 can be utilized for the assessment. Every one of these techniques depend on keeping a few information avoided preparation procedure. Contingent upon the objective amount, this takes into consideration a progressively practical assessment and a superior comprehension of the confinements of the AI model. Forget about one-bunch cross-approval evacuates a group having a place with the expelled group.

The accomplishment of such strategies relies for the most part upon the sum and nature of information that is accessible, and this ends up being significant difficulties. The hazardous must resolved tentatively in an exorbitant manner. The analysts from various orders ought to have the option to discover and get to information, just a regularly acknowledged configuration. It permits utilization information, generally, outcomes frequently disposed. Be that as it may, as information frequently similarly as significant for AI calculations as a social change toward the distribution of ineffective essential. The certain orders more drawn out convention of information based research, such databases as of now. A comparable, information develops item yet fundamental a distribution regularly. The long run outcomes in a misuse of assets as different specialists are then required to rehash the work. At long last, each and every disposed of estimation will be painfully missed in future AI applications [8].

Conclusion

This survey presented the significant element of an AI calculation is the portrayal of the information in an appropriate structure. Highlights in material science must have the option to catch all the applicable data, important to recognize diverse nuclear or gem. A procedure, indicated highlight building, may basic deciding nucler, may include complex changes, for example, a development of spiral appropriation capacities (RDFs) in a specific premise, or may require collections dependent on. Weyl grids are characterized by the inward item.

References

  1. Lalithmani K, Punitha A. Detection of oral cancer using deep neural based adaptive fuzzy system in data mining techniques. Int J Rec Tech Eng. 2019 Feb; 7:397-404.
  2. Google Schoalr

  3. Praveenkumar GD, Dharmalingam M. Detection of skin cancer using artificial neural network classifier. Int J Res Appl Sci Eng Technol. 2018; 6(3):1884-9..
  4.  Google Schoalr

  5. Udayakumar E, Santhi S, Vetrivelan P, et al. TB screening using SVM and CBC techniques. Curr Pediatr Res. 2017.
  6. Indexd at, Google Schoalr

  7. Schmidt J, Marques MR, Botti S, et al. Recent advances and applications of machine learning in solid-state materials science. Npj Comp Mater. 2019; 5(1):1-36.
  8.  Google Schoalr, Cross Ref

  9. Seiwerth S, Štambuk N, Konjevoda P, et al. Immunohistochemical analysis and prognostic value of cathepsin D determination in laryngeal squamous cell carcinoma. J Chem Inf  Comput Sci. 2000; 40(3):545-49.
  10. Indexd at, Google Schoalr, Cross Ref

  11. Udayakumar E, Santhi S, Gowrishankar R, et al. An unified Reeb analysis for cortical surface reconstruction of MRI images. Biomed Pharm J. 2017; 10(2):939-45.
  12. Google Schoalr, Cross Ref

  13. Murugan R. Implementation of deep learning neural network for retinal images. InHandbook of Research on Applications and Implementations of Machine Learning Techniques 2020; 77-95. IGI Global.
  14. Google Schoalr, Cross Ref

  15. Udayakumar E, Santhi S, Yogeshwaran JR, et al. Automatic detection of diabetic retinopathy through optic disc using morphological methods. Asian J  Pharm  Clin Res, Innovare Academic Sciences. 2017; 10(4).
  16. Indexed at, Google Schoalr, Cross Ref

  17. Mohsen H, El-Dahshan ES, El-Horbaty ES, et al. Classification using deep learning neural networks for brain tumors. Future Comput Infor J. 2018; 1:3(1):68-71.
  18. Google Schoalr, Cross Ref

  19. Ramesh C. A Review on diagnosis of Malignant Melanoma from Benign Lesion by using BPNN and ABCD Rule Parameters. Int Res J  Pharm. 2018; 9(10):1-7.
  20. Google Schoalr

  21. Lundervold AS, Lundervold A. An overview of deep learning in medical imaging focusing on MRI. Zeitschrift für Medizinische Physik. 2019; 29(2):102-27.
  22. Indexed at, Google Schoalr, Cross Ref

  23. Konstantinos Kamnitsas, “Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation”. Medical Image Anal. 2017; 36:61–78.
  24. Indexed at, Google Schoalr, Cross Ref

  25. Udayakumar E, Yogeshwaran K, Ramesh C, et al. An efficient tissue segmentation of neonatal brain magnetic resonance imaging. Res J  Pharm  Technol. 2019;12(6):2963-966.
  26. Google Schoalr, Cross Ref

  27. Cruz JA, Wishart DS. Applications of machine learning in cancer prediction and prognosis. Cancer Infor. 2006; 2:117693510600200030.
  28. Indexed at, Google Schoalr, Cross Ref

  29. Datta P, Rohilla R. An Introduction to Deep Learning Applications In MRI Images. In2019 2nd International Conference on Power Energy, Environment and Intelligent Control (PEEIC) 2019 Oct 18 (pp. 458-465). IEEE.
  30. Sivaganesan S. Design and development of smart glucose monitoring system. Int J Pharma  Biosciences. 2017; 8(3):631-38.
  31. Google Schoalr

  32. De Silva SF, Alcorn J. Flaxseed lignans as important dietary polyphenols for cancer prevention and treatment: Chemistry, pharmacokinetics, and molecular targets. Pharma. 2019; 12(2):68.
  33. Indexed at, Google Schoalr, Cross Ref

  34. Ramesh C, Udayakumar E, Yogeshwaran K, et al. Detection and segmentation of optic disc in fundus images. Int J Curr Pharm Res, Innovare Academic Scie. 2018; 10(5).
  35. Google Schoalr, Cross Ref

  36. Sindhumathy S. Analysis of Magnetic Resonance Image Segmentation using spatial fuzzy clustering algorithm. J Glob Pharma Technol. 2018; 10(12):88-94.
  37. Google Schoalr

  38. Udayakumar E, Santhi S, Vetrivelan P, et al. An Identification of efficient vessel feature for Endoscopic Analysis. Res J of Pharm Technol. 2017; 10(8):2633-636.
  39. Google Schoalr, Cross Ref

  40. Matos CR, Xavier MJ, Barreto LS, et al. Principal component analysis of X-ray diffraction patterns to yield morphological classification of brucite particles. Anal Chem. 2007; 79(5):2091-095.
  41. Indexed at, Google Schoalr, Cross Ref

Get the App