Category: Body

Introduction

Anatomy undoubtedly forms the basis of any medical curricula as well serves a strong component for good clinical practice. However, literature pertaining to anatomy education is increasingly reporting of a deterioration of anatomy knowledge not only among undergraduate students but also among clinical practitioners. Partly in response to these observations and due to changes in demands of medical professions, anatomy curricula have experienced a major paradigm shift from passive didactic and teacher centered approach to active, clinical based and student-centered approach. However comparative retrospective analysis of these different approaches has failed to identify a significant difference in terms of student performance. A possible explanation is that both approaches have failed to emphasize key competencies required to be known by the student. This has led to a deprivation of a strong anatomy foundation, a common observation made amongst most medical undergraduate students. The authors herewith propose a competency level based approach for teaching anatomy whereby the understanding of anatomy is allowed to evolve within the student based on pre defined competency levels. The proposed approach enables the student to develop a strong core knowledge via a teacher centered, didactic commencement and ends in a complete student centered, clinical based approach allowing for application and synthesis of new knowledge based on the already laid strong foundation.

The big picture vs details

Different approaches used in teaching anatomy, whether traditional or novel is observed to share one thing in common: the focus on detailed anatomy of a selected region. Regional anatomy taught in traditional curricula for example presents the upper limb under the main subdivisions of pectoral region, shoulder region, the arm, forearm and the hand. This approach is seen to allow the student to focus on detailed anatomy of each sub region discussed and leaves the student to amalgamate the knowledge gained so as to perceive the upper limb as a single functional unit. Careful analysis of “problem based learning” and other modern reforms to the traditional curricula is also observed to be lacking in this sense, thus leaving the challenge of amalgamation to the student.

The authors believe that the possession of a holistic understanding of anatomy is the key to successful anatomy learning. Good clinical practice requires the understanding of different anatomical entities as a single functional unit, reflecting the role in nature. Possessing detailed knowledge of regional anatomy will therefor only cater partly to the demands of the practice of anatomy in the clinical setting. The remining greater proportion thus relies on the student’s capability of self-amalgamating the acquired knowledge which woefully is not observed amongst the majority of the student population today. It is in this context that authors also argue on the threshold of such amalgamating capacities expected from undergraduate medical students.

The proposed approach for teaching anatomy is designed in such a way so as to substantially rid the student of the need of self-amalgamation. This is brought about through pre identified competency levels with each defined level designed to perceive this broader anatomical picture. Thus, at any given time it is expected that a selected student will possess a holistic understanding of a learnt functional unit appropriate of his/her competency level. The gradual course of a student through increasing competency levels will enable him/her to add required details to the broader picture whereby a deeper understanding is generated.

The evolving nature

Two major criteria (center and discipline) will be used when defining each competency level. Commencing from a teacher centered mode of deliverance the teaching methodology is gradually shifted toward the student with increasing competency level. Thus, competency level I which is identified as a complete teacher centered approach makes a gradual transit towards a fully student-centered approach on reaching competency level IV. The teaching learning methodology will also undergo a similar transition reflecting the change of center. Lectures and prosected specimen observation employed as the main teaching learning methodology at competency level I will change to student presentations and dissections when reaching the competency level IV. A simultaneous shift will also be conducted with regard to the type of anatomy taught. Commencing with core structural anatomy at competency level I, the student will course through descriptive to functional and clinical anatomy at competency level IV. Sub disciplines of histology and embryology is also expected to be presented in a similar manner with core knowledge being presented at competency level II and III and with abnormal development, histopathology and molecular basis being introduced at competency level IV. The gradual transition of the aforementioned criteria exemplifies the dynamic nature of the proposed approach when compared to traditional as well as novel teaching methods such as PBL.

Vertical integration

In the setting of basic sciences, integrated learning has proven to be effective both in terms of factual comprehension and promoting life long learning. However, studies conducted in this regard has yet to answer questions pertaining to the extend as well as exact points of such integrations. Retrospective analysis of student performance in most problem-based learning curricula which accommodates a marked degree of vertical integration has failed to identify a significant improvement in terms of student performance. In contrary to the expectations such approaches have been observed to leave students with sporadic knowledge. The authors thus present the argument that a strong core anatomy knowledge is mandatory and forms the foundation for successful vertical integration. In bringing this ideology to light, the proposed approach herewith suggests to emphasize functional and clinical anatomy learning during latter stages of competency levels, a stage where the student has already mastered his core knowledge and is ready to apply.

Structuring formative and summative assessment

While summative assessment is considered to evaluate a student’s learning progress, formative assessment can be considered as an approach to reinforce learning. Formative assessment also plays a key role in motivating the students to learn thus allowing for continuous improvement. In this perspective formative assessment can be considered to be a crucial component of a curricula and conduction of a systematic formative assessment not only is beneficial directly to the student but also allows for identification of areas to be improved within a curriculum. However, conducting a successful formative assessment demands a sensible partition of the learnt subject matter in such a way the student himself can identify his/her weaknesses and apply the experience during the next level. The competency level-based teaching approach presented herewith facilitate this process of partition enabling a successful formative assessment to be carried out.

Summative assessments are usually applied at the end of a period of instruction to measure the outcome of student learning. Such assessments ideally require a structure covering all aspects of knowledge gained over the specified time period. However, a grey area exists in defining such a structure within a traditional or a novel approach such as PBL. The competency level based approach presented herewith also facilitates the structuring process of a summative assessment via the outcome based different weightages which can be allocated to each competency level.

Chapter 1. Introduction

Though we may take for granted our brain’s ability to recognize the faces of friends, family, and acquaintances, it is actually an extraordinary gift. Designing an algorithm that can effectively scan through a series of digitized photographs or still video images of faces and detect all occurrences of a previously encountered face is a monumental task. This challenge and many others are the focus of a broad area of computer science research known as facial recognition. The discipline of facial recognition spans the subjects of graphics and artificial intelligence, and it has been the subject of decades of research and the product of significant government and corporate investment. Face recognition systems have been conducted now for almost 50 years, which makes it to be one of the researches in the area of pattern recognition & computer vision due to its numerous practical applications in the area of biometrics, Information security, access control, law enforcement, smart cards, and surveillance system.

Biometric-based techniques have emerged as the most promising option for recognizing individuals in recent years since, instead of certifying people and allowing them access to physical and virtual domains based on passwords, PINs, smart cards, plastic cards, tokens, keys and so, these methods examine an individual’s physiological and/or behavioral characteristics in order to determine and/or ascertain his/her identity. Passwords and PINs are difficult to remember and can be stolen or guessed; cards, tokens, keys, and the like can be misplaced, forgotten, or duplicated; magnetic cards can become corrupted and unclear. However, an individual’s biological traits cannot be misplaced, forgotten, stolen, or forged, however, a facial recognition system is a computer application capable of identifying a person from a digital image or video frame from a video source. The system of verifying and identifying the human face is among the few methods of biometric identification which gets the merits of a high level of accuracy and non-intrusiveness. This is because facial recognition provides details about the age, personal identity, gender, emotional state, and mood of a person. A great deal of achievements has been recorded in the area of facial recognition over the years. However, despite these achievements, facial recognition still stands as an active research area due to the changeability perceived in facial appearance as a result of illumination, expression occlusion, pose, age, and plastic surgery.

In the area of face recognition, several approaches have been proposed to address the challenges of illumination, pose, expression, aging, and disguise. However plastic surgery-based face recognition is still a lesser explored area. Thus the use of face recognition for surgical faces introduces a new challenge for designing future face recognition systems. [1]. Plastic surgery is a sophisticated operational technique that is used across the world for improving facial appearance. For instance to remove acne scars, to become white, to remove dark circles, and many more. Plastic surgery can be broadly classified in two different categories such as global plastic surgery and local plastic surgery. Global surgery changes the complete facial structure whereas in local plastic surgery certain parts of the face are changed. Recognizing a face after plastic surgery might lead to the rejection of genuine users or the acceptance of impostors. To this challenge yet much literature is not available. Very few researchers now have contributed in this field. Many researchers have shown a comparative study of different face recognition algorithms for plastic surgery. Based on the experimentation carried out but it has been concluded that face recognition algorithms such as PCA, FDA, GF, LLA, LBP, and GNN have shown a recognition rate of not more than 40% for local plastic surgery. Moreover, for global surgery, it was merely up to 10%. Among all the algorithms, the geometrical feature-based approach has proven to a great extent comparatively for local plastic surgery.

One challenge that is affecting the verification and identification of human faces using face detection algorithms is facial plastic surgery. A human face that passes through surgery alters the features to be used by these algorithms in verifying and identifying a face. When a given face undergoes surgery, criminals or evaders hide their identities and reside within the society smoothly, which means the standard of face recognition is compromised making it an issue to be dealt with using technology. Hence, in order to fort-nail these trepidations, recognition of human faces by computer algorithms had to spread its tentacles and address this issue successfully. Statistical data has shown that the use of plastic surgery to alter faces in humans is growing exponentially [1]. This is because of the improvement in technology which is making plastic surgery less painful, fast, and affordable for many within society. Furthermore, two key issues have to be considered:

• The overall speed of the system from detection to recognition should be acceptable.
• The accuracy should be high.

Problem Definition

The face recognition problem can be formulated as follows: Given an input face image and a database of face images of known individuals, how can we verify or determine the identity of the person in the input image?

Why Use the Face for Recognition

Biometric-based techniques have emerged as the most promising option for recognizing individuals in recent years since, instead of authenticating people and granting them access to physical and virtual domains based on passwords, PINs, smart cards, plastic cards, tokens, keys, and so forth, these methods examine an individual’s physiological and/or behavioral characteristics in order to determine and/or ascertain his identity. Passwords and PINs are hard to remember and can be stolen or guessed; cards, tokens, keys, and the like can be misplaced, forgotten, purloined, or duplicated; magnetic cards can become corrupted and unreadable. However, an individual’s biological traits cannot be misplaced, forgotten, stolen or forged.

Biometric-based technologies include identification based on physiological characteristics (such as the face, fingerprints, finger geometry, hand geometry, hand veins, palm, iris, retina, ear, and voice) and behavioral traits (such as gait, signature and keystroke dynamics) [1]. Face recognition appears to offer several advantages over other biometric methods, a few of which are outlined here: Almost all these technologies require some voluntary action by the user, i.e., the user needs to place his hand on a hand-rest for fingerprinting or hand geometry detection and has to stand in a fixed position in front of a camera for iris or retina identification. However, face recognition can be done passively without any explicit action or participation on the part of the user since face images can be acquired from a distance by a camera. This is particularly beneficial for security and surveillance purposes. Furthermore, data acquisition in general is fraught with problems for other biometrics: techniques that rely on hands and fingers can be rendered useless if the epidermis tissue is damaged in some way (i.e., bruised or cracked). Iris and retina identification require expensive equipment and are much too sensitive to any body motion. Voice recognition is susceptible to background noises in public places and auditory fluctuations on a phone line or tape recording. Signatures can be modified or forged. However, facial images can be easily obtained with a couple of inexpensive fixed cameras. Good face recognition algorithms and appropriate preprocessing of the images can compensate for noise and slight variations in orientation, scale, and illumination. Finally, technologies that require multiple individuals to use the same equipment to capture their biological characteristics potentially expose the user to the transmission of germs and impurities from other users. However, face recognition is totally non-intrusive and does not carry any such health risks. (Rabia, 2009).

Chapter 2. State of the Art

Face recognition is a rapidly growing area of research and innovation. Asides from much other applicability it is high in demand for security and police investigation purposes. Plastic surgery allowing criminals or evaders to hide their identities and reside in society soundly is an issue to be dealt with by the goon of technology. Hence, to overcome the challenge face recognition technology had to evolve and address this issue successfully.

Plastic surgery is a process of alteration or reconstruction of facial defects or improving the aspects like cleaning off the birthmarks, and pockmarks and adjusting the disfiguring defects. Although facial surgery is advantageous for some it can be misused by some who have indulged in some kind of crime and wants to conceal his/her identity. Many countries have incorporated facial data into the electronic passport with fingerprint information and iris information for security reasons. However, due to plastic surgery facial texture, shape and countenance can be altered and the security becomes unapt [9]. In all types of plastic surgery, there is one common thing that is all are gone through some kind of facial modification and diverge from the original. Plastic surgery is broadly divided into two types of classification.

A. Depend on the area of impact

Based on the percentage of surgical procedures, the classification of surgery is done. It depends on the modification of features and the degree of modification. Plastic surgery derived under this type of classification is of the following types:

1. Local surgery: This kind of surgery is used for correcting the skin texture, removing birthmarks, and correcting anomalies, aging effects, and accidental marks. Local surgery is also known as disease-correcting plastic surgery. It changes only some regions of face locally instead of the entire region globally.
2. Global surgery: This kind of surgery can be performed to alter facial texture entirely. This is also known as complete facelift surgery. In the situation of lethal flame and bruise, this type of surgery is mostly suggested.

B. Depend on the surgical approach

It is based on the process of surgery in which the person can do some kind of surgery according to need. Following are plastic surgery based on the surgical procedure.

• Eyelid surgery: It is basically used to correct defects and abnormalities of the eyelids. It remakes the eyelids which are overlapped the region of the eyebrow and upper chop area.
• Nose surgery: It is primarily used for modifying or reshaping of the nose. It is mainly preferred for decorative purposes like changing the shape of the nose according to the mouth to look good.
• Ear surgery: It is used for correcting the abnormality of Pima or ear. It can be used for decorative purposes or functional purposes.
• Lip augmentation: This is mainly used for beautifying purposes. It basically magnifies the form of the lips for better facial appearances.
• Skin peeling: Skin peeling is a global surgery and by using laser technology it regenerates the whole face or removes wrinkles. Using this type of surgery, the face gets changed entirely and gives the younger look [12].

Face recognition is a process of verifying or identifying a person from a video frame or image. A face recognition system is beneficial in the area of security mostly. The mirror points on the face on opposite sides of the central axis, pupillary separation, and some other basic features are mostly used for verifying the identity of a person. And then there are certain features that cannot be altered even after plastic surgery like the shape of the zygomatic bone, nodal points on the face, and the pupillary distance [8].

In 2010 Singh et al. observed the six existing face recognition algorithms and have shown that the performances of these algorithms are downgrading on the plastic surgery database [7]. The correlativeness between faces before and after plastic surgery was studied comprehensively by K. R. Singh et al. (2011). In that, they can classify the facial image using near set theory [1]. In 2011, Lakshiprabha et al. present an approach for face recognition using Gabor and LBP for feature extraction on the face region and eye region and achieved better accuracies [2]. Aruni Singh et al. done the contrast of various face verifying approaches on a dummy dataset of faces and they have shown the critical analysis of various algorithms on the same sets of data (2012) [3]. Face recognition presents a challenging problem in the field of image analysis and computer vision, and as such has received a great deal of attention over the last few years because of its many applications in various domains. Face recognition techniques can be broadly divided into three categories based on the face data acquisition methodology: methods that operate on intensity images; those that deal with video sequences; and those that require other sensory data such as 3D information or infra-red imagery.

Chapter 3. Aim of the Research

The main aim of the research work is to fully explore those challenges attached to facial recognition after plastic surgery and find out the possible solution.

Objectives of the Research

• To identify the reasons behind the lower rate of recognition when plastic surgery is performed on a given face and possible solutions.
• To develop an algorithm that will increase the recognition rate in facial recognition after plastic surgery.
• To highlight the area of feature research in facial recognition after plastic surgery.

The approach of the Research

The approach of Pankaj Dadure in 2018, will be studied and modified, the approach uses Local Binary Pattern(LBP) and Principal Component Analysis(PCA). Possibly when little modification is done in his work the recognition rate can be improved.

Research Methodology

In the proposed approach, LBP combined features the given pixel of LBP pattern is outlined as an ordered set of binary comparisons and using the following equation resulting value can be obtained [10][11]. Extracted from the face and periocular region to perform well and PCA algorithm is used for dimension reduction. Using a Euclidean distance classification is done. Features extracted from the face region are compared first and then features extracted from the periocular region are compared.

A. Data collection and preprocessing

For the recognition of face images, a plastic surgery database would be used [7]. In the plastic surgery database, it contains 1800 face images, 900 images are non-surgery face images and 900 are surgery face images. The images of faces have some noise and irregularities. So that the preprocessing operations like extracted features. So, here PCA is used to reduce the dimensionality. The extracted features are decumbent to noise. Using PCA, this problem can be minimized. The PCA algorithm has the following step:

1. Assemble a training set of images (M number of face images)
2. Resampled all the images to common pixel resolution

B. Local binary pattern

Local binary pattern (LBP) provides an efficient way for texture description. LBP is faster as compared to any other feature extraction algorithm. It is a non-parametric approach and very demanding in the domain of machine vision and image processing [8][9]. Consider a 3*3 pixel with (pC , qc) intensity value be Ic and local texture as L = l(t0, t1, t2, t3, t4, t5, t6, t7) where tn (n=0,1,2,3,4,5,6,7) corresponds to the grey values of the eight encompassing pixels. These encompassing pixels are threshold with the middle value tc as l(r(t0 – tc), – – – – – r(t7 – tc)) and therefore the r(x) is outlined.

Feature extraction using LBP

C. Principle component analysis (PCA)

PCA is used for feature extraction and reduces the dimensionality of the images. Which transfigure a number of associated pixel values into a number of disassociated pixel values called as Eigenfaces. It also calculates an optimized and compact description of the dataset. Sometimes the extracted features are large in size which leads to a memory problem, computation problem, and many more. PCA procedure is as follows.

1. Assemble a training set of images (M number of face images)
2. Resampled all the images to common pixel resolution R*C
3. Individual image is converted to a single vector or single row which contains R*C elements. i.e. [R1 R2 R3 – – – – – – RC]
4. The training set is then accumulated into a single matrix S which consists of M column. i.e.
5. Calculate mean image m using the following formula.
6. Subtract mean m from matrix T and from that matrix A (centroid image matrix) is obtained.
7. Calculate surrogate matrix S and covariance matrix C. Where C = A* A| and S = A| * A
8. Calculate eigenvalues and eigenvectors of surrogate matrix S.
9. Calculate eigenvalues and eigenvectors of the covariance matrix with the help of eigenvalues of S using the following formula. Vc = A * Vs and Uc = A * Us. Where Vc and Uc are eigenvectors and eigenvalues of the covariance matrix and Vs and Us are eigenvectors and eigenvalues of the surrogate matrix.
10. An eigenvector of the covariance matrix is eigenfaces.

D. Periocular region

The periocular region includes the iris, eyes, eyelids, eyelashes, and part of the eyebrows. Recognition using periocular biometrics is an emerging research area Eyelids, lateral canthus, medial canthus, lid folds, and surrounding area of the eyes are known as the periocular region and these points consider as discriminant in nature. There is no separate database is present for the periocular region. So that can be obtained from face images by cropping the periocular region. There are three different ways to perform periocular biometrics is overlapping, non-overlapping, and strip [2]. Using the periocular region a significant accuracy is obtained for face recognition.

Chapter 4. Current Work and Preliminary Results

The literature for the work was carried out, the plastic surgery database was obtained from IIIT New Delhi India. Now the research work is at the replication stage.

Chapter 5. Work Plan and Implications

Inline quantum of the work a time frame was allocated to each activity within which it will be executed step by step which is represented in the chart below.

Figure 1 Activity chart for the Work

Step 1 (Information gathering stage)

This will be the initial stage of the research, it involves gathering enough materials such as journals, write-ups, conference papers, review articles, textbooks, lecture notes, etc. that will enable the literature of the research work to have a solid foundation and to have deep inside of the work dataset gathering inclusive.

Step 2 (Replication stage)

At this stage the previous work by Pankaj Dadure., will be studied and the algorithm for the work will be replicated. The result of the work will be used to identify the shortcomings, however, facts will be established based on the result obtained. My research work will be carried out based on the limitations of the replicated result. A seminar paper will be presented.

Step 3 (Experimental stage)

This is where the work be done it will involve enhancing the algorithm of Pankaj Dadure., which will eliminate the drawback of the existing work and facts will be devoid on the result by comparing it side by side i.e the enhanced algorithm will run side by side with the replicated algorithm so that result will be compared. Another seminar paper will be presented.

Step 4 (Analysis stage)

At this stage, the two results will be analyzed and by so doing a good summary and conclusion could be derived based on the facts established from the results. However on the bases of analysis, the work it will show what has been added to the former work and the contributions to knowledge. Another seminar paper will be presented.

Chapter 6. Conclusions

The Conclusions would be restated based on the objectives of the work project, a recap of the research approach would also be stated, and clarified in a few words indicating my findings, why it is scientifically valuable to find it out would be answered, and on what basis would the work expect to evaluate the validity of the results.