Knowledge Connect @ SASTRA
Not a member yet
201 research outputs found
Sort by
AI in Healthcare: Early Diagnosis of Skin Cancer Using Medical Image Processing and Deep Neural Networks
No full textSeveral cancer types are commonly prevalent, and skin cancer is one among them, becoming even more widespread worldwide in the last few decades. To diagnose skin cancer at an early stage and obtain appropriate therapy to treat it, there is a demand to know more about the disease’s characteristics or severity. Skin cancer is caused mainly by various reasons, including damage of the sun or tanning beds by ultraviolet light exposure.
Failing to treat skin cancer might substantially impair an individual’s quality of life as the victim. They likely to experience physical issues linked with the deformities caused by psychological distress and the spreading of cancer cells to other healthy organs. Image processing and Artificial Intelligence (AI) play a crucial role in the early detection of melanoma skin cancer by analysing skin images for irregularities.
This technology helps in identifying potential cancerous lesions, leading to early diagnosis and potentially lifesaving treatment interventions. The ultimate goal of this research is to aid the melanoma skin cancer diagnosis in its early stage. Using the revolutionary AI algorithms and image processing techniques introduced the novel methods that will assist the clinical diagnosis with fair decisions.
Several artefacts are present, and the low image quality makes it challenging to evaluate the clinical features of malignant melanoma. Additionally, it is challenging to discern between benign lesions and malignant melanomas. This research initially proposed the MELIIGAN (Melanoma Information Improved Generative Adversarial Network) architecture to extract fine features from intermediate skin lesion images for early noninvasive diagnosis to overcome a few clinical diagnosis obstacles.
Undiagnosed skin lesions, also known as intermediate skin lesions with image enhancement. Secondly, proposed a new DDCNN-F (Double Decker Convolutional Neural Network – Feature fusion) framework for melanoma classification that includes a novel ‘F’ Flag feature for early detection. This novel ‘F’ indicator efficiently distinguishes benign skin lesions from malignant ones known as melanoma. Additionally, it deals with artefacts such as occluded hair. The scope of the study is extended to deal with multimodal images, and the wavelet method of extracting eight statistical features and seven distinct entropy features from a segmented image is utilised as an early diagnosis tool for a computer-aided diagnosis system.
Finally, the interpretable findings are used for multiclass skin malignancies to make oppressive choices about the clinical diagnosis of cancer in an early period by metadata ingestion in the proposed YOLOv7-XAI paradigm. The suggested research obtained a 9.1% improvement relative to the current state of the art in accurately classifying worrisome lesions. The proposed research shows a 93.75% accurate detection of melanoma and a 7.34% reduction in mistakes. Accuracy with multimodal image classification has been increased to 93.62%, and the multiclass classification model with explainability gave an accurate output of 96.89% and a precision of 94.60%
Role of Customized Visual and Multifactorial Interventions in Preventing Falls Among Individuals With Visual Impairment in South India
Get PDFFalls are the second leading cause of unintentional injuries and extensively studied among the elderly. Hence, intervention studies for falls prevention are restricted to the elderly with or without visual impairment (VI). Even among studies with VI, strategies for treatable VI are widely explored. Hence this study aimed to understand falls and interventions that may reduce falls in all the age groups especially among those with untreatable VI.
This mixed methods study included identification of subjects aged ≥ 18 years with VI from clinical records of Medical Research Foundation, Chennai and details of previous history of falls were enquired over telephone. In-depth interviews were conducted for a subset of subjects to understand perspectives of falls considering reduction of falls as behavior. Theoretical Domains Framework, Behavior Change Wheel and Socio- Ecological Model were used to analyse the results.
Effectiveness of proposed strategy was tested for subset of subjects with untreatable VI and history of falls. Falls rate was considered as primary outcome and fear of falling, visual reaction time, balance assessment, temporal gait parameters, obstacle foot placement were considered as secondary outcome measures. Out of 888 subjects, 18.0% (n=160) reported having fallen at least once and among them, 45.6% (n=73) had multiple falls.
Overall, fallers were more among ‘young’ (21.1%) than ‘middle-aged’ (16.0%) and ‘elderly’ (17.3%). Increased severity of VI, Side vision difficulty, diseases such retinitis pigmentosa and glaucoma showed increased risk of falls. Overall ‘Fear of falling’ was reported by 11.3% (n=100), out of which 30.0% (n=48) were by fallers. SMART intervention strategy (S-Support from caregivers/assistive devices, M-Monitoring of falls, A-Awareness about falls, R- Reduction of home hazards, T-Training-Training orientation and mobility, balance exercises and yoga therapy) was proposed from in-depth interviews which showed reduction of falls rate by 62% with incidence rate 0.381 (95% CI 0.240-0.606, p\u3c 0.001) for cases.
Fear of falling score, visual reaction time, indoor walking speed, indoor and outdoor walking steps per minutes improved in cases compared to controls after intervention. This mixed methods study concluded that the SMART strategy identified used behavior change models is effective in reducing falls among adults with untreatable VI in India
A Critical Study on the Effectiveness of Mediation as a Means for the Settlement of Matrimonial Disputes With Special Reference To Tamilnadu
Get PDFThe research works evaluates the effectiveness of Mediation as a means for the settlement of matrimonial disputes in the State of Tamilnadu. The research work travels through the essential features of Mediation and acceptability of Mediation as a tool for settlement of matrimonial disputes in the State of Tamilnadu. The research work elucidates the special features of matrimonial disputes and the social and psychological factors which are interplaying such disputes. The researcher through this work attempt to unfold the underlying problems in the application of Mediation in the settlement of matrimonial disputes with specific reference to the State of Tamilnadu. The researcher in her work found that the Indian legislature and judiciary encourages Mediation for resolving civil disputes in general and matrimonial disputes in particular.
In the research work such judicial pronouncements are being analysed and the positive attitude of Indian Judiciary is brought out. Indian family law has its roots in personal laws which are the offshoot from religious texts and commentaries. Hence, the researcher found meaningful to make an outline of the nature of matrimonial disputes in different personal laws in addition to Special Marriage Act, 1954. The researcher elucidates the elements supporting the ‘Mediation’ and ‘Conciliation’ in the above mentioned personal laws and thereby traced the evolution of Mediation as a tool for the settlement of matrimonial conflicts from the personal Laws.
Thereafter the researcher analysed the judicial interpretation of the specific provisions of the Family Court Act of 1984 and the Code of Civil Procedure of 1908 regarding Mediation for the settlement of matrimonial disputes. The researcher brings out the Judicial enthusiasm to encourage Mediation. However, the researcher felt that an empirical study of the success rate of the Mediation is worth for this work and would add its value.
The researcher confined their study on the effectiveness of the Mediation in settling the matrimonial causes in the State of Tamilnadu. While the State of Tamilnadu is being considered as Universe, the data from five districts are chosen on the basis of rural- urban distinction; is the sample from the universe. The data is being collected through different means which includes requisition from Family Courts and other Courts, application under Right to Information Act, 2005.
The primary issues of confidentiality and lack of legal obligation on the part of judicial forums and Mediation Centres, remained as hurdles for the researcher throughout her empirical study. However, the academic nature of the research work tempted the kindness of the Hon’ble Madras High Court in considering the request of the researcher in granting data from Family Courts in all the five districts namely Chennai, Coimbatore, Madurai, Tiruchirappalli and Thanjavur.
Mediation as a method is being encouraged by both the legislature and judiciary due to speedy disposal of the cases and win-win situation. In addition to that in matrimonial disputes involving multiple socio-psychological dimensions, an orthodox and stringent Court atmosphere is not considered suitable by the judiciary. However, in the empirical study conducted by the researcher in the five districts mentioned above less than 20% matrimonial causes are only settled through Mediation.
The researcher attempted to intrude into the reason for the failure of Mediation. The researcher reveals the following impasses through her work as reason for the failure. They are
1. Emotional Impasses
2. Substantial Impasses
3. Procedural Impasses
The empirical study and the testing of hypothesis along with analysis for the reasons for failure are being discussed in Chapter V of the research work. In the last chapter, the researcher elucidates her conclusions and recommendations
Motion Based Analysis of Ultrasound Imaging for the Study of Musculoskeletal Tissue Bio Mechanics
No full textUltrasound image analysis plays an important role in diagnosing musculoskeletal injuries and monitoring rehabilitation exercises. The first and foremost step in this analysis involves segmentation of region of interest from the ultrasound images. The segmentation of the musculoskeletal tissues from the ultrasound images is challenging due to the inherent drawback present in ultrasound like : (1) Poor image quality due to image corruption by speckle noise, shadows, and attenuation. (2) Dis-continuous boundaries due to orientation dependence during the acquisition of image. (3) Low contrast between nearby anatomical structures. Hence in order to overcome these drawbacks, there is a need for powerful segmentation algorithms for computerized segmentation of region of interest. The existing segmentation algorithms used for ultrasound images of musculoskeletal tissues are model based segmentation, machine learning and deep learning methods. Even though these methods are algorithmically distinctive they all rely on the same type of input information and try to distinguish the region of interest based on an analysis of signal intensity, texture, and shape features. However, in the case of ultrasound images such information might not be sufficient for boundaries. In such cases video analytics where the pattern of movement is also considered could enhance the analysis, however this approach has not as yet been tested in the case of medical imaging. Musculoskeletal tissues can be classified into two types, based on whether it is attached or detached from the surrounding tissues. When the musculoskeletal tissue is detached from the surrounding tissue, the tissue movement is simple and a simple pixel displacement information between a pair of frames is enough to segment the tissue based on movement information from the pair of frames. Example of such tissue is tendon. In the ultrasound video of tendon, tendon appears to move in the opposite direction to the surrounding tissues. This motion information is deployed for the segmentation of tendon from the surrounding tissues. The algorithm used for computing displacement between pair of frames are NCC and optical flow. The segmented output from the computerized segmentation is compared with the manual segmentation for calculation of accuracy. The algorithm having highest accuracy is considered as an accurate method for segmentation of tendon. When the musculoskeletal tissue is attached to the surrounding tissue, the movement information is complex and a simple displacement information from a pair of frames is not enough for tissue segmentation. Strain and average displacement are calculated from the multiple frames using which map of movement is created for these tissues. The map of movement from strain and average displacement helps in tissue segmentation. Example of such tissues is bone. The regions are tracked in the multiple frames using NCC and optical flow algorithm and map of movement is created from the motion information from these multiple frames. From the map of movement bone is having the highest movement compared to the surrounding tissues. This motion information is deployed for the segmentation of the bone from the surrounding tissues. The computerized segmented output are compared with the manually segmented output for determination of accuracy. The algorithm having highest accuracy is considered as an accurate method for segmentation of bone. From the segmentation of tendon and bone, algorithm having the highest accuracy is optical flow algorithm
Fecal Genomics Provides Insights On Connectivity Of Endangered Wildlife In A Human Dominated Landscape
No full textMany mammalian species, facing severe historic range reductions, now inhabit fragmented habitats. Habitat fragmentation involves the reduction of habitat into smaller patches, increasing distances between them, and eventually leading to small isolated populations. Species\u27 responses to habitat alteration may vary based on factors such as body size, dispersal ability, and abundance. Understanding these differences is crucial for effective multispecies conservation planning. Genomic data, within a landscape genetic framework, are increasingly applied to evaluate the effects of habitat fragmentation and understand influential landscape features for connectivity. The quantity of data enhances the power and precision of assessing landscape impacts on gene flow. Recent advances in next-generation sequencing (NGS) make harnessing genomic data increasingly viable, extending its application to non-model wild species through non-invasive sampling methods. However, non-invasive samples pose challenges related to DNA quality and quantity, which is often compromised.
In this study, I developed methods enabling the use of fecal DNA to generate genome-wide data for endangered species. Using the Jungle cat as a model system, we attempted and achieved substantial enhancement in the proportion of host DNA. We demonstrate that using a higher number of markers provides more power than using a high number of individuals. Using the same standardized enrichment method, I attempted to understand the genetic connectivity of two large herbivores, gaur and sambar, to comprehend how different species, even with similar habitat associations, respond to various landscape features. We found that both gaur and sambar exhibit population structure in central India and are negatively impacted by roads and land-use changes. We also observed species-specific responses to various landscape variables, highlighting small and isolated populations for both species that require conservation intervention. In the process, we developed a double-digest restriction-associated DNA sequencing (ddRAD) protocol to suit low input host DNA concentration in generating empirical genomic data. Finally, I applied these methods to elucidate fine-scale multispecies connectivity for eight different species (Gaur, Sambar, Muntjac, Jungle cat, Dhole, Tiger, Leopard, and Sloth bear) and examined how different species, based on their body size, trophic level, social structure, abundance, and dispersal distance, are impacted by habitat fragmentation. We found no general pattern of resistance based on body size. However, landscape features offer higher resistance to herbivore movement compared to carnivores, and social species are more impacted by various landscape features compared to solitary species.
This study represents the first attempt where non-invasive samples were used to generate genome-wide data for several endangered species, investigating multi-species patterns of connectivity in a human-dominated landscape. The enrichment method applied to a range of endangered species is able to elucidate adequate genome-wide data for investigating fine-scale patterns of connectivity. Furthermore, multi-species patterns reveal that while some landscape elements affect all species, others appear to be species-specific. Incorporating such knowledge into conservation planning could potentially reduce local extinction risks and, more broadly, slow down ongoing biodiversity loss
Mathematical Concepts and Coding Theory in Kumārasambhavam of Kālidāsa
No full textKālidāsa, a renowned Sanskrit poet and playwright from ancient India, is known for his profound understanding and depiction of various aspects of human life, nature, and the universe in his works. While his writings primarily revolve around themes of love, romance, and human emotions, there are several instances where he incorporates mathematical concepts and astronomical references.
His meticulous usage of such abstractions and symbolism adds depth and richness to his poetry. These elements serve as tools to enhance the imagery and aesthetics of his literary works, reflecting the interplay between art and science in ancient Indian culture. These subtle allusions in his works demonstrate the interconnectedness of different disciplines and the universal appeal of mathematical concepts.
They showcase his appreciation for numerical symbolism, symmetry, geometry, proportions, and time calculations. His poetic descriptions and metaphors often draw upon mathematical postulations to enhance the beauty and imagery within his writings.All these prove his authority in diverse scientific disciplines.
This thesis brings to light, the hidden treasures of the Mathematical Concepts including Astronomy and Astrology in Kumārasambhavam of Kālidāsa, their interconnectedness with the earthly realm and their relevance to our day- to-day life, as they remain unexplored. An earnest attempt is made to apply the standard decoding techniques to the verses to unravel many results from the various branches of Mathematics like Arithmetic, Number theory, Decimals and Fractions, Prime numbers and its properties, Decimal and Hexa decimal Number system and others
Investigation on the Impact of Loading Effect of Fruit Juices on the Performance of Pulsed Electric Field Generators
No full textPulsed Electric Field (PEF) treatment is one of the efficient non-thermal food processing techniques which is being preferred as a replacement for thermal pasteurization methods. The effectiveness of PEF treatment was measured in terms of reduction in the microbial load in the food, extension of shelf life of the food and retention of nutritional properties of the food. The successful implementation of the PEF treatment depends upon various aspects such as design of pulse generator, parameters of pulse generator, shape and size of the treatment chamber and most importantly the characteristics of each food items. Design and fabrication of a suitable pulse generator for PEF treatment is mainly governed by the type and volume of food to be treated. For the processing of food items using PEF technique, different shapes of pulse voltages such as square pulse, exponential pulse and oscillatory pulse voltages were preferred.
However, each type of food offers different loading effect to the pulse generator due to its varying impedance based on the type of food and volume of food to be treated. Identifying suitable type of pulse shape and pulse generation circuit to get optimum treatment for a given food type is an important task. In this work, the main focus was to analyze the impact of different volumes of juice samples on the performance of pulse generator circuit. The performance of pulse generator was investigated by analyzing the loading effect of juice samples on pulse generator as a simulation study, followed by numerical computation to optimize the key process parameter and validation of results using a hardware circuit.
In order to analyze the loading effect, different shapes of pulse voltages such as square pulse, exponential pulse and oscillatory pulse voltages were generated using power electronic converter circuits and the same was applied to different volumes of juice samples. The juice samples were considered as an R-C equivalent circuit and the equivalent circuit was connected as a load to pulse. Two distinctly dimensioned test chambers with parallel plate electrodes were fabricated for different volumes ranging from 1 ml to 500 ml.
The generated pulse voltages were applied to the juice samples and response of the pulse generator were measured in-terms of voltage, electric field intensity, pulse power consumption and specific energy dissipation. The loading effect were analyzed for both the test chambers (named, Test Chamber 1 & Test Chamber 2) dimensions. From the analysis, it was observed that the change in pulse output voltage magnitude and the change in pulse shape with respect to change in juice volumes is evident for test chamber 1. In addition, the electric field intensity decreases with the increase in volume.
In order to get the appreciable field intensity between the electrodes of test chamber, the dimensions of the first test chamber were modified and the new R-C parallel branch model was considered as a load across the PEF generator circuits. It was evident from the results of test chamber 2 that the field intensity obtained for all the volumes were higher when compared to test chamber 1. Particularly, the results of bipolar square pulse generator were higher compared to other generator circuits. Though the electric field intensity applied to the sample was appreciable, there is a reduction in the voltage magnitude with increase in sample volume.
Among all the process parameters, specific energy dissipation was the key process parameter which results in maximum log reduction. To optimize the effect of treatment chamber design and its volumes on the key process parameter [specific energy dissipation], Response surface methodology with the Box-Behnken design was applied. For this analysis, the features were selected with the typical range of distance, voltage, conductivity, pulse duration and volume. The corresponding response of specific energy dissipation was obtained.
From the results, it was clear that most significant effects were observed from the gap distance between electrodes when compared to other parameters. Most importantly, there was no significant effect on response due to the volume with other parameters. It was found from the results that to achieve effective microbial inactivation, the gap spacing needs to be maintained less and the same should remain unchanged with the increase in the volume of the treatment chamber.
For the validation of loading effect analysis and for the treatment of fruit juice, a high voltage bipolar square pulse generator was fabricated and the generated 1 kV pulse was applied across the different volumes from 1 ml, 10 ml and 100 ml of apple juice sample. From the treatment results, the microbial load in the apple juice gets reduced and the inactivation efficiency of 71.5% was achieved with a log reduction of 4.8
Generative AI-Based Optimized Recommender System for Debt Collection Using Large Language Models
No full textReducing the percentage of defaulters who often skip payments throughout the debt collection process might help minimize losses in the banking industry. The debt collection process should be optimized to reduce the rate of defaulters and improve collection rates. Traditional Machine Learning algorithms focused on credit risk analysis, defaulter prediction, and forecasting the recovery rate of debt collection. Researchers are not currently prioritizing the analysis of debt collectors’ performance. The debt collector’s primary responsibility is to retrieve outstanding debts from consumers on behalf of the debt collection firm.
Examining debt collectors’ performance is essential to enhance collection efficiency in the dynamic debt collection process. This research uses the continual learning approach to categorize the debt collector’s performance using the Enhanced Deep Q Network with Continual Learning (EDQN-CL). These techniques adapt to evolving data patterns over time and incrementally update new performance categories as needed. The metrics used for analyzing the debt collectors’ performance such as collection percentage, and visit patterns (date, time of visit, time interval between visits, and number of visits). The proposed algorithm achieves 13.56% higher accuracy than the existing system for employee performance.
If the customer reaches higher levels of overdue, then the customer will be allocated to the external collection agency. The proposed recommendation systems prevent the customer from being transferred to an external agency by preventing the customers from reaching higher levels of overdue. This thesis systematically optimized debt collection management by evaluating the debt collectors’ performance and categorizing the credit risk analysis.
A flexible recommendation system built using a hybrid actor-critic algorithm with transfer learning to enhance the debt collection agency’s collection rate. To adapt to these dynamic changes in the customer risk category, the proposed Reinforcement Learning (RL) methodology creates a customer allocation file that matches the risk of the customer with the appropriate debt collector. Moreover, it suggested recommendations such as visit count, actual statements, date, time of visits, and the time interval between the visits. After implementing this FRS-DRL for six months in a real-world situation, the collection rate increased to 20.34% and the number of visits decreased to 16.30%.
Text generation has advanced with the emergence of Large Language Models (LLM) in the field of artificial intelligence. This research explores the use of prompt engineering to enhance the text-based explainable AI of Debt Collection Management (XAI-DCM) using LLM. To get the intended results, several prompting techniques were used, including multiple-shot and zero-shot prompting. Conducted the comparative analysis between the existing system with LLM-based explanation generation. The proposed XAI-DCM model achieves a 16.03% higher BLEU score and a 12.80% ROUGE score than the existing knowledge graph systems.
The implementation of intelligent chatbots for debt collecting has the potential to significantly enhance the success rate of the debt recovery procedure. This research focuses on the development of chatbots by leveraging different prompting techniques. Occasionally LLM can generate some undesired or irrelevant response, called a hallucinated response. Additionally, this research investigated hallucinations through the use of Parameter Efficient Fine Tuning (PEFT) techniques, including Low-Rank adaptation (LoRa). This model sets automatic payment reminders regarding the due date and, based on the customer response, offers flexible payment, understanding the financial situation of the customers.
Moreover, Demonstrated the effectiveness of prompt engineering through a comparative analysis between two LLM such as GPT 3.5 and Llama 2 7B chat model in terms of ROUGE and BLEU score. In Zero-shot prompting, the Llama chatbot achieves a 0.46% higher BLEU-4 score. Whereas in Multiple shot prompting, the GPT 3.5 model achieves 42% improvement in the BLEU-4 score. This thesis made the comparative analysis with the existing work chatbot system, and the proposed DCM chatbot model achieves a higher BLEU-4 score of 1.88% for Kaggle’s chatbot dataset.
The fine-tuned model using LoRa achieves a 20.77% improvement in the BLEU - 4 score. Figure 1 gives the overall debt collection optimization process. The proposed FRS - DRL system provides the optimal recommendations without human intervention and increases the collection rate by 20.34% while reducing visits by 16.30%. Debt collectors achieved a higher collection rate when they adhered closely to the recommendations provided by the FRS-DRL
High Power Capacitive Wireless Power Transfer System for Future Transportation Electrification
No full textWireless Power Transfer (WPT) technologies, including Inductive Power Transfer (IPT) and Capacitive Power Transfer (CPT), offer promising solutions for Electric Vehicle (EV) charging applications. Compared to conventional IPT, a CPT method has no eddy current losses, lower electromagnetic interference (EMI), low system cost and weight, less misalignment problems, which gives the motivation to start this research work. The Proposed CPT method uses aluminum plates that generate electric fields by applying voltage on the plates to transfer the power from the transmitter (ground side) to the receiver (vehicle side).
This research introduces novel advancements in CPT systems to enhance power transfer efficiency and reliability for EV charging. A key focus is the Static Capacitive Power Transfer (SCPT) system, where a novel vertical six-plate arrangement is proposed to improve mutual capacitance, power transfer level, and efficiency. The vertical configuration of this arrangement reduces the size and weight of the coupler, offering advantages over conventional horizontal arrangements.
The transmitter side comprises four aluminium plates, while the remaining two plates, featuring mica, are strategically placed on the receiver side. Finite Element Analysis (FEA) using ANSYS Maxwell simulation is employed to determine plate dimensions and coupling capacitances, followed by LTspice simulations to validate the circuit model. Experimental verification at a transmitter distance of 300 mm demonstrates impressive results, with an output power of 6.06 kW and an efficiency of 92.3% achieved at a 60 mm air gap, contributing significantly to SCPT system advancement.
Additionally, this research proposes a novel vertical Dynamic Capacitive Power Transfer (DCPT) system for rail transit and on-road charging, aiming to enhance power transfer efficiency and CPT performance. The system utilizes a unique four-plate arrangement with optimized coupling capacitances and dielectric materials to improve output power. Experimental validation shows promising results, with an output power of 2.76 kW and a maximum efficiency of 92.6% achieved at a frequency of 1 MHz, highlighting the potential of DCPT for practical implementation in dynamic charging scenarios.
The presence of foreign objects, such as metallic ones, liquids, and other conductive materials, can be dangerous and reduce the effectiveness of the CPT system. For the widescale adoption of CPT, it is essential to develop an accurate and trustworthy Foreign Object Detection (FOD) system. Therefore, in addition this research proposes a FOD system based on deep learning techniques, designed to accurately identify foreign objects in CPT systems. The implementation of such a system holds the promise of enhancing the safety and reliability of CPT systems, contributing to its broader adoption in the realm of EVs. Overall, this research contributes to the advancement of CPT systems for EV charging, offering improvements in performance, efficiency, and safety, and holding promise for practical implementation in real-world applications
Design of an Integrated System for the Protection of Patient Health Information in Medical Images
No full textElectronic clinical data such as Patient Health Information (PHI) is stored by adapting the Digital Imaging and Communications in Medicine (DICOM) standard to build an integrated healthcare system where medical images from various sources can be interlinked. The healthcare industry poses a threat from hackers, and the information fetched by hackers provides more money than the other information stolen. This thesis deals with designing and analysing encryption algorithms, key generation mechanisms, and information-hiding schemes for protecting PHI.
Sensitive multimedia information of all forms is encrypted, with a key, before storage and transmission to protect it from illegal use and manipulation of data. Since digital images are more significant, effectively encrypting the content, specifically medical images, is crucial. Even a little data manipulation in medical image diagnosis may lead to misinterpretation. For this, an algorithm is devised to be suitable for encrypting DICOM and other images.
With the onset of any pandemic, the medical image database is bound to increase. These medical images are prone to attack by hackers for their medical data and PHI. To safeguard these medical images, a new algorithm is proposed. The algorithm involves secretly embedding the patient identification number into the medical image and encrypting the medical image, protecting the patient\u27s identity and medical condition from hackers. The encryption algorithm involved a single stage of confusion and two stages of diffusion. The confusion operation was performed using the key generated by the Bülban map. The first stage of diffusion was done in the transform domain, using 5/3 transformation, and the second stage of diffusion was performed in the spatial domain by altering the pixel values using the key.
Transferring multimedia content without detection is gaining traction over a while. Using Steganography to hide secret data, viz. passwords, medical information, and secret messages, has become the norm. This work involves hiding confidential information in the images. It aims to increase the payload capacity by adaptively choosing the embedding pixel location, using hamming distance, and increasing the number of bits embedded per pixel using threshold while retaining the image quality. The embedding of secret information is random, using a key, generated using a Combined Logistical Tent Map, governed by equations and reordered using a Cartesian product generator. The algorithm is designed to perform embedding for images of various modalities, viz., Grayscale, RGB and DICOM.
The impact of hacking into medical images for PHI reverberates through the hospitals and their services. The need for PHI protection gives rise to the requirement of a practical algorithm for hiding the PHI within the medical image and recovering both PHI and the cover medical image. In this work, an adequate reversible data hiding using Bicubic Spline Interpolation, pixel value differencing, and adaptive data hiding is designed