Description:FORM 2 THE PATENTS ACT, 1970 (39 of 1970) & THE PATENT RULES, 2003 COMPLETE SPECIFICATION (See Section 10 and Rule 13) Title of invention: SYSTEMS AND METHODS FOR IN BODY MICROWAVE IMAGING OF A SUBJECT Applicant: Tata Consultancy Services Limited A company Incorporated in India under the Companies Act, 1956 Having address: Nirmal Building, 9th Floor, Nariman Point, Mumbai 400021, Maharashtra, India The following specification particularly describes the invention and the manner in which it is to be performed. TECHNICAL FIELD The disclosure herein generally relates to microwave imaging (MWI) techniques, and, more particularly, to systems and methods for in body microwave imaging of a subject. BACKGROUND Cancer (e.g., breast cancer) is one of the most prevalent diseases in the world (e.g., predominantly occurring in women). Detecting cancer early can significantly reduce the mortality rate, but this still remains a challenge owing to shortcomings in early screening and detection with existing modalities. Detection of cancer is typically done using screening methods such as X-ray mammography, Magnetic Resonance Imaging (MRI) and Ultrasound imaging (US), out of which X-ray mammography is considered as a standard detection method. But these conventional methods have their own limitations such as compression discomfort, inherent health risks, expensive, and consumes more time and effort. SUMMARY Embodiments of the present disclosure present technological improvements as solutions to one or more of the above-mentioned technical problems recognized by the inventors in conventional systems. For example, in one aspect, there is provided a processor implemented method for an in-body microwave (MW) imaging of a subject. The method comprises obtaining, via one or more hardware processors, an initial coarse microwave (MW) data pertaining to a set of predefined uniform locations scanned around a specific portion of a body of a subject; sequentially recommending, by using a trained reinforcement learning (RL) agent, via the one or more hardware processors, one or more subsequent locations pertaining to the specific portion of the body of the subject using the initial coarse microwave data to obtain a set of subsequent MW data; vectorizing and stacking, via the one or more hardware processors, the initial coarse microwave data and the set of subsequent MW data to obtain a measurement matrix; estimating, via the one or more hardware processors, a Kernel K using a reference point scatterer of dielectric value e_0 and an area A_r at a set of reference coordinate placed in a region of interest, wherein the Kernel K estimated at the set of reference coordinate is used to estimate the Kernel K at remaining coordinates in the region of interest based on an Euclidean distance between the set of reference coordinate and the remaining coordinates; and estimating, via the one or more hardware processors, an unknown dielectric constant of the region of interest in the specific portion using an inverse model formulation obtained from the measurement matrix and the Kernel K, wherein the unknown dielectric constant indicates a degree of severity of a disease associated with the region of interest in the specific portion. In an embodiment, the trained reinforcement learning (RL) agent is obtained by: defining a state set, an action set, an environment, and a reward pertaining to acquisition of MW data, wherein the state set is defined by a set of reconstructed microwave (MW) images that is used by the RL agent to select an action, wherein an initial state obtained from an initial coarse MW data enables the RL agent to select an optimal action from the action set, wherein the action set comprises a set of positions from which the set of subsequent MW data is acquired, wherein MW data pertaining to a specific action is collected based on a location being recommended, the environment reconstructs a next state of the region of interest based on the MW data collected from a set of cumulative actions selected by the RL agent, wherein a reward is computed by comparing a current state with a ground truth state obtained by using an entire MW data obtained from a plurality of possible acquisition positions, and wherein the current state, the action, the next state, and the reward constitutes a replay memory buffer data being stored in the replay memory buffer; randomly sampling data amongst the replay memory buffer data stored in a replay memory buffer to obtain sampled data; computing a mean square error (MSE) loss using the randomly sampled data based on a pre-defined equation; and training one or more network parameters of a Double Deep Q Network (DDQN), wherein the DDQN serves as the trained RL agent, wherein during the training, the MSE loss serves as feedback to the RL agent for a given-state-action pair; and deploying the trained RL agent in the environment for scanning to obtain one or more optimal action indices required for estimating the unknown dielectric constant. In an embodiment, the step of training the one or more network parameters of the DDQN is preceded by: receiving a MW image; generating a value function for one or more action sets based on the MW image; and selecting at least one action from the one or more action sets based on an associated maximum value function. In an embodiment, a dielectric constant (value) and an area of a first type of cells (normal cells) and a second type of cells (area of abnormal cells is sparse in nature) in the region of interest are different from each other. In an embodiment, the unknown dielectric constant of the region of interest estimated uses a sparsity constraint which is based on the dielectric constant and the area. In an embodiment, the unknown dielectric constant of the region of interest is estimated using an iterative threshold technique. In an embodiment, number of MW data to be acquired for an episode is based on a scanning duration, and the scanning duration is one of a pre-determined duration or an empirically determined duration. In an embodiment, the region of interest comprises a tissue. In another aspect, there is provided a processor implemented system for in body microwave (MW) imaging of a subject. The system comprises: a memory storing instructions; one or more communication interfaces; and one or more hardware processors coupled to the memory via the one or more communication interfaces, wherein the one or more hardware processors are configured by the instructions to: obtain an initial coarse microwave (MW) data pertaining to a set of predefined uniform locations scanned around a specific portion of a body of a subject; sequentially recommend, by using a trained reinforcement learning (RL) agent, one or more subsequent locations pertaining to the specific portion of the body of the subject using the initial coarse microwave data to obtain a set of subsequent MW data; vectorize and stack the initial coarse microwave data and the set of subsequent MW data to obtain a measurement matrix; estimate a Kernel K using a reference point scatterer of dielectric value e_0 and an area A_r at a set of reference coordinate placed in a region of interest, wherein the Kernel K estimated at the set of reference coordinate is used to estimate the Kernel K at remaining coordinates in the region of interest based on an Euclidean distance between the set of reference coordinate and the remaining coordinates; and estimate an unknown dielectric constant of the region of interest in the specific portion using an inverse model formulation obtained from the measurement matrix and the Kernel K, wherein the unknown dielectric constant indicates a degree of severity of a disease associated with the region of interest in the specific portion. In an embodiment, the trained reinforcement learning (RL) agent is obtained by: defining a state set, an action set, an environment, and a reward pertaining to acquisition of MW data, wherein the state set is defined by a set of reconstructed microwave (MW) images that is used by the RL agent to select an action, wherein an initial state obtained from an initial coarse MW data enables the RL agent to select an optimal action from the action set, wherein the action set comprises a set of positions from which the set of subsequent MW data is acquired, wherein MW data pertaining to a specific action is collected based on a location being recommended, the environment reconstructs a next state of the region of interest based on the MW data collected from a set of cumulative actions selected by the RL agent, wherein a reward is computed by comparing a current state with a ground truth state obtained by using an entire MW data obtained from a plurality of possible acquisition positions, and wherein the current state, the action, the next state, and the reward constitutes a replay memory buffer data being stored in the replay memory buffer; randomly sampling data amongst the replay memory buffer data stored in a replay memory buffer to obtain sampled data; computing a mean square error (MSE) loss using the randomly sampled data based on a pre-defined equation; and training one or more network parameters of a Double Deep Q Network (DDQN), wherein the DDQN serves as the trained RL agent, wherein during the training, the MSE loss serves as feedback to the RL agent for a given-state-action pair; and deploying the trained RL agent in the environment for scanning to obtain one or more optimal action indices required for estimating the unknown dielectric constant. In an embodiment, prior to training the one or more network parameters of the DDQN, the one or more hardware processors are configured by the instructions to: receive a MW image; generate a value function for one or more action sets based on the MW image; and select at least one action from the one or more action sets based on an associated maximum value function. In an embodiment, a dielectric constant (value) and an area of a first type of cells (normal cells) and a second type of cells (area of abnormal cells is sparse in nature) in the region of interest are different from each other. In an embodiment, the unknown dielectric constant of the region of interest estimated uses a sparsity constraint which is based on the dielectric constant and the area. In an embodiment, the unknown dielectric constant of the region of interest is estimated using an iterative threshold technique. In an embodiment, number of MW data to be acquired for an episode is based on a scanning duration, and the scanning duration is one of a pre-determined duration or an empirically determined duration. In an embodiment, the region of interest comprises a tissue. In yet another aspect, there are provided one or more non-transitory machine-readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause an in-body microwave (MW) imaging of a subject by obtaining an initial coarse microwave (MW) data pertaining to a set of predefined uniform locations scanned around a specific portion of a body of a subject; sequentially recommending, by using a trained reinforcement learning (RL) agent, one or more subsequent locations pertaining to the specific portion of the body of the subject using the initial coarse microwave data to obtain a set of subsequent MW data; vectorizing and stacking the initial coarse microwave data and the set of subsequent MW data to obtain a measurement matrix; estimating a Kernel K using a reference point scatterer of dielectric value e_0 and an area A_r at a set of reference coordinate placed in a region of interest, wherein the Kernel K estimated at the set of reference coordinate is used to estimate the Kernel K at remaining coordinates in the region of interest based on an Euclidean distance between the set of reference coordinate and the remaining coordinates; and estimating an unknown dielectric constant of the region of interest in the specific portion using an inverse model formulation obtained from the measurement matrix and the Kernel K, wherein the unknown dielectric constant indicates a degree of severity of a disease associated with the region of interest in the specific portion. In an embodiment, the trained reinforcement learning (RL) agent is obtained by: defining a state set, an action set, an environment, and a reward pertaining to acquisition of MW data, wherein the state set is defined by a set of reconstructed microwave (MW) images that is used by the RL agent to select an action, wherein an initial state obtained from an initial coarse MW data enables the RL agent to select an optimal action from the action set, wherein the action set comprises a set of positions from which the set of subsequent MW data is acquired, wherein MW data pertaining to a specific action is collected based on a location being recommended, the environment reconstructs a next state of the region of interest based on the MW data collected from a set of cumulative actions selected by the RL agent, wherein a reward is computed by comparing a current state with a ground truth state obtained by using an entire MW data obtained from a plurality of possible acquisition positions, and wherein the current state, the action, the next state, and the reward constitutes a replay memory buffer data being stored in the replay memory buffer; randomly sampling data amongst the replay memory buffer data stored in a replay memory buffer to obtain sampled data; computing a mean square error (MSE) loss using the randomly sampled data based on a pre-defined equation; and training one or more network parameters of a Double Deep Q Network (DDQN), wherein the DDQN serves as the trained RL agent, wherein during the training, the MSE loss serves as feedback to the RL agent for a given-state-action pair; and deploying the trained RL agent in the environment for scanning to obtain one or more optimal action indices required for estimating the unknown dielectric constant. In an embodiment, the step of training the one or more network parameters of the DDQN is preceded by: receiving a MW image; generating a value function for one or more action sets based on the MW image; and selecting at least one action from the one or more action sets based on an associated maximum value function. In an embodiment, a dielectric constant (value) and an area of a first type of cells (normal cells) and a second type of cells (area of abnormal cells is sparse in nature) in the region of interest are different from each other. In an embodiment, the unknown dielectric constant of the region of interest estimated uses a sparsity constraint which is based on the dielectric constant and the area. In an embodiment, the unknown dielectric constant of the region of interest is estimated using an iterative threshold technique. In an embodiment, number of MW data to be acquired for an episode is based on a scanning duration, and the scanning duration is one of a pre-determined duration or an empirically determined duration. In an embodiment, the region of interest comprises a tissue. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed. BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are incorporated in and constitute a part of this disclosure, illustrate exemplary embodiments and, together with the description, serve to explain the disclosed principles: FIG. 1 depicts an exemplary system for an in-body microwave (MW) imaging of a subject, in accordance with an embodiment of the present disclosure. FIGS. 2(a) and 2(b) (collectively referred as FIG. 2) depicts (a) an illustration of a subject undergoing detection test, and (b) a detailed view demonstrating the N_a optimized locations selected by the RL agent from a total of N possible radar locations around the breast tissue, in accordance with an embodiment of the present disclosure. FIG. 3 depicts an exemplary flow chart illustrating a method for an in-body microwave (MW) imaging of a subject, using the system of FIG. 1, in accordance with an embodiment of the present disclosure. FIG. 4 depicts a block diagram of a deep reinforcement learning (DRL or RL agent) comprised in the system of FIG. 1, in accordance with an embodiment of the present disclosure. FIG. 5 depicts a Double Deep Q Network (DDQN) value architecture for training a reinforcement learning (RL) agent, in accordance with an embodiment of the present disclosure. FIG. 6 shows a reconstructed image using different algorithms, in accordance with an embodiment of the present disclosure. FIGS. 7A through 7C (collectively referred as FIG. 7) show visual MWI obtained using all 72 measurements, using only 24 uniform spaced fixed measurements and the method of the present disclosure using reinforcement learning (RL) acquisition with N_a=24 respectively, in accordance with an embodiment of the present disclosure. FIG. 8 shows visual comparison of the method of the present disclosure against Delay-And-Sum-Deep RL (DAS-DRL) and Delay-Multiply-and-Sum-Deep RL (DMAS-DRL), in accordance with an embodiment of the present disclosure. DETAILED DESCRIPTION OF EMBODIMENTS Exemplary embodiments are described with reference to the accompanying drawings. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. Wherever convenient, the same reference numbers are used throughout the drawings to refer to the same or like parts. While examples and features of disclosed principles are described herein, modifications, adaptations, and other implementations are possible without departing from the scope of the disclosed embodiments. Cancer (e.g., breast cancer) is one of the most prevalent diseases in the world (e.g., predominantly occurring in women). Detecting breast cancer early can significantly reduce the mortality rate, but this still remains a challenge owing to shortcomings in early screening and detection with existing modalities. Detection of breast cancer is typically done using screening methods such as X-ray mammography, Magnetic Resonance Imaging (MRI) and Ultrasound imaging (US), out of which X-ray mammography is considered as a standard detection method. But these conventional methods have their own limitations such as compression discomfort, inherent health risks, expensive, and consumes more time and effort (e.g., refer “N. AlSawaftah, S. El-Abed, S. Dhou, and A. Zakaria, “Microwave imaging for early breast cancer detection: Current state, challenges, and future directions,” Journal of Imaging, vol. 8, no. 5, p. 123, 2022.”). Recently, Microwave Imaging (MWI) based techniques which can overcome some of the above-mentioned limitations have been explored in literature (e.g., refer N. AlSawaftah et. al). MWI relies on the change in electrical properties when excited with electromagnetic waves. It has been observed that tumor cells have more water content as compared to normal cells and hence have higher dielectric properties of around 8-10% more than the normal cells (e.g., refer “T. Sugitani, S.-i. Kubota, S.-i. Kuroki, K. Sogo, K. Arihiro, M. Okada, T. Kadoya, M. Hide, M. Oda, and T. Kikkawa, “Complex permittivities of breast tumor tissues obtained from cancer surgeries,” Applied Physics Letters, vol. 104, no. 25, p. 253702, 2014.”). MWI is based on the principle of radar that excites electromagnetic waves and reflections from the breast are captured at different predefined locations. Further, these collected measurements are processed using various algorithms to reconstruct the MWI of the breast. Another literature (e.g., refer “X. Li and S. Hagness, “A confocal microwave imaging algorithm for breast cancer detection,” IEEE Microwave and Wireless Components Letters, vol. 11, no. 3, pp. 130–132, 2001.”) used Delay-And-Sum (DAS) which makes use of shifted time delay at different antenna positions. This is a fast and effective technique to reconstruct the image but results in significant clutter artifacts (e.g., refer “N. AlSawaftah, S. El-Abed, S. Dhou, and A. Zakaria, “Microwave imaging for early breast cancer detection: Current state, challenges, and future directions,” Journal of Imaging, vol. 8, no. 5, p. 123, 2022”). Therefore, various improvements have been made on DAS resulting in different variants like Delay-Multiply-And-Sum (DMAS) (e.g., refer “H. B. Lim, N. T. T. Nhung, E.-P. Li, and N. D. Thang, “Confocal microwave imaging for breast cancer detection: Delay-multiply-and sum image reconstruction algorithm,” IEEE Transactions on Biomedical Engineering, vol. 55, no. 6, pp. 1697–1704, 2008”), Improved Delay-And-Sum (IDAS) (e.g., refer “M. Klemm, I. Craddock, J. Leendertz, A. Preece, and R. Benjamin, “Improved delay-and-sum beamforming algorithm for breast cancer detection,” International Journal of Antennas and Propagation, vol. 2008, 2008.”), etc. An evaluation of these algorithms on clinical patients can be found in yet another conventional literature (e.g., refer “M. A. Elahi, D. O’Loughlin, B. R. Lavoie, M. Glavin, E. Jones, E. C. Fear, and M. O’Halloran, “Evaluation of image reconstruction algorithms for confocal microwave imaging: Application to patient data,” Sensors, vol. 18, no. 6, p. 1678, 2018.”), where only DAS and DMAS were consistent with clinical reports, with DMAS having significantly reduced clutter. Further, these techniques require dense radar measurements to obtain a good quality MWI (e.g., refer Li et. al, Lim et. al, Klemm et. al, “M. A. Elahi, D. O’Loughlin, B. R. Lavoie, M. Glavin, E. Jones, E. C. Fear, and M. O’Halloran, “Evaluation of image reconstruction algorithms for confocal microwave imaging: Application to patient data,” Sensors, vol. 18, no. 6, p. 1678, 2018.”, “N. K. Nikolova, Introduction to microwave imaging. Cambridge University Press, 2017.”, and “D. Tajik, F. Foroutan, D. S. Shumakov, A. D. Pitcher, and N. K. Nikolova, “Real-time microwave imaging of a compressed breast phantom with planar scanning,” IEEE Journal of Electromagnetics, RF and Microwaves in Medicine and Biology, vol. 2, no. 3, pp. 154–162, 2018.). This makes the entire system more time-consuming and hence is not preferrable in the present case. On the other hand, more recently, N. K. Nikolova and Tijak et. al proposed Quantitative Microwave Imaging (QMI) techniques by employing the point spread function (PSF). Direct inversion techniques such as in N. K. Nikolova is typically used to solve QMI technique but this is computationally more complex and also error prone. Further, to reduce computational complexity, Tijak et. al described 2D FFT based technique, but this requires a 2D grid-based scanning with very dense radar measurements. While QMI based approach looks promising, the above limitations have to be efficiently addressed to make it deployment friendly. Embodiments of the present disclosure provide systems that implement a microwave imaging-based tumor detection approach referred to as “Enhanced Microwave imaging for efficient breast Tumor Detection” or method of the present disclosure. The method of the present disclosure addresses the aforementioned MWI limitations by using a computationally efficient model-based reconstruction algorithm and an intelligent radar scanning mechanism to reduce the scan duration. Firstly, the system of the present disclosure formulates the model-based MWI (herein referred as method of the present disclosure) as an inverse image reconstruction problem by building the forward model using the PSF (which may be obtained via calibration). Since the tumor content is sparse and localized to fewer regions, the system of the present disclosure solves the inverse problem efficiently by using sparsity as a prior. Next, to reduce the number of scanning measurements, an intelligent scanning mechanism based on Deep Reinforcement Learning (DRL) approach is implemented to optimize the radar acquisition locations. The number of measurements (say based on maximum duration) or scans in a given episode of tumor detection is fixed and a coarse uniform scan is firstly performed to obtain an initial MWI. This coarse level image helps the RL agent to optimally suggest the remaining acquisition points. The method of the present disclosure has been benchmarked against the existing/conventional methods using an open dataset collected using 3D breast phantoms having tumor. Both the visual results and Signal to Mean Ratio (SMR) is provided to compare the performance of the method of the present disclosure with the other standard/conventional DAS and DMAS approaches. The results clearly show that the method of the present disclosure provides improved tumor localized image with reduced clutter and shows up to 2 times SMR improvement over other existing techniques. Further, only a marginal visual deterioration is observed with the method of the present disclosure despite using only 33% of the total measurements. Referring now to the drawings, and more particularly to FIGS. 1 through 8, where similar reference characters denote corresponding features consistently throughout the figures, there are shown preferred embodiments and these embodiments are described in the context of the following exemplary system and/or method. FIG. 1 depicts an exemplary system 100 for an in-body microwave (MW) imaging of a subject, in accordance with an embodiment of the present disclosure. The system 100 may also be referred as ‘an in-body imaging system’, ‘an imaging system’, or ‘an unknown dielectric constant estimation system’, and may be interchangeably used herein. In an embodiment, the system 100 includes one or more hardware processors 104, communication interface device(s) or input/output (I/O) interface(s) 106 (also referred as interface(s)), and one or more data storage devices or memory 102 operatively coupled to the one or more hardware processors 104. The one or more processors 104 may be one or more software processing components and/or hardware processors. In an embodiment, the hardware processors can be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, and/or any devices that manipulate signals based on operational instructions. Among other capabilities, the processor(s) is/are configured to fetch and execute computer-readable instructions stored in the memory. In an embodiment, the system 100 can be implemented in a variety of computing systems, such as laptop computers, notebooks, hand-held devices (e.g., smartphones, tablet phones, mobile communication devices, and the like), workstations, mainframe computers, servers, a network cloud, and the like. The I/O interface device(s) 106 can include a variety of software and hardware interfaces, for example, a web interface, a graphical user interface, and the like and can facilitate multiple communications within a wide variety of networks N/W and protocol types, including wired networks, for example, LAN, cable, etc., and wireless networks, such as WLAN, cellular, or satellite. In an embodiment, the I/O interface device(s) can include one or more ports for connecting a number of devices to one another or to another server. The memory 102 may include any computer-readable medium known in the art including, for example, volatile memory, such as static random-access memory (SRAM) and dynamic-random access memory (DRAM), and/or non-volatile memory, such as read only memory (ROM), erasable programmable ROM, flash memories, hard disks, optical disks, and magnetic tapes. In an embodiment, a database 108 is comprised in the memory 102, wherein the database 108 comprises information pertaining to coarse microwave (MW) data pertaining to a set of predefined uniform locations scanned around a specific portion (e.g., breast) of a body of a subject, one or more subsequent locations pertaining to the specific portion of the body of the subject using the initial coarse microwave data, measurement matrix, Kernel K, unknown dielectric constant, and the like. The database 108 further comprises one or more Double Deep Q Networks which when trained serve as one or more trained reinforcement learning (RL) agent, and the like. The memory 102 further comprises (or may further comprise) information pertaining to input(s)/output(s) of each step performed by the systems and methods of the present disclosure. In other words, input(s) fed at each step and output(s) generated at each step are comprised in the memory 102 and can be utilized in further processing and analysis. FIGS. 2(a) and 2(b) (collectively referred as FIG. 2), with reference to FIG. 1, depicts (a) an illustration of a subject undergoing detection test, and (b) a detailed view demonstrating the N_a optimized locations selected by the RL agent from a total of N possible radar locations around the breast tissue, in accordance with an embodiment of the present disclosure. More specifically, FIG. 2(a) illustrates the typical deployment the method, where the subject simply lie on the bed by placing the breast inside a cup. The radar which is placed below takes measurements at different antenna positions in a circular manner. The simplified view of radar scan is shown in FIG. 2(b), where the radar measurements taken at different locations are fed to a model-based MWI reconstruction algorithm to detect tumors. The RL agent intelligently chooses N_a locations from N (N_a