Search

The presence of an additional supernumerary distolingual root in the mandibular molar is termed Radix Entomolaris (RE). It is common in the mandibular first molar, but its occurrence in the mandibular second molar is scarcely reported in the literature. Two-dimensional imaging can diagnose anatomical root canal variations when taken with different horizontal angulations. With the recent innovations in three-dimensional diagnostic imaging, cone beam computed tomography (CBCT) can aid in unfolding the complexities of the root canal system. Accurate diagnosis by CBCT leads to the success of endodontic treatment. A rare case of radix entomolaris in the mandibular second molar in an Indian female is reported here with three of its CBCT findings. Imaging in all three dimensions, oblique sections and volume rendering by CBCT proved to be of additional help to understand the complex anatomy and root canal morphology of the tooth.

FPGAs are well-suited for accelerating deep learning (DL) applications owing to the rapidly changing algorithms, network architectures and computation requirements in this field. However, the generic building blocks available on traditional FPGAs limit the acceleration that can be achieved. Many modifications to FPGA architecture have been proposed and deployed including adding specialized artificial intelligence (AI) processing engines, adding support for smaller precision math like 8-bit fixed point and IEEE half-precision (fp16) in DSP slices, adding shadow multipliers in logic blocks, etc. In this paper, we describe replacing a portion of the FPGA’s programmable logic area with Tensor Slices. These slices have a systolic array of processing elements at their heart that support multiple tensor operations, multiple dynamically-selectable precisions and can be dynamically fractured into individual multipliers and MACs (multiply-and-accumulate). These slices have a local crossbar at the inputs that helps with easing the routing pressure caused by a large block on the FPGA. Adding these DL-specific coarse-grained hard blocks to FPGAs increases their compute density and makes them even better hardware accelerators for DL applications, while still keeping the vast majority of the real estate on the FPGA programmable at fine-grain.

Photobiomodulation (PBM) or formerly known as low-level laser therapy (LLLT) is nothing but low-dosage biophotonics for therapy which presents an advancing new era of regenerative modalities in dental implantology. It utilizes light emitting diodes (LEDs), broad light sources and lasers for this purpose. Photobiomodulation is intended for relieving pain and inammation, regulating immune responses as well as stimulating wound healing and tissue regeneration. This helps in combatting the main pathological causes of implant failures that are, peri-implant mucositis and peri-implantitis as well as helps promote osseointegration and improve stability of implants. While the scope of photobiomodulation has been thoroughly investigated in in-vitro and animal studies, human clinical trials are still scarce which makes appropriate protocol formation with respect to dosage and mode of delivery among other parameters difcult. A recently introduced and potential application of photobiomodulation in the eld of implantology aims to deliver the positive effects of biophotonics through in-situ ambulatory PBM therapy called the smart dental implant (SDI) system. This comprehensive review presents the current and future trends in the application of photobiomodulation in the eld of dental implantology.

Ectodermal dysplasia comprises a hereditary disorder which occurs as a consequence of disturbances in the ectoderm of the developing embryo. The triad of nail dystrophy, alopecia or hypotrichosis and palmoplantar hyperkeratosis is also accompanied by a lack of sweat glands and a partial or complete absence of primary and/or permanent dentition. A case reports illustrating the prosthetic rehabilitation of a young girl with anhidrotic ectodermal dysplasia associated with complete anodontia is presented. Since the oral rehabilitation of such case is often difcult; particularly in pediatric patients, treatment should be administered by a multidisciplinary team involving pediatric dentistry, orthodontics, prosthodontics and oral-maxillofacial surgery

Cone beam computed tomography (CBCT, also referred to as C-arm computed tomography [CT], cone beam volume CT, or at panel CT) is a medical imaging technique of X-ray CT where the X-rays are divergent,forming a cone. It provides a complete 3D view of the maxilla, mandible, teeth, and supporting structures with relatively high resolution allowing a more accurate diagnosis, treatment planning and monitoring, and analysis of outcomes than conventional 2D images, along with low radiation exposure to the patient. This article is intended to elaborate and enunciate on the various applications and benets of CBCT, in the realm of maxillofacial prosthodontics, over and beyond its obvious benets in the rehabilitation of patients with implants. A web-based search for relevant articles in this specic area of interest was also conducted.The selected articles were critically reviewed and the data acquired were systematically compiled.

Purpose This paper aims to devise a first-of-its-kind methodology to determine the design, operating conditions and actuation strategy of pneumatic artificial muscles (PAMs) for assistive robotic applications. This requires extensive characterization, data set generation and meaningful modelling between PAM characteristics and design variables. Such a characterization should cover a wide range of design and operation parameters. This is a stepping stone towards generating a design guide for this highly popular compliant actuator, just like any conventional element of a mechanism. Design/methodology/approach Characterization of a large pool of custom fabricated PAMs of varying designs is performed to determine their static and dynamic behaviours. Metaheuristic optimizer-based artificial neural network (ANN) structures are used to determine eight different models representing PAM behaviour. The assistance of knee flexion during level walking is targeted for evaluating the applicability of the developed actuator by attaching a PAM across the joint. Accordingly, the PAM design and the actuation strategy are optimized through a tabletop emulator. Findings The dependence of passive length, static contraction, dynamic step response for inflation and deflation of the PAMs on their design dimensions and operating parameters is successfully modelled by the ANNs. The efficacy of these models is investigated to successfully optimize the PAM design, operation parameters and actuation strategy for using a PAM in assisting knee flexion in human gait. Originality/value Characterization of static and the dynamic behaviour of a large pool of PAMs with varying designs over a wide range of operating conditions is the novel feature in this article. A lucid customizable fabrication technique is discussed to obtain a wide variety of PAM designs. Metaheuristic-based ANNs are used for tackling high non-linearity in data while modelling the PAM behaviour. An innovative tabletop emulator is used for investigating the utility of the models in the possible application of PAMs in assistive robotics.

There is always need for the earliest diagnosis of enteric fever to start the appropriate treatment for this life threatening illness. The diagnosis becomes difficult because of inherent limitation due to the long standing vague complaints. Some findings like acalculus cholecystitis, hepato-splenomegaly, enlarged mesenteric lymph nodes (MLN) and free fluid can easily be picked up by ultrasonography (USG). USG is easily available, without any ionizing radiation, non-invasive and economical. Hemoculture and Widal tests are time consuming and the cost is quite high. The study was conducted with 100 paediatric patients aged 1-year to 15-year. All were having enteric fever based on typhi dot-IgM presented to the paediatric department of the hospital. USG was conducted for the evaluation of the various relevant findings. Blood samples for CBC, LFT and blood culture were taken on the same day as that of USG evaluation. USG evaluation of confirmed enteric fever cases revealed that 25% had gall bladder findings, 23% had hepato-splenomegaly, 11% had hepatomegaly, 1% had enlarged mesenteric lymph nodes (MLN) and 3% had thickened caecum wall and ileocaecal junction. USG can diagnose many findings which are related to enteric fever and are helpful in starting the early management. Hemoculture and other investigations may delay the treatment which can lead to various complications.

There have been advancements in sensing technologies in dentistry. Clinical applications in dentistry including Diagnosis as well monitoring disease and health in dental tissues are made easy with the advanced technologies in imaging and sensing tools. These recent advances in optical sources and detectors have unique applications including non-destructive testing, high speed and cost-effective electronic circuits, and novel signal processing methods. Some of the recent advances include optical coherence tomography, occlusal stabilization splint with stress-sensor system for bruxism, periodontal probe based on the fluorescent fiber position sensor, dental implantable temperature sensor for real time diagnosis of infectious periodontal disease, diamond probe/ perio2000 system, dental pulse oximeter with customized sensor holder, periodontal probe based on the flouroscent fiber position sensor etc.

Purpose: Science mapping is commonly known as Scientometrics; is used in combination with data visualization to analyze a considerable form of bibliographic resources to study and map the scientific field and thereby serving investigators to concentrate on under investigated areas of research. This scientometric study was aimed: 1) to evaluate the level of investigation in the field of periodontitis and osteoporosis, and 2) to recognize the journals, and countries in this field. Methods: The top hundred articles with most citations were selected by electronic searches in the MEDLINE database, and the citations were cross checked by Google Scholar database. The total number of publications per year along with their yearly growth was calculated. The following data regarding the journal, country of origin, and level of evidence were analysed. Results: 467 authors within 77 clusters contributed to these 100 articles. The annual rate of growth of publications in the period from 2011 to 2020 was 6.27%. The year range of these publications started from 1992 till 2019, the second decade of the new millennium showed the highest number of articles (39 out of 100 articles). Conclusion: Though the number of articles published on the present topic has risen in the last two decades, the proportion of studies with LOE1 is as low as 5%. Therefore, recommendations are made to the researchers to lay emphasis on producing more studies with LOE1 in future.

The present study aims to enrich the specialized literature by proposing and calculating a new flash-flood propagation susceptibility index (FFPSI). Thus, firstly the Flash-Flood Potential Index (F...

Nitrate is a major pollutant in groundwater whose main source is municipal wastewater and agricultural activities. In the present study, Bayesian approaches such as Bayesian generalized linear model (BGLM), Bayesian regularized neural network (BRNN), Bayesian additive regression tree (BART), and Bayesian ridge regression (BRR) were used to model groundwater nitrate contamination in a semiarid region Marvdasht watershed, Fars province, Iran. Eleven groundwater (GW) nitrate conditioning factors have been taken as input parameters for predictive modeling. The results showed that the Bayesian models used in this study were all competent to model groundwater nitrate and the BART model with R2 = 0.83 was more efficient than the other models. The result of variable importance showed that potassium (K) has the highest importance in the models followed by rainfall, altitude, groundwater depth, and distance from the residential area. The results of the study can support the decision-making process to control and reduce the sources of nitrate pollution.

At present, people have been waiting anxiously for the system that will satisfy their needs more clearly. The majority of the companies in the restaurant sector is looking for some software that improves the quality of delivery and increasing revenue. In a typical system, the waiter records the customer's order, and then put the order in to the kitchen, and then billing is done, which spends a lot of time and might lead to errors. The goal of this project is to automate the process of ordering food from the table in a restaurant and reserving a table, as well as to improve the quality of their customer service. Smart Restaurant is a concept of a restaurant's business that is based on the use of state-of-the-art technology from reservation to the order and the customer sign a relationship with you. The traditional restaurant will be replaced with the help of a smartphone, a tablet, or a graphical user interface and interactive touch screens. Customers will have to order their food via mobile phones by scanning a QR code on a table, as they are directly linked to the kitchen via a central server. In addition, our records are permanently stored on a central server, which can then be used for the marketing, distribution and sales. A smart restaurant, decrease the number of staff who work at the hotel-services, so as to increase the profit margins. The kitchen is going to be one of the interfaces which will be provided in accordance with the following priority (first-come-first-served). This app will help you to find more detailed information on the restaurant and its services, as well as the ease of ordering food and reservation table.

Weightless neural networks (WNNs) are an alternative pattern recognition technique where RAM nodes function as neurons. As both training and inference require mostly table lookups, few additions, and no multiplications, WNNs are suitable for high-performance and low-power embedded applications. This work introduces a novel approach to implement WiSARD, the leading WNN state-of-the-art architecture, completely eliminating memories and arithmetic circuits and utilizing only logic functions. The approach creates compressed minimized implementations by converting trained WNN nodes from lookup tables to logic functions. The proposed LogicWiSARD is implemented in FPGA and ASIC technologies to illustrate its suitability for edge inference. Experimental results show more than 80% reduction in energy consumption when the proposed LogicWiSARD model is compared with a multilayer perceptron network (MLP) of equivalent accuracy. Compared to previous work on FPGA implementations for WNNs, convolutional neural networks, and binary neural networks, the energy savings of LogicWiSARD range between 32.2% and 99.6%. info:eu-repo/semantics/acceptedVersion

Groundwater being an essential resource is not easily available in some parts of the world. The present study, aimed at procuring better prediction maps for groundwater potential zones, is based on a novel approach combining the use of k-fold cross-validation method and the implementation of four scenarios, each comprising of six machine learning models, ANFIS (Adaptive Neuro Fuzzy Inference System) and five other ensembles of it, ANFIS-Firefly, ANFIS-Bees, ANFIS-GA, ANFIS-DE and ANFIS-ACO. Ada Boost Model has played a vital role in determining the collinearity among the fourteen conditioning factors, which are, Lithology, Slope, TST, TRI, LULC, HAND, Curvature, Distance to Stream, Distance to Fault, Rainfall, Fault Density, Drainage Density, Elevation and Aspect. The AUCROC (Area Under Curve – Receiver Operating Characteristics) approach was employed as a model evaluation metric along with Accuracy, Sensitivity and Specificity. Among the models, ANFIS-DE showed the most promising results, acquiring the highest average values among the four scenarios for AUC (0.934), Accuracy (0.987), Sensitivity (0.985) and Specificity (0.985). Promising results of this study gives the necessary incentive for further applying this approach for groundwater zonation of other areas of the world as well as other areas of hydrogeological studies.

The crystalline quality of GaN layer grown on 4H-SiC substrate using Metal Organic Chemical Vapor Deposition (MOCVD) has been improved by reducing the growth rate of GaN. Depth resolved CL (cathodoluminescence) study was also done to analyze the defects in the GaN layers. Two samples, one with fast growth rate (∼0.7 nm/s; Sample-A) of GaN and other with slower growth rate (∼0.2 nm/s; Sample-B) were grown. The surface morphology and crystalline quality of GaN epi-layers have been studied by Atomic Force Microscopy (AFM) and High Resolution X-Ray Diffraction (HRXRD), respectively. It was observed that growth rate of the GaN layer strongly influenced its crystalline quality and surface morphology. The GaN layer grown with slow growth rate exhibited improved surface morphology with lower rms roughness of ∼0.18 nm and lower surface pits density. X-ray rocking curve measurement for (0 0 2) and (1 0 2) planes showed that both screw and edge dislocation densities were reduced for the GaN grown with slow growth rate. Depth profiling by CL spectroscopy revealed intense Yellow (YL) and Blue luminescence (BL) band in Sample-A, while in Sample B, the YL and BL bands were much weaker. The higher yellow and blue luminescence in the GaN grown at higher growth rate may be correlated with nitrogen vacancies (VN) and two charged state of CN(−/0 and 0/+) acceptor.

Molybdenum Trioxide (MoO3) films are grown on Si(100) substrates by reactive RF magnetron sputtering in plasma containing a mixture of Argon and Oxygen, using a pure Molybdenum target. In this paper, we report the deposition of (MoO3) films on Si(100) substrates under varying gas flow (O2 + Ar gas) (20 sccm to 30 sccm with the duration of deposition~ 1hr) by RF reactive magnetron sputtering at room temperature. To get crystalline MoO3 films annealing in O2 environment at 500 °C for 4 h is done. Phase formation and orientation of the film is characterized by Glancing incidence X-ray diffraction (GIXRD). The identification of the orthorhombic MoO3 phase is investigated by XRD and Raman spectroscopy. Raman lines at 819 cm-1 and 995 cm-1 are due to the (A1g, B1g) symmetric stretching (Mo-O–Mo) bond and asymmetric stretching band (Mo=O) respectively. Surface morphology and cross-sectional image of the deposited thin films were investigated by FE-SEM image. UV-Visible reflectance and cross-sectional FE-SEM image confirm the thickness of the MoO3 films with oxygen-rich and oxygen deficient phase formation occur. Reverse leakage current density of 20 sccm 1hr sample is low (1×10-6 mA/cm2) as compared to 30 sccm 1hr sample (1×10-3 mA/cm2). The higher leakage is due to crack formation during the ex-situ annealing of MoO3 films. This MoO3 films can be used in Gas sensing and switching devices.

Generally, it is very difficult to grow large diameter 4H-SiC single crystal with single polytype by Physical Vapor Transport (PVT) growth method and mostly it ends up with the presence of some other polytypes (viz. 6H, 15R). This paper presents the various comprehensive polytype identification techniques in SiC wafer grown by PVT method. Characterization techniques, viz. X-Ray diffraction, Scanning electron microscopy, Cathodoluminescence (CL) and Raman spectroscopy, are used in the present study in order to identify the presence of 4H, 6H and 15R polytype region in SiC wafer. Raman mapping (using phonon frequencies at 150, 171, 203 cm−1 for 6H, 15R and 4H, respectively) and X-Ray Topography [using grazing incidence asymmetric plane (11–2 8) for 4H-SiC, (11–2 12) for 6H-SiC and (11–2 30) for 15H-SiC] and CL spectra (defect state peak positions at ~ 500 nm and 580 nm for 4H and 6H, respectively) are proposed to distinguish the switching of polytypes in a large area SiC wafer. The polytype switching in SiC ingot may occur because of temperature fluctuations during the sublimation growth process.

In this study, the GaN films were deposited in the multilayered structures under different growth conditions. SIMS analysis showed that intrinsic carbon incorporation around two orders can be controlled effectively by varying the growth parameters. Defects analysis by cross-sectional cathodoluminescence (X-CL) is demonstrated as a useful analytical technique for unambiguously understanding the effect of incorporation of the intrinsic carbon concentration on the yellow luminescence (YL) and blue luminescence (BL) band intensities. It was inferred from the experimental data that the observed BL and YL intensities are due to the electronic transitions through the isolated CN and CNON complex, respectively. This finding is in contrast to the theoretically published literature that two charged states of CN are responsible for both BL and YL which sought that the BL emission to be always smaller in comparison to YL emission. Our experimental data, for all different growth conditions showed the prominent BL in comparison to YL except for the growth at sufficiently lower temperature which is considered to be more conducive for higher oxygen incorporation into the stoichiometry. This increases the probability of CNON formation which is manifested in the enhancement of YL intensity over BL as the formation energy of CNON complex is lower than the CN defect. Thus, X-CL analysis method can be key to address the challenges of understanding the impurity character in multi-thin layered GaN structures.

Non-Hodgkin’s Lymphoma (NHL) is very commonly associated with immunosuppressive states. Its occurrence in an immunocompetent person is scarcely reported in literature. Although NHL is known to occur in the jaws, mandibular involvement with invasion in bone is an uncommon entity. Dental surgeons should be aware that NHL can present in the oral cavity as a non-tender swelling or ulceration and should be included in the differential diagnosis of oral swellings and ulcers. Histopathological examination, assessment of immune status of the patient and search for primary or metastatic lesions should be a part of the diagnostic work up. A case of NHL of the mandible in an immunocompetent patient with follow up is presented here.

INTRODUCTION Reduction mammaplasty is one of the most common plastic surgery procedures performed. No study has evaluated whether geriatric patients are at greater risk for developing postoperative complications relative to nongeriatric patients. METHODS The American College of Surgeons National Surgical Quality Improvement Project (NSQIP) database was reviewed for reduction mammaplasty procedures from 2005 to 2017. Hypothesis testing for demographics, comorbidities, and postoperative complications between geriatric and nongeriatric patients was performed. Statistically significant differences were then evaluated with multivariate logistic regression analysis. RESULTS A total of 25,909 reduction mammaplasties were collected by NSQIP from 2005 to 2017, with 1897 patients older than 64 years (8% of all cases). The average age for geriatric breast reduction patients was 69 years versus 41.5 years for nongeriatric patients. Rates of comorbidities including diabetes, smoking status, dyspnea status, and hypertension medication were statistically different between the groups. Rates of deep venous thrombosis (DVT) and pulmonary embolism (PE) were significantly higher in geriatric versus nongeriatric patients, respectively (0.32 vs 0.06, P < 0.001 and 0.37 vs 0.09, P < 0.001). Multivariate analysis demonstrated geriatric patients had a 4.2 and 3.9 times higher risk of developing a DVT and PE than nongeriatric patients (C.I. 1.5-11.6, P = 0.006 and C.I. 1.6-9.8, P = 0.004). DISCUSSION This study represents the largest evaluation of geriatric reduction mammaplasties in the United States. Although rare, geriatric age confers a 4 times increased risk for developing DVT and PE relative to nongeriatric patients all while adjusting for preoperative risk factors in reduction mammaplasty. Plastic surgeons should consider counseling their geriatric patients regarding these increased risks.

To compare the rates of hormonal evaluation in patients who had CT reports describing adrenal incidentalomas with and without a specific recommendation for hormonal evaluation. We performed a retrospective review of adult outpatients without a history of cancer who had a CT report describing an incidental adrenal nodule. Radiology reports were reviewed to determine whether a standardized macro was used which gave specific recommendations for hormonal evaluation and endocrinology consultation. If no macro was used it was determined whether the report had a recommendation for hormonal evaluation and endocrinology consultation. A standardized macro recommending hormonal evaluation and endocrinology referral was used in 45/129 (34.8%) reports that described an incidental adrenal nodule. A recommendation for hormonal evaluation was made in 5/84 (6.0%) reports without a macro. Hormonal evaluation was performed in 24/50 (48.0%) patients whose reports recommended it and in 11/79 (13.9%) patients whose reports did not (p

This paper presents formation of interconnected mesoporous structure using low energy (50 keV) Ar+-ion irradiation normal to the GaSb surface for the ion fluence 1 × 1016 and 5 × 1017 ions/cm2. The...

BackgroundPatients with lower extremity wounds from diabetes mellitus or peripheral artery disease (PAD) have a risk of amputation as high as 25%. In patients with arterial disease, revascularization decreases the risk of amputation. We aimed to determine if the early assessment of arterial perfusion correlates with the risk of amputation.MethodsWe retrospectively reviewed patients referred to the vascular clinic over 18 months with Rutherford Grade 5 and 6 chronic limb-threatening ischemia to determine if patients had a pulse exam done at the time the wound was identified and when ankle brachial index (ABI) testing to evaluate perfusion was performed. Kaplan Meier analysis was used to determine if the timing of ABI testing affected the time to revascularization, wound healing, and risk of amputation.ResultsNinety-three patients with lower extremity wounds were identified. Of these, 59 patients (63%) did not have a pulse exam performed by their primary care provider when the wound was identified. Patients were classified by when they underwent ankle brachial index testing to assess arterial perfusion. Twenty-four had early ABI (

Background: Osteoporosis can be seen in children because of chronic medical illness or medication and this must be evaluated. This evaluation results in a lower risk of fractures by early intervention and subsequent management. In the paediatric age group, the radiologist must play an important role in interpreting the DEXA (Dual Energy X-ray Absorptiometery) examination results. The role includes being a statistician, an orthopaedic specialist, and a clinical pathologist. The various factors that influence bone mineral density have been investigated and compared to the available standard database. Bone mineral density (BMD) contributes to a more detailed picture of the overall bone health. There are numerous factors that influence a child's growing skeleton. The radiologist should be familiar with the technical aspects of acquiring this information. The interpretation is always based on information other than the DEXA scan. Z-scoring and other biochemical evaluation parameters are used to diagnose osteoporosis in children. Adult T-scores differ significantly from those that of children. Methods: Twenty pediatric patients (n=20) were subjected to DEXA examination from January to May 2017 who were referred from pediatric department. All these children were from 10-19 years age group with mixed gender. These patients were having symptoms of common ailments as that of respiratory, renal, gastrointestinal and others. Results: Ten out of twenty cases (50%) were found to have osteopenia or osteoporosis. The sensitivity of picking up the bone density was high in estimation of lumbosacral spine as compared to the femoral neck region. One case of mal-absorption syndrome was found to have only osteopenia (10%), while others were found to have osteoporosis (90 percent). Among the later group, tuberculosis (40%) was discovered to be the most common positive ailment, followed by mal-absorption (30%), kidney disease (10%), diabetes mellitus (10%), and worm infestation (10%). Males were dominating in this group (72.7%) as compared to females (27.3%). Three (15%) were declared to be within the normal range. Conclusion: DEXA scan Z-scoring estimation is a highly sensitive method for diagnosing osteoporosis and osteopenia in children. This is non invasive with minimal non-harmful radiation exposure. The evaluation is beneficial in chronic diseases such as tuberculosis, diabetes mellitus, malabsorption syndrome, and kidney disease in diagnosing decreased BMD. This aids in the prevention of fractures in these vulnerable patients by timely management.

With the prevalence of deep learning (DL) in many applications, researchers are investigating different ways of optimizing FPGA architecture and CAD to achieve better quality-of-results (QoR) on DL-based workloads. In this optimization process, benchmark circuits are an essential component; the QoR achieved on a set of benchmarks is the main driver for architecture and CAD design choices. However, current academic benchmark suites are inadequate, as they do not capture any designs from the DL domain. This work presents a new suite of DL acceleration benchmark circuits for FPGA architecture and CAD research, called Koios. This suite of 19 circuits covers a wide variety of accelerated neural networks, design sizes, implementation styles, abstraction levels, and numerical precisions. These designs are larger, more data parallel, more heterogeneous, more deeply pipelined, and utilize more FPGA architectural features compared to existing open-source benchmarks. This enables researchers to pin-point architectural inefficiencies for this class of workloads and optimize CAD tools on more realistic benchmarks that stress the CAD algorithms in different ways. In this paper, we describe the designs in our benchmark suite, present results of running them through the Verilog-to-Routing (VTR) flow using a recent FPGA architecture model, and identify key insights from the resulting metrics. On average, our benchmarks have 3.7x more netlist primitives, 1.8x and 4.7x higher DSP and BRAM densities, and 1.7x higher frequency with 1.9x more near-critical paths compared to the widely-used VTR suite. Finally, we present two example case studies showing how architectural exploration for DL-optimized FPGAs can be performed using our new benchmark suite.