Home-made Database Research Articles

Recognition Performance Analysis of a Multimodal Biometric System Based on the Fusion of 3D Ultrasound Hand-Geometry and Palmprint

Multimodal biometric systems are often used in a wide variety of applications where high security is required. Such systems show several merits in terms of universality and recognition rate compared to unimodal systems. Among several acquisition technologies, ultrasound bears great potential in high secure access applications because it allows the acquisition of 3D information about the human body and is able to verify liveness of the sample. In this work, recognition performances of a multimodal system obtained by fusing palmprint and hand-geometry 3D features, which are extracted from the same collected volumetric image, are extensively evaluated. Several fusion techniques based on the weighted score sum rule and on a wide variety of possible combinations of palmprint and hand geometry scores are experimented with. Recognition performances of the various methods are evaluated and compared through verification and identification experiments carried out on a homemade database employed in previous works. Verification results demonstrated that the fusion, in most cases, produces a noticeable improvement compared to unimodal systems: an EER value of 0.06% is achieved in at least five cases against values of 1.18% and 0.63% obtained in the best case for unimodal palmprint and hand geometry, respectively. The analysis also revealed that the best fusion results do not include any combination between the best scores of unimodal characteristics. Identification experiments, carried out for the methods that provided the best verification results, consistently demonstrated an identification rate of 100%, against 98% and 91% obtained in the best case for unimodal palmprint and hand geometry, respectively.

Coupling between high-resolution mass spectrometry and focalized data-analysis methods provides the identification of new putative glycosidic non-anthocyanic flavonoids in grape.

The biochemical diversity of flavonoids is based on glycosylation, methylation, acylation, and many other modifications of the flavonoid backbone. Liquid chromatography coupled to high-resolution mass spectrometry demonstrated to be a powerful approach to gain new insights into the flavonoid composition of many plant species, including grapes. Among different metabolomic approaches, suspect screening analysis relies on the construction of a specific database and on ultra-high performance liquid chromatography/quadrupole time-of-flight (UHPLC/QTOF) analysis to find new compounds of oenological interest. A homemade database containing mass data information retrieved from the literature specific for plant flavonoid derivatives (GrapeFlavMet) was constructed. Tandem mass spectrometry analysis of V. vinifera and hybrid grape extracts was performed, and MS/MS fragmentation allowed to assign the putative flavonoid chemical structure to various identification levels, as established by the Metabolomics Standard Initiative. By this approach, putative flavonoid derivatives with different glycosylation and acylation patterns were identified. They include three pentoside derivatives of tetrahydroxy-flavone, tetrahydroxy-flavanone and myricetin isomers, a putative dihydrorhamnetin hexoside derivative, three cinchonain isomers (phenylpropanoid-substituted flavan-3-ols with antidiabetic properties), and two syringetin isomer derivatives (acetyl- and p-coumaroyl-hexoside). Two acetyl-hexoside derivatives of dihydrorhamnetin and pentahydroxy-methoxy-flavanone, and three derivatives of tetrahydroxy-dimethoxy-flavanone (acetyl, p-coumaroyl, and caffeoyl-hexoside) were tentatively annotated. Most of the compounds were identified in grape for the first time, while two putative syringetin derivatives previously proposed in the literature were confirmed. These findings deepen the current knowledge on grape flavonoids, suggesting more connections at the biochemical level.

MALDI-TOF MS application for identification of filamentous fungi

Introduction. MALDI-TOF identification method is widely used in microbiology due to its accuracy and rapid results achievement. However, applying this method to mold fungi faces some difficulties and is not always effective. Purpose of the study. The aim of the study was to evaluate the profits of using the long cultivation and protein extraction protocol in routine identification of mold fungi isolates from environment. Materials and methods. The analysis of molds museum collection from Centre for Strategic Planning of FMBA of Russia was performed by MALDI-TOF mass-spectrometry Biotyper (Bruker Daltonics) with cultivation in liquid media and long optimized protein extraction protocol with acetonitrile and formic acid. Results. One hundred thirty seven isolates were analyzed. Quality spectra were achieved for 71.5% of samples. Identification with MBT Filamentous Fungi Library database with the high confidence score (> 1.7) was achieved for 55% of isolates (26% with score >2). Samples analyzed included members of nineteen families and 27 genera. 16% of samples were not identified despite producing high-quality spectra. Limitations. When studying the possibility of using the time-of-flight mass spectrometry method to identify mold fungi, a sample of 137 isolates of mold fungi from the environment was analyzed, which is a sufficient reference sample. The analyzed samples included representatives of 19 families and 27 genera, which makes it possible to apply the findings to at least these representatives of micellar fungi. In this study 22 samples with good quality spectra, were not identified with MBT Filamentous Fungi Library database. In the future studies, these samples, along with other samples like that, will be identified by genetic molecular methods and added to the new home-made database for filamentous fungi MALDI-TOF identification. Conclusion. Effective identification of filamentous fungi by mass-spectrometry methods requires pure culture achieved from liquid media, long optimized protocol of protein extraction and building an in-house database of spectra not presented in Bruker database.

Automatically Guided Selection of a Set of Underwater Calibration Images

The 3D reconstruction of underwater scenes from overlapping images requires modeling the sensor. While underwater self-calibration gives good results when coupled with multi-view algorithms, calibration or pre-calibration with a pattern is still necessary when scenes are weakly textured or if there are not enough points of view of the same points; however, detecting patterns on underwater images or obtaining a good distribution of these patterns on a dataset is not an easy task. Thus, we propose a methodology to guide the acquisition of a relevant underwater calibration dataset. This process is intended to provide feedback in near real-time to the operator to guide the acquisition and stop it when a sufficient number of relevant calibration images have been reached. To perform this, pattern detection must be optimized both in time and success rate. We propose three variations of optimized detection algorithms, each of which takes into account different hardware capabilities. We present the results obtained on a homemade database composed of 60,000 images taken both in pools and at sea.

Multimodal Biometric Recognition Based on 3D Ultrasound Palmprint-Hand Geometry Fusion

In recent years, multimodal biometric systems are increasingly employed in many application field due to several advantages in terms of universality, recognition rate. and security. Among various acquisition technologies, Ultrasound shows important merits, because it allows obtaining volumetric images of the human body and hence a more accurate description of characteristics and to verify liveness. In this work, a multimodal ultrasound recognition system based on the fusion between 3D hand geometry and 3D palmprint features is proposed and experimentally evaluated. The system acquires a volumetric image of the whole hand and for both characteristics, several 2D images are extracted at different depth levels. From each image, 2D features are extracted and then properly combined to achieve a 3D template. Recognition performances are evaluated through verification and identification experiments by employing a homemade database. Experiments are carried out first for the two unimodal biometrics and successively, by fusing the two modalities at score level. Results have shown that fusion is able to dramatically improve the recognition performances of the single biometrics, achieving an Equal Error Rate of 0.08% and an identification rate of 100%.

Discovery and quantification of lipoamino acids in bacteria

Improving knowledge about metabolites produced by the microbiota is a key point to understand its role in human health and disease. Among them, lipoamino acid (LpAA) containing asparagine and their derivatives are bacterial metabolites which could have an impact on the host. In this study, our aim was to extend the characterization of this family. We developed a semi-targeted workflow to identify and quantify new candidates. First, the sample preparation and analytical conditions using liquid chromatography (LC) coupled to high resolution mass spectrometry (HRMS) were optimized. Using a theoretical homemade database, HRMS raw data were manually queried. This strategy allowed us to find 25 new LpAA conjugated to Asn, Gln, Asp, Glu, His, Leu, Ile, Lys, Phe, Trp and Val amino acids. These metabolites were then fully characterized by MS2, and compared to the pure synthesized standards to validate annotation. Finally, a quantitative method was developed by LC coupled to a triple quadrupole instrument, and linearity and limit of quantification were determined. 14 new LpAA were quantified in gram positive bacteria, Lactobacilus animalis, and 12 LpAA in Escherichia coli strain Nissle 1917.

Research On Pre-Training Method and Generalization Ability of Big Data Recognition Model of the Internet of Things

The Internet of Things and big data are currently hot concepts and research fields. The mining, classification, and recognition of big data in the Internet of Things system are the key links that are widely of concern at present. The artificial neural network is beneficial for multi-dimensional data classification and recognition because of its strong feature extraction and self-learning ability. Pre-training is an effective method to address the gradient diffusion problem in deep neural networks and could result in better generalization. This article focuses on the performance of supervised pre-training that uses labelled data. In particular, this pre-training procedure is a simulation that shows the changes in judgment patterns as they progress from primary to mature within the human brain. In this article, the state-of-the-art of neural network pre-training is reviewed. Then, the principles of the auto-encoder and supervised pre-training are introduced in detail. Furthermore, an extended structure of supervised pre-training is proposed. A set of experiments are carried out to compare the performances of different pre-training methods. These experiments include a comparison between the original and pre-trained networks as well as a comparison between the networks with two types of sub-network structures. In addition, a homemade database is established to analyze the influence of pre-training on the generalization ability of neural networks. Finally, an ordinary convolutional neural network is used to verify the applicability of supervised pre-training.

Discovery of minor quality evaluation marker compounds for Chinese patent medicine products using a two-leveled metabolomics strategy

Chinese patent medicines (CPMs) are popularly used in clinical practice. Though the composition is complex, the quality of CPM is usually evaluated by the contents of a few main compounds. In this study, a two-leveled metabolomics strategy was proposed to discover minor marker compounds for different CPM products. Zhenqi Fuzheng (ZQFZ) granule was studied an example, where 15 batches from 3 producers were analyzed. The samples were separated using UHPLC on an Acquity UPLC® HSS T3 column, and then detected using Q-Orbitrap-MS. In the first level, 1475 common peaks were extracted and 95 compounds were identified using diagnostic ions and a homemade database. In the second level, the data were subjected to a two-way hierarchical clustering analysis and screened by variable importance value. In total 14 marker compounds were discovered which were responsible for the grouping of different ZQFZ products. Echinacoside (22), oleoside (13), loganic acid (5), salidroside (7), ligustrosidic acid (42), 6α-hydroxygeniposide (28), and oleoside 11-methyl ester (15) could be used to reflect the quality difference for ZQFZ granule products. The proposed strategy could also contribute to the discovery of quality control markers for other CPMs.

A Feasible 3D Ultrasound Palmprint Recognition System for Secure Access Control Applications

Biometric recognition systems based on 3D palmprint captured with optical technology have been widely investigated in the last decade; however, they can provide information about the external skin surface only. This limit can be overcome by Ultrasound, which allows gaining information on the depth of palm lines and can verify the liveness of the sample, making the recognition systems very hard to fake. In this work, a feasible palmprint recognition system based on 3D ultrasound images is proposed. Unlike previous wet setups, the coupling between probe and human is realized through a gel pad, which permits a comfortable and precise positioning of the hand by the user. Collected 3D images are processed to generate 2D palmprint images at various under-skin depths. 2D features are then extracted from these images, experimenting with different procedures, and are merged to define a 3D template that contains lines’ depth information. Recognition performances were evaluated by performing verification and identification experiments on a home-made database composed of 423 samples from 55 volunteers. An EER rate of 0.36% and an identification accuracy of 100% are obtained. The suitability of the proposed system in secure access control applications is finally discussed.

Rapid Screening and Quantification of Multi-Class Multi-Residue Veterinary Drugs in Pork by a Modified Quechers Protocol Coupled to UPLC-QOrbitrap HRMS

Background:A rapid and simple analytical method for the screening and quantification of multi-residues was established by a quick, easy, cheap, effective, rugged and safe (QuEChERS) approach coupled to ultra-performance liquid chromatography and electrospray ionization quadrupole orbitrap high-resolution mass spectrometry (UPLC-Q-Orbitrap HRMS). A total number of 59 veterinary drugs were investigated, which belonged to 12 classes, such as β-agonist, quinolones, sulfonamides, tetracyclines, lincomycin series, triphenylmethane, nitroimidazoles, macrolides, amide alcohols, quinoxalines, steroid hormone and sedatives.Methods:The factors which influence the determination of veterinary drugs residues, such as mobile phase, extract solvent, clean up sorbent, and re-dissolved solvent, were optimized by the single factor experiment. The method was sufficiently validated by using the parameters of linearity, sensitivity, accuracy, and repeatability.Results:The response of the detector was linear for 59 veterinary drug residues in extensive range (two to three orders of magnitude) with a high coefficient of determination (R2) (0.9995-0.9998). The limit of quantification (LOQ) ranged from 0.1μg/kg to 2.0μg/kg for 59 veterinary drug residues in pork samples. The repeatability was in the range of 1.0%-9.5%. Average recoveries of 59 veterinary drugs at three spiked levels ranged from 53.7%-117.8% with relative standard deviation (RSD) of 1.9%-13.9%. The full MS scan coupled with data-dependent MS/MS mode was applied for screening the target compounds to simultaneously obtain the accurate mass of parent ion and the mass spectrum of fragments. Elemental composition, accurate mass, and retention time and characteristic fragment ions were used to establish a homemade database.Conclusion:The ability of the homemade database was verified by analyzing the real pork samples, and the result was satisfactory.

ARTEMIS, THE 14C AMS FACILITY OF THE LMC14 NATIONAL LABORATORY: A STATUS REPORT ON QUALITY CONTROL AND MICROSAMPLE PROCEDURES

ABSTRACTQuality control procedures have been developed at the Laboratoire de Mesure du Carbone 14 (LMC14) national laboratory throughout the years of operation. Routine procedures are applied to sample preparation depending on their composition and their size. The tuning of the ARTEMIS AMS facility, hosted by the LMC14 laboratory, uses an accurate procedure. A batch of unknown samples is measured with accompanying samples (primary and secondary standards and blanks), which give a powerful set of data to control the quality of each measurement. A homemade database has been created to store the sample information and study the evolution of the accompanying samples. The LMC14 laboratory participated in the Sixth International Radiocarbon Intercomparison, SIRI. The results are presented here, with statistical tests to assess the quality of the preparations and measurements done at the LMC14 national laboratory. To obtain a reliable radiocarbon (14C) age by AMS, 1 mg of sample is required in routine analysis. Recently, the LMC14 developed a new procedure dedicated to microsamples, allowing the size of samples to be reduced and contributing to opening 14C dating to materials that were previously unreachable. This new procedure has been successfully tested on valuable Cultural Heritage samples: lead white mural paintings.

3-D Ultrasound Palmprint Recognition System Based on Principal Lines Extracted at Several Under Skin Depths

In recent years, several palmprint recognition procedures that use 3-D information on the palm surface have been proposed in order to overcome some limitations of 2-D palmprint. However, 3-D optical images provide information on the external skin surface only. Ultrasound waves have the capability to penetrate into human tissue and, therefore, can provide new kinds of 3-D information from acquired palmprints. In this paper, a 3-D ultrasound palmprint recognition system that accounts for principal line depth is proposed. It exploits several 2-D palmprints, extracted at different under skin depths with a classic procedure, opportunely combined to achieve a 3-D template. An ad hoc matching criterion is then defined in order to account for line depth information when comparing 3-D templates. Experiments of both verification and identification, performed on a homemade database, have shown that the 3-D recognition procedure exhibits very good recognition rates. The proposed 3-D technique also benefits from other advantages of ultrasound, including not being sensitive to many kinds of surface contaminations and its capability to detect liveness during the acquisition phase, which makes it very difficult to counterfeit.

Experimental Validation of a Reliable Palmprint Recognition System Based on 2D Ultrasound Images

Ultrasound has been trialed in biometric recognition systems for many years, and at present different types of ultrasound fingerprint readers are being produced and integrated in portable devices. An important merit of the ultrasound is its ability to image the internal structure of the hand, which can guarantee improved recognition rates and resistance to spoofing attacks. In addition, ambient noise like changes of illumination, humidity, or temperature, as well as oil or ink stains on the skin do not affect the ultrasound image. In this work, a palmprint recognition system based on ultrasound images is proposed and experimentally validated. The system uses a gel pad to obtain acoustic coupling between the ultrasound probe and the user’s hand. The collected volumetric image is processed to extract 2D palmprints at various under-skin depths. Features are extracted from one of these 2D palmprints using a line-based procedure. Recognition performances of the proposed system were evaluated by performing both verification and identification experiments on a home-made database containing 281 samples collected from 32 different volunteers. An equal error rate of 0.38% and an identification rate of 100% were achieved. These results are very satisfactory, even if obtained with a relatively small database. A discussion on the causes of bad acquisitions is also presented, and a possible solution to further optimize the acquisition system is suggested.

Robust real-time hand detection and localization for space human–robot interaction based on deep learning

Hand gestures are quite suitable for space human–robot interaction (SHRI) because of their natural and convenient features. While the detection and localization of hands are the premise and foundation for SHRI based on hand gestures. But hand gestures are very complicated and hand sizes are very small in some images. These problems make the robust real-time hand detection and localization very difficult. In this paper, a feature-map-fused single shot multibox detector (FF-SSD) which is a deep learning network is designed to deal with the problems of hand detection and localization in SHRI. First, the background of the method is introduced in this paper, including an astronaut assistant robot platform, the difficulties of hand detection and localization, and introduction of the state-of-the-art deep learning networks for object detection and localization. Then, the FF-SSD is proposed for detecting and localizing hands especially pony-size hands. This network takes into consideration both accuracy and speed with balanced performance. And in the experiment part, the FF-SSD is trained and tested on hand databases which include a homemade database and two public databases. At last, the superiority of the proposed method is demonstrated compared with the state-of-the-art methods.

Investigation of pesticides and their transformation products in the Júcar River Hydrographical Basin (Spain) by wide-scope high-resolution mass spectrometry screening

The Water Framework Directive 2000/60/EC implemented by the European Union established as the main objectives to achieve a “good ecological and chemical status” of the surface water and a “good quantitative and chemical status” of groundwater bodies. One of the major pressures affecting water bodies comes from the use of pesticides and their potential presence in the water ecosystems. For this purpose, the reliable determination of pesticides and their transformation products (TPs) in natural waters (both surface and groundwater) is required. The high number of compounds potentially reaching the aquatic environment makes extraordinary difficult, if not impossible, to investigate all these compounds even using the most powerful analytical techniques. Among these, liquid chromatography coupled to high-resolution mass spectrometry is emphasized due to its strong potential for detection and identification of many organic contaminants thanks to the accurate-mass full spectrum acquisition data. This work focuses on wide-scope screening of many pesticides and their TPs in surface water and groundwater samples, collected between March and May 2017, in the Júcar River Hydrographical Basin, Spain. For this purpose, a home-made database containing more than 500 pesticides and TPs was employed. Analyses performed by liquid chromatography coupled to quadrupole-time of flight mass spectrometry (LC-QTOF MS) allowed the identification of up to 27 pesticides and 6 TPs. The most detected compounds in groundwater were the herbicides atrazine, simazine, terbuthylazine, and their TPs (atrazine-desethyl, terbumeton-desethyl and terbuthylazine-desethyl). Regarding surface water, the fungicides carbendazim, thiabendazole and imazalil, the herbicide terbutryn and the TP terbumeton-desethyl were also detected. These results illustrate the wide use of these compounds (in the present or in the recent past) in the area under study and the vulnerability of the water bodies, and are in accordance with previous findings in other water bodies of the different Spanish Hydrographic systems.

Ripeness Classification of Bananas Using an Artificial Neural Network

The quality of fresh banana fruit is a main concern for consumers and fruit industrial companies. The effectiveness and fast classification of banana’s maturity stage are the most decisive factors in determining its quality. It is necessary to design and implement image processing tools for correct ripening stage classification of the different fresh incoming banana bunches. Ripeness in banana fruit generally affects the eating quality and the market price of the fruit. In this paper, an automatic computer vision system is proposed to identify the ripening stages of bananas. First, a four-class homemade database is prepared. Second, an artificial neural network-based framework which uses color, development of brown spots, and Tamura statistical texture features is employed to classify and grade banana fruit ripening stage. Results and the performance of the proposed system are compared with various techniques such as the SVM, the naive Bayes, the KNN, the decision tree, and discriminant analysis classifiers. Results reveal that the proposed system has the highest overall recognition rate, which is 97.75%, among other techniques.

Comprehensive steroid profiling by liquid chromatography coupled to high resolution mass spectrometry

A steroidomics workflow has been developed in the objective of monitoring a wide range (n >150) of steroids in urine. The proposed workflow relies on the optimization of an adequate SPE extraction step followed by an UHPLC-HRMS/MS simultaneous analysis of both free and conjugated forms of C18, C19 and C21 steroid hormones. On the basis of 44 selected steroids, representative of main classes of steroids constituting the steroidome, the performances of the developed workflow were evaluated in terms of selectivity, repeatability (< 13%) and linearity (R2> 0.985 in the concentration range [0.01–10 ng/mL]). As metabolites identification and characterization constitute the bottleneck of such profiling approaches, a homemade database was created encompassing a large number of characterized free and conjugated steroids (n> 150) for putative steroid-like biomarkers identification purposes. The efficiency of the workflow in highlighting fine modifications within the urinary steroidome was assessed in the frame of an anabolic treatment involving an intra-muscular administration of boldenone undecylenate (2 mg/kg) to veals (n=6) and the investigation of potential steroid biomarkers. Besides monitoring known phase II metabolites of boldenone in the bovine specie, namely, boldenone glucuronide and sulfate, the applied strategy also permitted to observe, upon boldenone administration, a modified profile of epiboldenone glucuronide. Furthermore, 31 signals corresponding to non-identified steroid species could also be highlighted as impacted upon the exogenous steroid treatment. This study is the first to simultaneously investigate both free and conjugated C18, C19 and C21 steroid hormones in their native form using UHPLC-HRMS/MS and allowing their comprehensive profiling. This strategy was probed in-vivo.

A convolutional neural network-based flame detection method in video sequence

Computer vision-based fire detection is one of the crucial tasks in modern surveillance system. In recent years, the convolutional neural network (CNN) has become an active topic because of its high accuracy recognition rate in a wide range of applications. How to reliably and effectively solve the problems of flame detection, however, has still been a challenging problem in practice. In this paper, we proposed a novel flame detection algorithm based on CNN in real time by processing the video data generated by an ordinary camera monitoring a scene. Firstly, to improve the efficiency of recognition, a candidate target area extraction algorithm is proposed for dealing with the suspected flame area. Secondly, the extracted feature maps of candidate areas are classified by the designed deep neural network model based on CNN. Finally, the corresponding alarm signal is obtained by the classification results. The experimental results show that the proposed method can effectively identify fire and achieve higher alarm rate in the homemade database. The proposed method can effectively realize the real-time performance of fire warning in practice.

Elucidation of non-intentionally added substances migrating from polyester-polyurethane lacquers using automated LC-HRMS data processing.

An untargeted strategy aiming at identifying non-intentionally added substances (NIAS) migrating from coatings was developed. This innovative approach was applied to two polyester-polyurethane lacquers, for which suppliers previously provided the identity of the monomers involved. Lacquers were extracted with acetonitrile and analyzed by liquid chromatography-high resolution mass spectrometry (LC-HRMS). Data, acquired in the full scan mode, were processed using an open-source R-environment (xcms and CAMERA packages) to list the detected features and deconvolute them in groups related to individual compounds. The most intense groups, accounting for more than 85% of cumulated feature intensities, were then investigated. A homemade database, populated with predicted polyester oligomer combinations from a relevant selection of diols and diacids, enabled highlighting the presence of 14 and 17 cyclic predicted polyester oligomers in the two lacquers, including three mutual combinations explained by common known monomers. Combination hypotheses were strengthened by chromatographic considerations and by the investigation of fragmentation patterns. Regarding unpredicted migrating substances, four monomers were hypothesised to explain several polyester or caprolactam oligomer series. Finally, considering both predicted and tentatively elucidated unpredicted oligomers, it was possible to assign hypotheses to features representing up to 82% and 90% of the cumulated intensities in the two lacquers, plus 9% and 3% (respectively) originating from the procedural blank. Graphical abstract Elucidation of non-intentionally added substances.

Quality assurance of a record-and-verify system.

With the introduction of more complex three-dimensional conformal radiotherapy and intensity-modulated radiotherapy techniques in clinical practice, the use of record-and-verify systems is recommended to improve the accuracy of radiotherapy treatments. The aim of the present study was to evaluate, for a commercial record-and-verify system, the efficiency, integration with the treatment planning system, and impact of manual checking of data. The most frequent errors or misses were also evaluated. The development of internal protocols to systematically implement new technologies has been identified as a priority in the departmental quality assurance process. Data electronically fed into the record-and-verify system were compared with those manually recorded in the clinical paper chart over a period of almost 6 years (October 2000 to December 2006). A total of 7768 treated patients was reviewed. The check was performed by using a homemade data base in which the errors are stratified as follows: 1) general section, 2) geometric and dosimetric section, and 3) delivered dose section. On a total of 7768 checked patients, one or more mismatches between treatment planning system data and record-and-verify system data or paper chart data were observed for 452 patients (5.8% of total number of inspected patients). The percentage of discrepancies out of the total was: 2.2% in the general section, 3.3% in the dosimetric and geometric section, and 4.2% in the delivered-dose section. Although record-and-verify systems assume a crucial role in the accuracy and reproducibility of radiation treatment, their inability to eradicate all the errors requires vigilance on the part of the radiation therapy and physics team.