Automatic Localization Research Articles

Reliability and accuracy of Artificial intelligence-based software for cephalometric diagnosis. A diagnostic study

BackgroundArtificial intelligence (AI) is revolutionizing cephalometric diagnosis in orthodontics, streamlining the patient assessments. This study aimed to assess the reliability, accuracy, and time consumption of artificial intelligence (AI)-based software compared to a conventional digital cephalometric analysis method on 2D lateral cephalogram.Methods408 lateral cephalometries were analysed using three methods: manual landmark localization, automatic localization, and semi-automatic localization with AI-based software. On each lateral cephalogram, 15 variables were selected, including skeletal, dental, and soft tissue measurements. The difference between the two AI-based software options (automatic and semi-automatic) was compared with the conventional digital technique. The time required to produce a complete cephalometric tracing was evaluated for each method using Student’s t-test.ResultsStatistically significant differences in the accuracy of landmark positioning were detected among the three different techniques (p < 0,01). However, it is noteworthy that almost all of these differences were not clinically significant. There was a small difference in accuracy between the semi-automatic AI-based option and conventional digital techniques. Regarding the time used for each technique, the automatic version was the fastest, followed by the semi-automatic option and the conventional digital technique. (p < 0,000).ConclusionsThe study showed a statistical difference in accuracy between the conventional digital technique and two AI-based software alternatives, but these differences were not clinically significant except for specific measurements. The semi-automatic option was more accurate than the automatic one and faster than conventional tracing. Further research is needed to confirm AI’s accuracy in cephalometric tracing.

Intelligent recognition and automatic localization of pipeline welds based on multi-vision system

Abstract Currently, the leakage detection of spacecraft pipeline welds relies on manual point-by-point inspection using a detection gun, which is inefficient and inadequate for the automation needs of spacecraft production. However, the accurate recognition and precise localization of widely distributed and small pipeline welds are crucial for automated detection. Therefore, this paper proposes a multi-vision detection and localization system that integrates global and local information, considering both comprehensive global 3D search and high-precision local 3D measurement. The improved YOLOv8 model is employed for pipeline weld recognition, which improves the recognition rate of welds. Based on the deep learning recognized and segmented welds, this paper proposes stereo matching and segmentation extraction methods for 3D localization and pipeline orientation determination. Additionally, the system integrates a robot to perform automated point-by-point inspection of welds within the area without collisions. The experimental results demonstrate the effectiveness of the improved YOLOv8 and the proposed methods for 3D weld localization and pipeline orientation determination. The maximum deviation of the spatial distance of fine weld positioning is 0.20 mm, and the repeatability of the 3D coordinates is around 0.1 mm. The system can perform precise localization and detection, meeting the requirements for automatic weld recognition and localization.

Bolt loosening assessment using ensemble vision models for automatic localization and feature extraction with target‐free perspective adaptation

AbstractBolt loosening assessment is crucial to identify early warnings of structural degradation and prevent catastrophic events. This paper proposes an automatic bolt loosening assessment methodology. First, a novel end‐to‐end ensemble vision model, Bolt‐FP‐Net, is proposed to reason the locations of bolts and their hexagonal feature patterns concurrently. Second, an adaptive target‐free perspective correction method is proposed to correct perspective distortion and enhance assessment accuracy. Finally, an iterative bolt loosening quantification is developed to estimate and refine the bolt loosening rotation. Experimental parametric studies indicated that the proposed Bolt‐FP‐Net can achieve excellent performance under different environmental conditions. Finally, a case study was conducted on steel bolt connections, which shows the proposed methodology can achieve high accuracy and real‐time speed in bolt loosening assessment.

ATD-300 an AI empowered miniature acoustic device for automatic gunshot attacks detection and image capture

Built on over 30 years of acoustic source recognition expertise in transient impulsives signature, recently highly empowered by neural networks, and wave propagation model driven algorithms, ACOEM has developed the ATD-300, an acoustic sensor for automatic gunshot detection, identification and localization. What sets this sensor apart from the competition is its reliability in terms of detection rate over false alarm and robust high resolution shooter pointing capability. Therefore it can be used directly by police departments without third-party human validation. Resulting from the design effort to distribute intelligence among a collaborative sensor network, processing and artificial intelligence algorithms are fully embedded locally, enabling an alert to be raised in less than three seconds and a camera to be pointed at the shooter with only one sensor. In addition, as the system is already installed in several American cities, we will show how a device-to-server distributed artificial intelligence approach continuously improves the performance of the solution, fed by real field data.

Automated localization of dike leakage outlets using UAV-borne thermography and YOLO-based object detectors

Leakage-induced soil erosion poses a major threat to dike failure, particularly during floods. Timely detection and notification of leakage outlets to dike management are crucial for ensuring dike safety. However, manual inspection, the current main approach for identifying leakage outlets, is costly, inefficient, and lacks spatial coverage. To achieve efficient and automatic localization of dike leakage outlets, an innovative strategy combining drones, infrared thermography, and deep learning is presented. Drones are employed for dikes’ surface sensing. Real-time images from these drones are sent to a server where well-trained detectors are deployed. Once a leakage outlet is detected, alarming information is remotely sent to dike managers. To realize this strategy, 4 thermal imagers were employed to image leaking outlets of several models and actual dikes. 9,231 hand-labeled thermal images with 13,387 leaking objects were selected for analysis. 19 detectors were trained using transfer learning. The best detector achieved a mean average precision of 95.8 % on the challenging test set. A full-scale embankment was constructed for leakage outlet detection tests. Various field tests confirmed the efficiency of the proposed leakage outlet localization method. In some tough conditions, the trained detector also evidently outperformed manual judgement. Results indicate that under typical circumstances, the localization error of the proposed method is within 5 m, demonstrating its practical reliability. Finally, the influencing factors and limits of the suggested strategy are thoroughly examined.

Secure smart grid implementation with automatic data integrity attack location prediction and exalted energy theft detection

Smart grid systems use digital communication technology to efficiently monitor and manage electricity distribution. However, there is a need for improve the robustness of smart grid systems. Hence a novel “Secure Smart Grid Implementation with Automatic Data Integrity Attack Location Prediction and Exalted Energy Theft Detection” has been proposed. Current detection methods are not good at identifying data integrity attacks. So a novel Automatic Deterministic Attack Location Prediction Model detects data integrity attacks using a Stacked Watermarking BiGRU-based Deep Deterministic Q network, extracting system matrix and state time series variation. Moreover, Energy theft detection algorithms, which are intended to detect dishonest clients, provide erroneous findings because of linear interpolation processes and randomised near misses, which ignore important data points. Thus a novel Exalted Energy Theft Detector with Secure Encoding is introduced where Amortized Multicode GAN method secures smart grids by detecting energy theft and preventing firmware code modifications, utilising fractional multi-level quantum encoding to identify injection locations. As a result, the proposed model has an accuracy of 97.1%, recall of 96.25%, error rate is 1.36%, F1-score is 96.20%, sensitivity is 96.10%, specificity is 96.10%, MAPE is 0.48%, MSE is 0.0001 and RMSE is 0.017.

Automatic location and recognition of horse freezing brand using rotational YOLOv5 deep learning network

Individual livestock identification is of great importance to precision livestock farming. Liquid nitrogen freezing labeled horse brand is an effective way for livestock individual identification. Along with various technological developments, deep-learning-based methods have been applied in such individual marking recognition. In this research, a deep learning method for oriented horse brand location and recognition was proposed. Firstly, Rotational YOLOv5 (R-YOLOv5) was adopted to locate the oriented horse brand, then the cropped images of the brand area were trained by YOLOv5 for number recognition. In the first step, unlike classical detection methods, R-YOLOv5 introduced the orientation into the YOLO framework by integrating Circle Smooth Label (CSL). Besides, Coordinate Attention (CA) was added to raise the attention to positional information in the network. These improvements enhanced the accuracy of detecting oriented brands. In the second step, number recognition was considered as a target detection task because of the requirement of accurate recognition. Finally, the whole brand number was obtained according to the sequences of each detection box position. The experiment results showed that R-YOLOv5 outperformed other rotating target detection algorithms, and the AP (Average Accuracy) was 95.6 %, the FLOPs were 17.4 G, the detection speed was 14.3 fps. As for the results of number recognition, the mAP (mean Average Accuracy) was 95.77 %, the weight size was 13.71 MB, and the detection speed was 68.6 fps. The two-step method can accurately identify brand numbers with complex backgrounds. It also provides a stable and lightweight method for livestock individual identification.

In-field disease symptom detection and localisation using explainable deep learning: Use case for downy mildew in grapevine

Diseases and pests in agriculture significantly impact crop yield and quality. Downy mildew (Plasmopara viticola) is a particular noteworthy example in grapevines. Traditional detection methods are laborious, subjective and time-consuming. Consequently, a technological solution based on artificial intelligence, would provide higher levels of reproducibility and sampling. The aim of this work was to develop an interpretable, automated method for detection and localisation of plant disease symptoms under field conditions. Images of the grapevine canopy were taken in 14 commercial vineyard plots under a range of lightning conditions, including both static and on-the-go settings. The images were processed using a sliding window, classifying sub-images into areas with and without downy mildew. Transfer learning, fine-tuning and data augmentation were employed to automate the classification, comparing convolutional neural networks (CNNs) and vision transformers (ViT). Subsequently, the trained model was integrated into the sliding window to localise regions within the canopy images exhibiting symptoms of downy mildew. Model predictions were interpreted using explainable artificial intelligence (XAI) methods. The EfficientNetV2S model achieved an accuracy of 91 % and an F1-score of 0.92 when classifying image areas and an Intersection over Union (IoU) of 0.83 when locating symptomatic areas. This method showed promising results, enabling automatic and explainable detection and localisation of plant diseases in complex conditions. The straightforward labelling process facilitated adaptation to new conditions, making it suitable for different crops and diseases. Integration into mobile platforms could enhance disease management and reduce the spread of pathogens, making a significant advance in agricultural technology.

MiLoPYP: self-supervised molecular pattern mining and particle localization in situ

Cryo-electron tomography allows the routine visualization of cellular landscapes in three dimensions at nanometer-range resolutions. When combined with single-particle tomography, it is possible to obtain near-atomic resolution structures of frequently occurring macromolecules within their native environment. Two outstanding challenges associated with cryo-electron tomography/single-particle tomography are the automatic identification and localization of proteins, tasks that are hindered by the molecular crowding inside cells, imaging distortions characteristic of cryo-electron tomography tomograms and the sheer size of tomographic datasets. Current methods suffer from low accuracy, demand extensive and time-consuming manual labeling or are limited to the detection of specific types of proteins. Here, we present MiLoPYP, a two-step dataset-specific contrastive learning-based framework that enables fast molecular pattern mining followed by accurate protein localization. MiLoPYP’s ability to effectively detect and localize a wide range of targets including globular and tubular complexes as well as large membrane proteins, will contribute to streamline and broaden the applicability of high-resolution workflows for in situ structure determination.

Automatic localization of anatomical landmarks in head cine fluoroscopy images via deep learning.

Fluoroscopy guided interventions (FGIs) pose a risk of prolonged radiation exposure; personalized patient dosimetry is necessary to improve patient safety during these procedures. However, current FGIs systems do not capture the precise exposure regions of the patient, making it challenging to perform patient-procedure-specific dosimetry. Thus, there is a pressing need to develop approaches to extract and use this information to enable personalized radiation dosimetry for interventional procedures. To propose a deep learning (DL) approach for the automatic localization of 3D anatomical landmarks on randomly collimated and magnified 2D head fluoroscopy images. The model was developed with datasets comprising 800000 pseudo 2D synthetic images (mixture of vessel-enhanced and non-enhancement), each with 55 annotated anatomical landmarks (two are landmarks for eye lenses), generated from 135 retrospectively collected head computed tomography (CT) volumetric data. Before training, dynamic random cropping was performed to mimic the varied field-size collimation in FGI procedures. Gaussian-distributed additive noise was applied to each individual image to enhance the robustness of the DL model in handling image degradation that may occur during clinical image acquisition in a clinical environment. The model was trained with 629370 synthetic images for approximately 275000 iterations and evaluated against a synthetic image test set and a clinical fluoroscopy test set. The model shows good performance in estimating in- and out-of-image landmark positions and shows feasibility to instantiate the skull shape. The model successfully detected 96.4% and 92.5% 2D and 3D landmarks, respectively, within a 10mm error on synthetic test images. It demonstrated an average of 3.6±2.3mm mean radial error and successfully detected 96.8% 2D landmarks within 10mm error on clinical fluoroscopy images. Our deep-learning model successfully localizes anatomical landmarks and estimates the gross shape of skull structures from collimated 2D projection views. This method may help identify the exposure region required for patient-specific organ dosimetry in FGIs procedures.

Machine-learning based location of the 2021 MW 7.4 Maduo, Qinghai, China earthquake sequence: Insight into intraplate seismogenesis

On 22 May 2021, an MW 7.4 earthquake occurred in Maduo County, Qinghai Province, China, which is located on the Kunlun Mountain Pass-Jiangcuo fault inside the Bayan Har block, providing a good opportunity to investigate seismogenesis of large intraplate earthquakes. We analyze two years of continuous seismic data from June 2021 to June 2023, which were recorded at 34 portable seismic stations of the MaduoArray deployed in the source zone by our group. The LOC-FLOW workflow of automatic detection and location is applied to construct a complete and high-precision seismic catalog for the region, which includes machine-learning phase picking (PhaseNet), earthquake phase association (REAL), velocity model updating and station correction (VELEST), absolute earthquake location (HypoInverse), and relative location (HypoDD). As a result, 78,832 Maduo aftershocks and other local earthquakes are detected and relocated precisely. Our results show that the length of the Maduo aftershock zone is ∼170 km, which is mainly distributed along the NWW-SEE oriented Kunlun Mountain Pass-Jiangcuo fault, and there is a horsetail bifurcation feature at the eastern end of the aftershock sequence. The seismogenic fault is nearly vertical, and local seismicity occurs on both sides of the fault. Our results also show that there is no seismic gap or aftershock sparse area in the region. Previous studies have revealed a low-velocity and high-conductivity anomaly below the source zone, reflecting fluids ascending from the lower crustal flow. These results provide new insights into the cause of the 2021 Maduo earthquake.

An Intelligent Model of Anomaly Detection and Anomaly Localization in Images Using Hybrid Heuristic Adaptive Multiscale Attention-Based DenseNet and Cascaded Variational Autoencoder

Anomaly detection and localization become necessary since humans are unable to spot visual frauds. Since it differs from the normal representation of shapes, colors, textures, size, and location, it does not have the potential to be employed in such practical applications. Hence, the automatic anomaly detection and localization framework is suggested for supporting situations like industrial systems, healthcare systems and so on. The detection and localization are the most intriguing and challenging concerns in image processing. In general, anomalies can manifest themselves in diverse ways. Detection is the process of providing information about whether an image contains modifications or not. On the other hand, localization is defined as locating the anomalies in images. Several traditional models have been deployed, but they still lack such constraints to evade performance enhancement. To sort out this issue, a novel adaptive method is recommended for anomaly detection and localization in images. Initially, the source images are acquired from standard public data sources. Once the required images are collected, such abnormal or anomalous activities are detected by adopting the new method known as Adaptive Multiscale Attention-based DenseNet (AMA-DeNet), where the hyperparameters are tuned by a hybrid algorithm known as the Hybrid Chimp Grasshopper Optimization Algorithm (HCGOA). Subsequently, the final stage is localizing the anomalies in the detected images. This is to be accomplished by using the Cascaded Variational Autoencoder (CVA), in which parameters are optimally chosen by HCGOA. After implementation, the experimental results are computed across diverse validating measures. Hence, the extensive results elucidate that the recommended model contains the potential to detect the abnormalities in the images accurately, thereby enhancing the efficiency of the new system.

Automatic localization of image semantic patches for crop disease recognition

Crop disease recognition plays a crucial role in agricultural production. However, disease images are large in scale and have a lot of redundant information, which reduces the effectiveness of deep neural networks in extracting diseases. To address the above issues and considering that not all image regions are relevant to disease recognition, this study proposes an efficient crop disease recognition method with dynamic reduction of image redundancy. The method is a two-stage process. In the first stage, we employ the lightweight CA-AnchorNet, which incorporates coordinate attention, to swiftly generate a feature map of the affected crop areas. Subsequently, class activation maps (CAMs) are utilized to identify the disease feature regions, highlighting areas that exhibit class discriminability. These regions are then mapped to a higher resolution from the original lower-resolution image, and the target patch is extracted. In the second stage, these local semantic patches, characterized by reduced spatial redundancy, are fed into the lightweight PatchNet for accurate recognition. PatchNet incorporates the Inception-C module and the ACON-C activation function. These features enhance the model's ability to express multi-scale and non-linear characteristics of crop disease features. This method does not require manually annotated region boxes and achieves an identification accuracy of 99.86 % on a 12-class crop disease dataset with complex environments. The parameter count is only 0.98 M. This method has the characteristics of accurate localization and low parameter count, and can be used for effective and high-precision recognition of crop diseases in complex environments.

Full-waveform automatic location of small seismic events in an underground mine using synthetic strain Greens tensors

A reliable method to locate microseismic evets is an important objective for routine mine seismic monitoring. We developed an algorithm that can automatically and reliably locate an event from a single uniaxial trace. Signals created by microseismic events are recorded as seismograms that consist of direct P- and S-waves as well as the waves reflected off the underground excavations. We take advantage of the extra complexity these reflections introduce as they make the signal more unique. In our investigation we cross-correlated unaltered synthetic seismograms against trial waveforms created from a linear combination of appropriate waveforms from a library of numerically calculated strain Green's tensors. We use a single uniaxial synthetic seismogram to invert for the source location, six moment tensor components, and frequency. The resulting cost functions are constructed as maximal correlations at each node in the mine grid through the true source x-planes. This function is maximized by differential evolution to simultaneously yield the source parameters. Sensitivity tests such as shifting the mine plans and altering the background velocities of the media are investigated to test the robustness of the method. We contaminate the data with 40% white noise to test the resolution against simulated real recorded seismograms. The method is demonstrated using synthetic data calculated with a realistic underground 3D velocity model.

A fully automatic curve localization method for extracted spine

The automation of scoliosis positioning presents a challenging and often understated task, yet it holds fundamental significance for the automated analysis of spinal morphological anomalies. This paper introduces a novel spinal curve localization model for precisely differentiating the spinal curves and identifying their concave centers. The proposed model contains three components: i) custom spine central line model, to define the spine central line as a combination of several secant line sequences with different polarities; ii) custom curve model, to classify each spinal curve into one of 11 curves types and deduce each its concave centers by several custom formulas; and iii) adapted distance transform and quadratic line fitting algorithm coupled with custom secant line segment searching strategy (DTQL-LS), to search all line segments in the spine and group consecutive line segments with identical polarity into line sequence. Experimental results show that its positioning success rate is close to 99%. Furthermore, it exhibits significant time efficiency, with the average time to process a single image being less than 30 milliseconds. Moreover, even if some image boundaries are blurred, the center of the curve can still be accurately located.

Facial feature point detection under large range of face deformations

Facial Feature Point Detection (FFPD) plays a significant role in several face analysis tasks such as feature extraction and classification. This paper presents a Fully Automatic FFPD system using the application of Random Forest Regression Voting in a Constrained Local Model (RFRV-CLM) framework. A global detector is used to find the approximate positions of the facial region and eye centers. A sequence of local RFRV-CLMs are used to locate a detailed set of points around the facial features. Both global and local models use Random Forest Regression to vote for optimal positions. The system is evaluated in the task of facial expression localization using five different facial expression databases of different characteristics including age, intensity, 6-basic expressions, 22 compound expressions, static and dynamic images, and deliberate and spontaneous expressions. Quantitative results of the evaluation of automatic point localization against manual points (ground truth) demonstrated that the results of the proposed approach are encouraging and outperform the results of alternative techniques tested on the same databases.

Multiscale Seismic Monitoring in the Bedretto Underground Laboratory for Geosciences and Geoenergies (BULGG)

Abstract The Bedretto Underground Laboratory for Geoenergies and Geosciences (BULGG) is located in south-central Switzerland in the middle of a 5.2-km-long tunnel, which connects the Bedretto valley to the Furka railway tunnel. BULGG is a multidisciplinary laboratory that facilitates experiments and research across various fields. From a seismological perspective, a dense seismic network is deployed that allows real-time monitoring of both natural and induced seismicity occurring in the tunnel and the surroundings. In addition, a multilevel monitoring approach during experiments leads to the generation of real-time high-resolution earthquake catalogs issuing event-based alerts and is the input for a simple traffic light system (magnitude and/or ground-motion based), which provides essential information for the advanced traffic-light system (probabilistic approach). We have set up two separate real-time monitoring systems that monitor the background seismicity, as well as injection experiments, with both systems built on the SeisComP framework. The background monitoring, serving as the backbone network, includes broadband sensors at the surface and along the tunnel, as well as strong-motion sensors and high-frequency geophones along the tunnel and in boreholes. The sampling rate is divergent and depends on sensor type and proximity to faults (200–2000 Hz). Acoustic emission sensors and high-frequency accelerometers sampled at 200 kHz constitute the experimental setup that locates in multiple experimental volumes, which include fluid injections, extractions, and tunneling activities. All sensors transmit real-time data to a common server (SeedLink), which serves multiple clients for processing, real-time visualization, archiving via SeisComP, and risk control via dedicated software. A standardized workflow is applied to both background and experimental monitoring, encompassing automatic picking, automatic phase association and location, and magnitude estimation. Advanced methods are implemented in real time that include double-difference relocation and earthquake detection based on waveform cross correlation. BULGG provides a unique environment to implement novel methods in observational and network seismology across scales.

Fault localization by abstract interpretation and its applications

Fault localization aims to automatically identify the cause of an error in a program by localizing the error to a relatively small part of the program. In this paper, we present a novel technique for automated fault localization via error invariants inferred by abstract interpretation. An error invariant for a location in an error program over-approximates the reachable states at the given location that may produce the error, if the execution of the program is continued from that location. Error invariants can be used for statement-wise semantic slicing of error programs and for obtaining concise error explanations. We use an iterative refinement sequence of backward–forward static analyses by abstract interpretation to compute error invariants, which are designed to explain why an error program violates a particular assertion.Furthermore, we present a practical application of the fault localization technique for automatic repair of programs. Given an erroneous program, we first use the fault localization to automatically identify statements relevant for the error, and then repeatedly mutate the expressions in those relevant statements until a correct program that satisfies all assertions is found. All other statements classified by the fault localization as irrelevant for the error are not mutated in the program repair process. This way, we significantly reduce the search space of mutated programs without losing any potentially correct program, and so locate a repaired program much faster than a program repair without fault localization.We have developed a prototype tool for automatic fault localization and repair of C programs. We demonstrate the effectiveness of our approach to localize errors in realistic C programs, and to subsequently repair them. Moreover, we show that our approach based on combining fault localization and code mutations is significantly faster that the previous program repair approach without fault localization.

Breast Cancer Detection and Localizing the Mass Area Using Deep Learning

Breast cancer presents a substantial health obstacle since it is the most widespread invasive cancer and the second most common cause of death in women. Prompt identification is essential for effective intervention, rendering breast cancer screening a critical component of healthcare. Although mammography is frequently employed for screening purposes, the manual diagnosis performed by pathologists can be laborious and susceptible to mistakes. Regrettably, the majority of research prioritizes mass classification over mass localization, resulting in an uneven distribution of attention. In response to this problem, we suggest a groundbreaking approach that seeks to identify and pinpoint cancers in breast mammography pictures. This will allow medical experts to identify tumors more quickly and with greater precision. This paper presents a complex deep convolutional neural network design that incorporates advanced deep learning techniques such as U-Net and YOLO. The objective is to enable automatic detection and localization of breast lesions in mammography pictures. To assess the effectiveness of our model, we carried out a thorough review that included a range of performance criteria. We specifically evaluated the accuracy, precision, recall, F1-score, ROC curve, and R-squared error using the publicly available MIAS dataset. Our model performed exceptionally well, with an accuracy rate of 93.0% and an AUC (area under the curve) of 98.6% for the detection job. Moreover, for the localization task, our model achieved a remarkably high R-squared value of 97%. These findings highlight that deep learning can boost the efficiency and accuracy of diagnosing breast cancer. The automation of breast lesion detection and classification offered by our proposed method bears substantial benefits. By alleviating the workload burden on pathologists, it facilitates expedited and accurate breast cancer screening processes. As a result, the proposed approach holds promise for improving healthcare outcomes and bolstering the overall effectiveness of breast cancer detection and diagnosis.

Automatic localization of cochlear implant electrodes using cone beam computed tomography images

BackgroundCochlear implants (CI) are implantable medical devices that enable the perception of sounds and the understanding of speech by electrically stimulating the auditory nerve in case of inner ear damage. The stimulation takes place via an array of electrodes surgically inserted in the cochlea. After CI implantation, cone beam computed tomography (CBCT) is used to evaluate the position of the electrodes. Moreover, CBCT is used in research studies to investigate the relationship between the position of the electrodes and the hearing outcome of CI user. In clinical routine, the estimation of the position of the CI electrodes is done manually, which is very time-consuming.ResultsThe aim of this study was to optimize procedures of automatic electrode localization from CBCT data following CI implantation. For this, we analyzed the performance of automatic electrode localization for 150 CBCT data sets of 10 different types of electrode arrays. Our own implementation of the method by Noble and Dawant (Lecture notes in computer science (Including subseries lecture notes in artificial intelligence and lecture notes in bioinformatics), Springer, pp 152–159, 2015. https://doi.org/10.1007/978-3-319-24571-3_19) for automated electrode localization served as a benchmark for evaluation. Differences in the detection rate and the localization accuracy across types of electrode arrays were evaluated and errors were classified. Based on this analysis, we developed a strategy to optimize procedures of automatic electrode localization. It was shown that particularly distantly spaced electrodes in combination with a deep insertion can lead to apical–basal confusions in the localization procedure. This confusion prevents electrodes from being detected or assigned correctly, leading to a deterioration in localization accuracy.ConclusionsWe propose an extended cost function for automatic electrode localization methods that prevents double detection of electrodes to avoid apical–basal confusions. This significantly increased the detection rate by 11.15 percent points and improved the overall localization accuracy by 0.53 mm (1.75 voxels). In comparison to other methods, our proposed cost function does not require any prior knowledge about the individual cochlea anatomy.