Instance-Based Segmentation for Boundary Detection of Neuropathic Ulcers Through Mask-RCNN

<title>Abstract</title> <bold>Objectives</bold> Tooth numbering plays a significant role in dental care. The aim of this study was to evaluate the Mask Region-based Convolutional Neural Network (Mask R-CNN) model for tooth number detection, classification, and segmentation from smartphone photographs. <bold>Materials and Methods</bold> This retrospective study utilized the data collected in Finland during 2022. A total of 448 individuals, aged 13–78 years, participated in this study. Participants uploaded dental photographs taken by smartphone. The dataset consisted of upper and lower occlusal, right and left lateral and frontal views. Teeth were labelled using polygon annotation for the FDI World Dental Federation (FDI) notation. The final number of images and tooth polygons were 1,272 and 14,736 respectively. The Mask R-CNN model was used for tooth number recognition. Performance of trained models was measured based on the bounding box and class prediction for 28 tooth numbers, using PASCAL performance metric’s open toolset. The mean average precision (mAP) was calculated for each class with an intersection over union (IoU) threshold of 0.5. External validation was conducted using open Mendeley dataset from India, consisting of 40 preliminary selected images. <bold>Results</bold> The tooth number recognition model demonstrated an overall mAP of 88.2%, with a standard deviation of 0.0075 on 10-fold cross validation, which indicates stable performance of the model. The external validation showed excellent performance of the model, with a mAP of 84.1%. <bold>Conclusions</bold> The Mask R-CNN model developed in this study demonstrated excellent performance and good generalizability. Clinical Relevance: The developed mHealth system facilitates remote dental screening in real-world community settings.

Latest advances in artificial intelligence, particularly in object recognition and segmentation, provide unprecedented opportunities for precision agriculture. This work investigates the use of state-of-the-art AI models, namely Meta’s Segment Anything (SAM) and GroundingDino, for the task of grape cluster detection in vineyards. Three different methods aimed at enhancing the instance segmentation process are proposed: (i) SAM-Refine (SAM-R), which refines a previously proposed depth-based clustering approach, referred to as DepthSeg, using SAM; (ii) SAM-Segmentation (SAM-S), which integrates SAM with a pre-trained semantic segmentation model to improve cluster separation; and (iii) AutoSAM-Dino (ASD), which eliminates the need for manual labeling and transfer learning through the combined use of GroundingDino and SAM. Analysis is conducted on both the object counting and pixel-level segmentation accuracy against a manually labeled ground truth. Metrics such as mean Average Precision (mAP), Intersection over Union (IoU), and precision and recall are calculated to assess the system performance. Compared to the original DepthSeg algorithm, SAM-R slightly advances object counting (mAP: +0.5%) and excels in pixel-level segmentation (IoU: +17.0%). SAM-S, despite a mAP decrease, improves segmentation accuracy (IoU: +13.9%, Precision: +9.2%, Recall: +11.7%). Similarly, ASD, although with a lower mAP, shows significant accuracy enhancement (IoU: +7.8%, Precision: +4.2%, Recall: +4.9%). Additionally, from a labor effort point of view, instance segmentation techniques require much less time for training than manual labeling.

Object detection remains a cornerstone task in computer vision, with wide-ranging applications in autonomous driving, surveillance, and medical imaging. However, traditional methods rely heavily on large annotated datasets, limiting their adaptability in low-resource environments. We propose Hybrid Masked Autoencoder-Detection Transformer++ (HybridMAE-DETR++), a novel self-supervised object detection framework that synergizes Masked Autoencoders (MAEs) and DINOv2 for Vision Transformer (ViT) pretraining. Integrated with a Swin-ViT hybrid backbone and an enhanced DETR++ detection head, the framework significantly reduces dependence on annotated data while improving detection accuracy for small and occluded objects. Evaluated on COCO 2017 and Cityscapes, HybridMAE-DETR++ achieves 47.5% mean Average Precision (mAP) and 68.0% Intersection over Union (IoU) on COCO, and 53.2% mAP and 72.1% IoU on Cityscapes, outperforming DETR and other transformer-based baselines. Ablation and sensitivity analyses confirm the robustness of our hybrid pretraining strategy, and visualizations using Layer-weighted Class Activation Mapping (LayerCAM) and Gradient-weighted Class Activation Mapping++ (Grad-CAM++) validate model interpretability. Despite a moderate increase in training time, the precision gains justify the computational cost. This framework sets a new benchmark for label-efficient, interpretable object detection in real-world scenarios.

Echocardiography can conveniently, rapidly, and economically evaluate the structure and function of the heart, and has important value in the diagnosis and evaluation of cardiovascular diseases (CVDs). However, echocardiography still exhibits significant variability in image acquisition and diagnosis, with a heavy dependency on the operator's experience. Image quality affects disease diagnosis in the later stage, and even image quality assessment still has variability in human evaluation. This study aimed to develop an automated and real-time quality assessment system using deep learning (DL) techniques while decreasing the measurement error of left ventricular ejection fraction (LVEF). This study involved over 5,000 echocardiography datasets from 2,461 participants across 10 medical centers in China to build the model. A 5-point quality scoring system was used to assess the integrity, clarity, and alignment of anatomical structures in each echocardiogram view. Additionally, an innovative DL model was developed to autonomously detect these essential cardiac anatomical structures in real-time, subsequently providing quality score estimations and LVEF. A total of 175 participants from two distinct external medical centers were enrolled for model validation. This dataset was employed to assess the consistency and repeatability of quality score and ejection fraction (EF) measurements, and the assessments made by human experts were compared with those of our model. The developed model demonstrated exceptional performance, achieving Intersection over Union (IoU) scores exceeding 0.8 for left ventricular (LV) segmentation, a mean average precision when IoU >0.5 (mAP50) of 0.91 for cardiac anatomical structures detection, and a 0.96±0.05 accuracy in view classification. The quality scores assessed by the model closely matched those of human experts, indicating strong agreement. The weighted average precision and weighted average recall scores fell within the range of 0.5 to 0.6. Notably, there was no statistically significant difference in LVEF assessments between human experts and our model (P=0.09), as demonstrated by an intraclass correlation coefficient (ICC) analysis of 0.821, reflecting high-level consistency. When assessing echocardiograms with high-quality scores, the model demonstrated a significantly closer alignment and a higher correlation coefficient with human experts (R=0.90±0.04). This study demonstrates that artificial intelligence-assisted echocardiography scoring system aligns well with manual quality scoring. Through the supervision of real-time echocardiogram quality, the artificial intelligence model can assist doctors in providing more reproducible and consistent assessments of cardiac function.

Accurate detection and segmentation of greenhouse-grown capsicum is essential for advancing automation in harvesting and improving yield estimation. This study presents a comparative investigation of two deep learning approaches: the single-shot You Only Look Once version twelve (YOLOv12), representing the latest generation of object detection architectures, and the zero-shot You Only Look Once Extended (YOLOE), a vision-language model capable of prompt-based recognition without prior task-specific training. A dedicated dataset of high-resolution red, green, and blue (RGB) images was collected from a controlled greenhouse environment, manually annotated for both bounding boxes and pixel-level segmentation masks, and used for performance evaluation. The assessment was conducted using mean average precision (mAP), intersection over union (IoU), and average inference time per image. Experimental findings reveal that the single-shot YOLOv12 consistently achieves higher detection accuracy and produces more precise segmentation masks, particularly under challenging conditions such as partial occlusion and varying illumination. Conversely, the zero-shot YOLOE demonstrates greater adaptability for detecting previously unseen variants of capsicum without retraining, albeit with slightly lower segmentation precision. These outcomes highlight the potential of combining high-accuracy retrained models with adaptable zero-shot techniques to develop efficient, scalable, and low-maintenance solutions for autonomous greenhouse harvesting systems. Received: February 28, 2026Revised: March 23, 2026Accepted: March 31, 2026

BackgroundAcupuncture is a widely practiced traditional therapy, yet safety concerns, particularly needle breakage and retention, remain critical issues that can lead to complications such as infections, organ injury, or chronic pain. This study aimed to develop a deep learning model to monitor acupuncture needle insertion, detect instances of needle breakage, and prevent needle retention, ultimately improving patient safety and treatment outcomes.MethodsA deep learning model based on the YOLOv8 architecture was trained using a dataset comprising 192 images from a commercial image library and 73 clinical images captured during real-world acupuncture sessions. Images were preprocessed through cropping and annotation, and augmented to enhance model generalizability. Five-fold cross-validation was employed to ensure robust performance. Model evaluation metrics included precision, recall, F1 score, and mean average precision (mAP) at Intersection over Union (IoU) thresholds of 50% (mAP@50) and 50–95% (mAP@50–95).ResultsThe model demonstrated strong performance, achieving an average precision of 88.0% and a recall of 82.9%. The mean average precision was 88.6% at mAP@50 and 62.9% at mAP@50–95, indicating high reliability in detecting acupuncture needles across diverse scenarios. These results highlight the potential of the model to enhance clinical safety by minimizing risks associated with needle breakage and retention, regardless of practitioner experience or patient demographics.ConclusionsThe proposed YOLOv8-based deep learning model offers a reliable method for real-time needle monitoring in acupuncture. Its integration into clinical workflows can improve safety and efficiency, especially in underserved regions or settings with less experienced practitioners. Future research should validate the model with larger, more diverse datasets and explore its application in various healthcare settings.Trial registrationNot applicable; this study did not involve a healthcare intervention requiring registration. Data collection adhered to ethical standards with institutional approval (TCHIRB-11310004).

Segmentation of trabecular bone microdamage in Xray microCT images using a two-step deep learning method

The purpose of this study is to evaluate and compare the performance of modern deep learning techniques for automatically recognizing and segmenting multiple organ regions on 3D CT images. CT image segmentation is one of the important task in medical image analysis and is still very challenging. Deep learning approaches have demonstrated the capability of scene recognition and semantic segmentation on nature images and have been used to address segmentation problems of medical images. Although several works showed promising results of CT image segmentation by using deep learning approaches, there is no comprehensive evaluation of segmentation performance of the deep learning on segmenting multiple organs on different portions of CT scans. In this paper, we evaluated and compared the segmentation performance of two different deep learning approaches that used 2D- and 3D deep convolutional neural networks (CNN) without- and with a pre-processing step. A conventional approach that presents the state-of-the-art performance of CT image segmentation without deep learning was also used for comparison. A dataset that includes 240 CT images scanned on different portions of human bodies was used for performance evaluation. The maximum number of 17 types of organ regions in each CT scan were segmented automatically and compared to the human annotations by using ratio of intersection over union (IU) as the criterion. The experimental results demonstrated the IUs of the segmentation results had a mean value of 79% and 67% by averaging 17 types of organs that segmented by a 3D- and 2D deep CNN, respectively. All the results of the deep learning approaches showed a better accuracy and robustness than the conventional segmentation method that used probabilistic atlas and graph-cut methods. The effectiveness and the usefulness of deep learning approaches were demonstrated for solving multiple organs segmentation problem on 3D CT images.

The feather coverage of a laying hen is an important indicator of both its productivity and welfare. Conventional manual feather scoring procedures are laborious, subjective, and stressful for the hens. Thermography offers a modern alternative to addressing these problems. Thermal cameras capture radiative heat loss, which is comparatively greater Classification from featherless areas. Studies have been conducted to establish a standard temperature range that correlates to specific featherless areas. However, such temperature-based approaches have been inconsistent with each other. In contrast, this study used deep learning techniques to automatically assess dorsal feather scores using thermal images. Thermal images (n = 1575) of the dorsal body of cage-free laying hens with varying degrees of feather damage were captured. Manual feather scoring was performed, classifying the image into a feather score (0–2) according to the increasing severity of feather loss. A total of 1222 images were selected, filtering out images of lower quality. Two types of computer vision models, a classification model and an object detection model, were trained and evaluated. A custom convolutional neural network (CNN) was trained to classify thermal images into feather score categories. Additionally, we trained and optimized You Only Look Once (YOLO) object detection models to detect areas of feather damage and predict the feather score. The CNN model achieved an overall accuracy of 0.81, with high precision for severe feather loss. The YOLO-based object detection model was optimum using YOLO11n, which achieved a precision of 0.81, a recall of 0.73 and a mean average precision (mAP) at 0.5 intersection over union (IoU) of 0.84. Results show the potential of combining thermal imaging with deep learning techniques to perform objective, automatic, and scalable feather scoring procedures. Future studies should focus on data diversity, multiple part scoring, and semantic segmentation for robust performance.

Timber transportation surveillance is the key and important means of forest resource management. In order to surveille timber transportation in the forest road, we carry out timber transportation vehicle detection from the video of bayonet, with which we propose an improvement of SSD (Single Shot Multibox Detector) based on MXNet. We establish a timber transportation vehicle dataset for training and testing the ResNet50 and Darknet53 with three thresholds. As mAP (mean Average Precision) is closely related to IoU (Intersection over Union), the original calculation of which does not consider overlap and disjoint between prediction box and the ground truth, we find it contributes the major part of error-detection with SSD. Inspired by this, we combine a novel metric loss calculation called GIoU (Generalized Intersection over the Union) with SSD to improve mAP, which is proved to be greatly promoted in timber transportation vehicle detection.

U-Net and YOLOv8 Artificial Intelligence Models for Automated Recognition of Internal Jugular Veins and Radial Arteries: A Foundational Study for Artificial Intelligence-guided Vascular Cannulation inPoint-of-care Ultrasound.

Artificial intelligence has garnered significant attention in recent years as a rapidly advancing field of computer technology. With the continual advancement of computer hardware, deep learning has made breakthrough developments within the realm of artificial intelligence. Over the past few years, applying deep learning architecture in medicine and industrial anomaly inspection has significantly contributed to solving numerous challenges related to efficiency and accuracy. For excellent results in radiological, pathological, endoscopic, ultrasonic, and biochemical examinations, this paper utilizes deep learning combined with image processing to identify spinal canal and vertebral foramen dimensions. In existing research, technologies such as corrosion and expansion in magnetic resonance image (MRI) processing have also strengthened the accuracy of results. Indicators such as area and Intersection over Union (IoU) are also provided for assessment. Among them, the mean Average Precision (mAP) for identifying intervertebral foramen (IVF) and intervertebral disc (IVD) through YOLOv4 is 95.6%. Resnet50 mixing U-Net was employed to identify the spinal canal and intervertebral foramen and achieved IoU scores of 79.11% and 80.89%.

Multiple myeloma is a condition of cancer in the bone marrow that can lead to dysfunction of the body and fatal expression in the patient. Manual microscopic analysis of abnormal plasma cells, also known as multiple myeloma cells, is one of the most commonly used diagnostic methods for multiple myeloma. However, as it is a manual process, it consumes too much effort and time. Besides, it has a higher chance of human errors. This paper presents a computer-aided detection and segmentation of myeloma cells from microscopic images of the bone marrow aspiration. Two major contributions are presented in this paper. First, different Mask R-CNN models using different images, including original microscopic images, contrast-enhanced images and stained cell images, are developed to perform instance segmentation of multiple myeloma cells. As a second contribution, a deep-wise augmentation, a deep learning-based data augmentation method, is applied to increase the performance of Mask R-CNN models. Based on the experimental findings, the Mask R-CNN model using contrast-enhanced images combined with the proposed deep-wise data augmentation provides a superior performance compared to other models. It achieves a mean precision of 0.9973, mean recall of 0.8631, and mean intersection over union (IOU) of 0.9062.

Among the artificial intelligence based studies conducted in the field of agriculture, disease recognition methods founded on deep learning are observed to become widespread. Due to the diversity and regional specificity of many plant species, studies performed in this field are not at the desired level. Olive peacock spot disease of the olive plant which grows only in certain regions in the world is a widely encountered disease particularly in Turkey. The aim of this research is to develop an olive peacock spot disease detection system using a Single Shot Detector (SSD) which is one the popular deep learning architectures to support olive farmers. This study presents a data set consisting of 1460 olive leaves samples for the detection of olive peacock spot disease. All of the images of the olive leaves which produced under controlled conditions were collected from Aegean region of Turkey during spring and summer. The data set was trained with different intersection over union (IoU) threshold values using SSD architecture. A 96 % average precision (AP) value was obtained with IoU=0.5. As IOU value goes up from 0.5, erroneously classified olive peacock spot disease symptoms growed larger as well. The AP curve becomes flat when between 0.1 and 0.5, and it decreases when greater than 0.5. This analysis showed that the IoU significantly influenced the performance of SSD based model in detection of olive peacock spot disease. In addition to, trainings were performed by employing Pytorch library and a GUI was developed for the SSD based application using PyQt5 which is one of Pyhton's libraries. Results showed that the SSD was a robust tool for recognizing the olive peacock spot disease.

Bias in medical image segmentation can lead to unequal performance across demographic subgroups, raising concerns about fairness and reliability in clinical AI systems. While deep learning models have achieved high segmentation accuracy, ensuring equitable performance across race and gender remains a significant challenge, particularly in privacy-sensitive healthcare environments. This study investigates fairness-aware medical image segmentation for hip and knee radiographs using deep learning models evaluated in both centralized and Federated Learning (FL) settings. We introduce Curriculum Learning (CL) strategies and Progressive Loss (PL) functions to regulate sample difficulty during training. In addition, we propose two novel fairness-oriented federated learning algorithms, Federated Intersection over Union (FedIoU) and Federated Intersection over Union with Outlier Analysis (FedIoUoutlier). Experiments are conducted using multiple segmentation backbones and simulated multi-site data partitions derived from the Osteoarthritis Initiative dataset. Model performance is evaluated using Intersection over Union (IoU), IoU standard deviation, Skewed Error Ratio (SER), and Min-Max Disparity across race and gender subgroups. Statistical significance was verified using paired t-tests to compare per-sample IoU performance against baseline configurations. Across both hip and knee segmentation tasks, curriculum learning and progressive loss strategies consistently improved segmentation accuracy and reduced demographic performance disparities in centralized training. In federated settings, fairness-aware aggregation further enhanced performance. Notably, FedIoUoutlier combined with balanced curriculum learning and tiered progressive loss achieved the highest mean IoU while yielding the lowest SER and Min-Max Disparity, indicating improved fairness without sacrificing accuracy. In several configurations, federated models matched or exceeded the performance of optimized centralized models, with statistically significant improvements in per-sample IoU over baseline configurations. The results demonstrate that structured training strategies and fairness-aware federated aggregation can jointly improve accuracy, stability, and demographic fairness in medical image segmentation. By integrating curriculum learning, progressive loss, and novel FL algorithms, this work provides a practical pathway toward equitable and privacy-preserving AI systems for medical imaging.