-
Journal of Dental Research Aug 2021Accurate segmentation of the jaw (i.e., mandible and maxilla) and the teeth in cone beam computed tomography (CBCT) scans is essential for orthodontic diagnosis and...
Accurate segmentation of the jaw (i.e., mandible and maxilla) and the teeth in cone beam computed tomography (CBCT) scans is essential for orthodontic diagnosis and treatment planning. Although various (semi)automated methods have been proposed to segment the jaw or the teeth, there is still a lack of fully automated segmentation methods that can simultaneously segment both anatomic structures in CBCT scans (i.e., multiclass segmentation). In this study, we aimed to train and validate a mixed-scale dense (MS-D) convolutional neural network for multiclass segmentation of the jaw, the teeth, and the background in CBCT scans. Thirty CBCT scans were obtained from patients who had undergone orthodontic treatment. Gold standard segmentation labels were manually created by 4 dentists. As a benchmark, we also evaluated MS-D networks that segmented the jaw or the teeth (i.e., binary segmentation). All segmented CBCT scans were converted to virtual 3-dimensional (3D) models. The segmentation performance of all trained MS-D networks was assessed by the Dice similarity coefficient and surface deviation. The CBCT scans segmented by the MS-D network demonstrated a large overlap with the gold standard segmentations (Dice similarity coefficient: 0.934 ± 0.019, jaw; 0.945 ± 0.021, teeth). The MS-D network-based 3D models of the jaw and the teeth showed minor surface deviations when compared with the corresponding gold standard 3D models (0.390 ± 0.093 mm, jaw; 0.204 ± 0.061 mm, teeth). The MS-D network took approximately 25 s to segment 1 CBCT scan, whereas manual segmentation took about 5 h. This study showed that multiclass segmentation of jaw and teeth was accurate and its performance was comparable to binary segmentation. The MS-D network trained for multiclass segmentation would therefore make patient-specific orthodontic treatment more feasible by strongly reducing the time required to segment multiple anatomic structures in CBCT scans.
Topics: Cone-Beam Computed Tomography; Deep Learning; Humans; Image Processing, Computer-Assisted; Neural Networks, Computer; Orthodontics; Spiral Cone-Beam Computed Tomography
PubMed: 33783247
DOI: 10.1177/00220345211005338 -
The International Journal of... 2017The vertebrate head characteristically exhibits a complex pattern with sense organs, brain, paired eyes and jaw muscles, and the brain case is not found in other... (Review)
Review
The vertebrate head characteristically exhibits a complex pattern with sense organs, brain, paired eyes and jaw muscles, and the brain case is not found in other chordates. How the extant vertebrate head has evolved remains enigmatic. Historically, there have been two conflicting views on the origin of the vertebrate head, segmental and non-segmental views. According to the segmentalists, the vertebrate head is organized as a metameric structure composed of segments equivalent to those in the trunk; a metamere in the vertebrate head was assumed to consist of a somite, a branchial arch and a set of cranial nerves, considering that the head evolved from rostral segments of amphioxus-like ancestral vertebrates. Non-segmentalists, however, considered that the vertebrate head was not segmental. In that case, the ancestral state of the vertebrate head may be non-segmented, and rostral segments in amphioxus might have been secondarily gained, or extant vertebrates might have evolved through radical modifications of amphioxus-like ancestral vertebrate head. Comparative studies of mesodermal development in amphioxus and vertebrate gastrula embryos have revealed that mesodermal gene expressions become segregated into two domains anteroposteriorly to specify the head mesoderm and trunk mesoderm only in vertebrates; in this segregation, key genes such as delta and hairy, involved in segment formation, are expressed in the trunk mesoderm, but not in the head mesoderm, strongly suggesting that the head mesoderm of extant vertebrates is not segmented. Taken together, the above finding possibly adds a new insight into the origin of the vertebrate head; the vertebrate head mesoderm would have evolved through an anteroposterior polarization of the paraxial mesoderm if the ancestral vertebrate had been amphioxus-like.
Topics: Animals; Body Patterning; Cephalochordata; Gene Expression Regulation, Developmental; Head; Lancelets; Models, Biological; Somites; Vertebrates
PubMed: 29319111
DOI: 10.1387/ijdb.170121to -
BMC Oral Health Sep 2023Intra-oral scans and gypsum cast scans (OS) are widely used in orthodontics, prosthetics, implantology, and orthognathic surgery to plan patient-specific treatments,...
OBJECTIVE
Intra-oral scans and gypsum cast scans (OS) are widely used in orthodontics, prosthetics, implantology, and orthognathic surgery to plan patient-specific treatments, which require teeth segmentations with high accuracy and resolution. Manual teeth segmentation, the gold standard up until now, is time-consuming, tedious, and observer-dependent. This study aims to develop an automated teeth segmentation and labeling system using deep learning.
MATERIAL AND METHODS
As a reference, 1750 OS were manually segmented and labeled. A deep-learning approach based on PointCNN and 3D U-net in combination with a rule-based heuristic algorithm and a combinatorial search algorithm was trained and validated on 1400 OS. Subsequently, the trained algorithm was applied to a test set consisting of 350 OS. The intersection over union (IoU), as a measure of accuracy, was calculated to quantify the degree of similarity between the annotated ground truth and the model predictions.
RESULTS
The model achieved accurate teeth segmentations with a mean IoU score of 0.915. The FDI labels of the teeth were predicted with a mean accuracy of 0.894. The optical inspection showed excellent position agreements between the automatically and manually segmented teeth components. Minor flaws were mostly seen at the edges.
CONCLUSION
The proposed method forms a promising foundation for time-effective and observer-independent teeth segmentation and labeling on intra-oral scans.
CLINICAL SIGNIFICANCE
Deep learning may assist clinicians in virtual treatment planning in orthodontics, prosthetics, implantology, and orthognathic surgery. The impact of using such models in clinical practice should be explored.
Topics: Humans; Deep Learning; Algorithms; Calcium Sulfate; Dental Care; Physical Examination
PubMed: 37670290
DOI: 10.1186/s12903-023-03362-8 -
Methods in Molecular Biology (Clifton,... 2022Reverse genetics systems provide a powerful tool to generate recombinant arenavirus expressing reporters to facilitate the investigation of the arenavirus life cycle and...
Reverse genetics systems provide a powerful tool to generate recombinant arenavirus expressing reporters to facilitate the investigation of the arenavirus life cycle and also for the discovery of antiviral countermeasures. The plasmid-encoded viral ribonucleoprotein components initiate the transcription and replication of a plasmid-driven full-length viral genome, resulting in infectious virus. Thereby, this approach is ideal for the generation of recombinant arenaviruses expressing reporter genes that can be used as valid surrogates for virus replication. By splitting the small viral segment (S) into two viral segments (S1 and S2), each of them encoding a reporter gene, recombinant tri-segmented arenavirus can be rescued. Bi-reporter-expressing recombinant tri-segmented arenaviruses represent an excellent tool to study the biology of arenaviruses, including the identification and characterization of both prophylactic and therapeutic countermeasures for the treatment of arenaviral infections. In this chapter, we describe a detailed protocol on the generation and in vitro characterization of recombinant arenaviruses containing a tri-segment genome expressing two reporter genes based on the prototype member in the family, lymphocytic choriomeningitis virus (LCMV). Similar experimental approaches can be used for the generation of bi-reporter-expressing tri-segment recombinant viruses for other members in the arenavirus family.
Topics: Arenaviridae Infections; Genes, Reporter; Humans; Lymphocytic choriomeningitis virus; Reverse Genetics; Virus Replication
PubMed: 35821475
DOI: 10.1007/978-1-0716-2453-1_17 -
F1000Research 2018Studies of the vertebrate hindbrain have revealed parallel mechanisms that establish sharp segments with a distinct and homogeneous regional identity. Recent work has... (Review)
Review
Studies of the vertebrate hindbrain have revealed parallel mechanisms that establish sharp segments with a distinct and homogeneous regional identity. Recent work has revealed roles of cell identity regulation and its relationships with cell segregation. At early stages, there is overlapping expression at segment borders of the Egr2 and Hoxb1 transcription factors that specify distinct identities, which is resolved by reciprocal repression. Computer simulations show that this dynamic regulation of cell identity synergises with cell segregation to generate sharp borders. Some intermingling between segments occurs at early stages, and ectopic egr2-expressing cells switch identity to match their new neighbours. This switching is mediated by coupling between egr2 expression and the level of retinoic acid signalling, which acts in a community effect to maintain homogeneous segmental identity. These findings reveal an interplay between cell segregation and the dynamic regulation of cell identity in the formation of sharp patterns in the hindbrain and raise the question of whether similar mechanisms occur in other tissues.
Topics: Animals; Cell Separation; Humans; Rhombencephalon
PubMed: 30135723
DOI: 10.12688/f1000research.15391.1 -
BMC Medical Imaging Jul 2022Recent developments to segment and characterize the regions of interest (ROI) within medical images have led to promising shape analysis studies. However, the procedures...
BACKGROUND
Recent developments to segment and characterize the regions of interest (ROI) within medical images have led to promising shape analysis studies. However, the procedures to analyze the ROI are arbitrary and vary by study. A tool to translate the ROI to analyzable shape representations and features is greatly needed.
RESULTS
We developed SAFARI (shape analysis for AI-segmented images), an open-source R package with a user-friendly online tool kit for ROI labelling and shape feature extraction of segmented maps, provided by AI-algorithms or manual segmentation. We demonstrated that half of the shape features extracted by SAFARI were significantly associated with survival outcomes in a case study on 143 consecutive patients with stage I-IV lung cancer and another case study on 61 glioblastoma patients.
CONCLUSIONS
SAFARI is an efficient and easy-to-use toolkit for segmenting and analyzing ROI in medical images. It can be downloaded from the comprehensive R archive network (CRAN) and accessed at https://lce.biohpc.swmed.edu/safari/ .
Topics: Algorithms; Artificial Intelligence; Glioblastoma; Humans; Image Processing, Computer-Assisted
PubMed: 35869424
DOI: 10.1186/s12880-022-00849-8 -
The American Journal of Managed Care Feb 2022Empiric segmentation is a rapidly growing, learning health system approach that uses large health care system data sets to identify groups of high-risk patients who may... (Review)
Review
OBJECTIVES
Empiric segmentation is a rapidly growing, learning health system approach that uses large health care system data sets to identify groups of high-risk patients who may benefit from similar interventions. We aimed to review studies that used data-driven approaches to segment high-risk patient populations and describe how their designs and findings can inform health care leaders who are interested in applying similar techniques to their patient populations.
STUDY DESIGN
Structured literature review.
METHODS
We searched for original research articles published since 2000 that identified high-risk adult patient populations and applied data-driven analyses to segment the population. Two reviewers independently extracted study population source and criteria for high-risk designation, segmentation method, data types included, model selection criteria, and model results from the identified studies.
RESULTS
Our search identified 224 articles, 12 of which met criteria for full review. Of these, 8 segmented high-risk patients and 4 segmented diagnoses without assigning patients to unique groups. Studies segmenting patients more often had clinically interpretable results. Common groups were defined by high prevalence of diabetes, cardiovascular disease, psychiatric conditions including substance use disorders, and neurologic disease (eg, stroke). Few studies incorporated patients' functional or social factors. Resulting patient and diagnosis clusters varied in ways closely linked to the model inputs, patient population inclusion criteria, and health care system context.
CONCLUSIONS
Empiric segmentation can yield clinically relevant groups of patients with complex medical needs. Segmentation results are context dependent, suggesting the need for careful design and interpretation of segmentation models to ensure that results can inform clinical care and program design in the target setting.
Topics: Adult; Delivery of Health Care; Humans; Mental Disorders; Research Design
PubMed: 35139299
DOI: 10.37765/ajmc.2022.88752 -
The Lancet. Digital Health Sep 2022Artificial intelligence (AI) and deep learning have shown great potential in streamlining clinical tasks. However, most studies remain confined to in silico validation... (Observational Study)
Observational Study
BACKGROUND
Artificial intelligence (AI) and deep learning have shown great potential in streamlining clinical tasks. However, most studies remain confined to in silico validation in small internal cohorts, without external validation or data on real-world clinical utility. We developed a strategy for the clinical validation of deep learning models for segmenting primary non-small-cell lung cancer (NSCLC) tumours and involved lymph nodes in CT images, which is a time-intensive step in radiation treatment planning, with large variability among experts.
METHODS
In this observational study, CT images and segmentations were collected from eight internal and external sources from the USA, the Netherlands, Canada, and China, with patients from the Maastro and Harvard-RT1 datasets used for model discovery (segmented by a single expert). Validation consisted of interobserver and intraobserver benchmarking, primary validation, functional validation, and end-user testing on the following datasets: multi-delineation, Harvard-RT1, Harvard-RT2, RTOG-0617, NSCLC-radiogenomics, Lung-PET-CT-Dx, RIDER, and thorax phantom. Primary validation consisted of stepwise testing on increasingly external datasets using measures of overlap including volumetric dice (VD) and surface dice (SD). Functional validation explored dosimetric effect, model failure modes, test-retest stability, and accuracy. End-user testing with eight experts assessed automated segmentations in a simulated clinical setting.
FINDINGS
We included 2208 patients imaged between 2001 and 2015, with 787 patients used for model discovery and 1421 for model validation, including 28 patients for end-user testing. Models showed an improvement over the interobserver benchmark (multi-delineation dataset; VD 0·91 [IQR 0·83-0·92], p=0·0062; SD 0·86 [0·71-0·91], p=0·0005), and were within the intraobserver benchmark. For primary validation, AI performance on internal Harvard-RT1 data (segmented by the same expert who segmented the discovery data) was VD 0·83 (IQR 0·76-0·88) and SD 0·79 (0·68-0·88), within the interobserver benchmark. Performance on internal Harvard-RT2 data segmented by other experts was VD 0·70 (0·56-0·80) and SD 0·50 (0·34-0·71). Performance on RTOG-0617 clinical trial data was VD 0·71 (0·60-0·81) and SD 0·47 (0·35-0·59), with similar results on diagnostic radiology datasets NSCLC-radiogenomics and Lung-PET-CT-Dx. Despite these geometric overlap results, models yielded target volumes with equivalent radiation dose coverage to those of experts. We also found non-significant differences between de novo expert and AI-assisted segmentations. AI assistance led to a 65% reduction in segmentation time (5·4 min; p<0·0001) and a 32% reduction in interobserver variability (SD; p=0·013).
INTERPRETATION
We present a clinical validation strategy for AI models. We found that in silico geometric segmentation metrics might not correlate with clinical utility of the models. Experts' segmentation style and preference might affect model performance.
FUNDING
US National Institutes of Health and EU European Research Council.
Topics: Algorithms; Artificial Intelligence; Carcinoma, Non-Small-Cell Lung; Deep Learning; Humans; Lung Neoplasms; Positron Emission Tomography Computed Tomography; United States
PubMed: 36028289
DOI: 10.1016/S2589-7500(22)00129-7 -
Journal of Dentistry May 2023Quantitative analysis of the volume and shape of the temporomandibular joint (TMJ) using cone-beam computed tomography (CBCT) requires accurate segmentation of the...
OBJECTIVE
Quantitative analysis of the volume and shape of the temporomandibular joint (TMJ) using cone-beam computed tomography (CBCT) requires accurate segmentation of the mandibular condyles and the glenoid fossae. This study aimed to develop and validate an automated segmentation tool based on a deep learning algorithm for accurate 3D reconstruction of the TMJ.
MATERIALS AND METHODS
A three-step deep-learning approach based on a 3D U-net was developed to segment the condyles and glenoid fossae on CBCT datasets. Three 3D U-Nets were utilized for region of interest (ROI) determination, bone segmentation, and TMJ classification. The AI-based algorithm was trained and validated on 154 manually segmented CBCT images. Two independent observers and the AI algorithm segmented the TMJs of a test set of 8 CBCTs. The time required for the segmentation and accuracy metrics (intersection of union, DICE, etc.) was calculated to quantify the degree of similarity between the manual segmentations (ground truth) and the performances of the AI models.
RESULTS
The AI segmentation achieved an intersection over union (IoU) of 0.955 and 0.935 for the condyles and glenoid fossa, respectively. The IoU of the two independent observers for manual condyle segmentation were 0.895 and 0.928, respectively (p<0.05). The mean time required for the AI segmentation was 3.6 s (SD 0.9), whereas the two observers needed 378.9 s (SD 204.9) and 571.6 s (SD 257.4), respectively (p<0.001).
CONCLUSION
The AI-based automated segmentation tool segmented the mandibular condyles and glenoid fossae with high accuracy, speed, and consistency. Potential limited robustness and generalizability are risks that cannot be ruled out, as the algorithms were trained on scans from orthognathic surgery patients derived from just one type of CBCT scanner.
CLINICAL SIGNIFICANCE
The incorporation of the AI-based segmentation tool into diagnostic software could facilitate 3D qualitative and quantitative analysis of TMJs in a clinical setting, particularly for the diagnosis of TMJ disorders and longitudinal follow-up.
Topics: Humans; Deep Learning; Temporomandibular Joint; Mandibular Condyle; Temporomandibular Joint Disorders; Cone-Beam Computed Tomography; Image Processing, Computer-Assisted
PubMed: 36870441
DOI: 10.1016/j.jdent.2023.104475 -
Current Problems in Diagnostic Radiology 2023Hepatosplenomegaly is commonly diagnosed by radiologists based on single dimension measurements and heuristic cut-offs. Volumetric measurements may be more accurate for...
Hepatosplenomegaly is commonly diagnosed by radiologists based on single dimension measurements and heuristic cut-offs. Volumetric measurements may be more accurate for diagnosing organ enlargement. Artificial intelligence techniques may be able to automatically calculate liver and spleen volume and facilitate more accurate diagnosis. After IRB approval, 2 convolutional neural networks (CNN) were developed to automatically segment the liver and spleen on a training dataset comprised of 500 single-phase, contrast-enhanced CT abdomen and pelvis examinations. A separate dataset of ten thousand sequential examinations at a single institution was segmented with these CNNs. Performance was evaluated on a 1% subset and compared with manual segmentations using Sorensen-Dice coefficients and Pearson correlation coefficients. Radiologist reports were reviewed for diagnosis of hepatomegaly and splenomegaly and compared with calculated volumes. Abnormal enlargement was defined as greater than 2 standard deviations above the mean. Median Dice coefficients for liver and spleen segmentation were 0.988 and 0.981, respectively. Pearson correlation coefficients of CNN-derived estimates of organ volume against the gold-standard manual annotation were 0.999 for the liver and spleen (P < 0.001). Average liver volume was 1556.8 ± 498.7 cc and average spleen volume was 194.6 ± 123.0 cc. There were significant differences in average liver and spleen volumes between male and female patients. Thus, the volume thresholds for ground-truth determination of hepatomegaly and splenomegaly were determined separately for each sex. Radiologist classification of hepatomegaly was 65% sensitive, 91% specific, with a positive predictive value (PPV) of 23% and an negative predictive value (NPV) of 98%. Radiologist classification of splenomegaly was 68% sensitive, 97% specific, with a positive predictive value (PPV) of 50% and a negative predictive value (NPV) of 99%. Convolutional neural networks can accurately segment the liver and spleen and may be helpful to improve radiologist accuracy in the diagnosis of hepatomegaly and splenomegaly.
PubMed: 37277270
DOI: 10.1067/j.cpradiol.2023.05.005